The Effectiveness of Health Care Teams in the National Health Service

Report
Carol S. Borrill, Jean Carletta, Angela J. Carter, Jeremy F. Dawson, Simon Garrod, Anne Rees, Ann Richards, David Shapiro and Michael A. West

Aston Centre for Health Service Organization Research, Aston Business School, University of Aston Human Communications Research Centre, Universities of Glasgow and Edinburgh Psychological Therapies Research Centre, University of Leeds

Contents
______________________________________________ __
Key Findings Acknowledgements Health Care Team Effectiveness Project: Summary

Chapter 1

Teamwork, Communication and Effectiveness in Health Care: A Review Primary Health Care Team Research Methods and Sample Details Primary Health Care Team Results from Survey and External Ratings Qualitative Research: Developing Objectives and Effectiveness Measures for Primary Health Care Teams Community Mental Health Teams Research Methods and Sample Details Community Mental Health Teams Results from Survey and External Ratings Community Mental Health Teams Results from Qualitative Research Secondary Health Care Teams Research Methods and Sample Details Secondary Care Teams Ratings Meetings and Communication Research Methods Analysis of Communication in PHCT Teams Analysis of Communication in CHMT's Conclusions and Recommendations

Page 1

Chapter 2

Page 25

Chapter 3

Page 44

Chapter 4

Page 57

Chapter 5

Page 78

Chapter 6

Page 103

Chapter 7

Page 121

Chapter 8

Page 141 Page 157

Chapter 9 Chapter 10

Page 172 Page 182 Page 197 Page 215

Chapter 11 Chapter 12 Chapter 13

Appendix I

Survey Instruments/Rating Measures/Interview Schedules

Appendix II Appendix III

Knowing the way: Effectiveness in Primary Health Care Developing Effectiveness Measures for Primary Health Care Teams

Appendix IV Training Programme Tools and Techniques for Assessing Performance Bibliography

Acknowledgements
________________________________________________________

Liaison Officers:

Liz Meerabeau Sue Longsdate John Wilkinson

Advisory Group Members: Debbie Mellors NHS Executive Sarah Connors NHS Executive Jim Ford NHS Executive Bonnie Sibbald NHS Executive Eileen Robertson NHS Executive Sheila Roberts Department of Health Terry Breugha University of Leicester Anne Netton University of Kent Thelma Sackman NHS Executive Research Team: Dr Carol Borrill Aston Business School Aston University Birmingham Sam Bedlingham City University London Jean Carletta Human Communication Research Centre Edinburgh January 1997 - December 1999

June 1997 - December 1999

January 1997 - December 1999

Christine Carmichael June 1997 - February 1998 Institute of Work Psychology Sheffield University Sheffield Angela Carter January 1998 - December 1999 Institute of Work Psychology Sheffield University Sheffield Jeremy Dawson Aston Business School Aston University Birmingham July 1999 - December 1999

Simon Garrod January 1997 - December 1999 Human Communications Research Centre Glasgow University Glasgow Heidi Frazer-Krauss Medical School Glasgow University Glasgow January 1997 - June 1997

Anne Rees January 1997 - June 1997 Psychological Therapies Research Centre Leeds University Leeds Anne Richards January 1997 - December 1999 Psychological Therapies Research Centre Leeds University Leeds Carein Todd April 1997 - May 1998 Institute of Work Psychology Sheffield University Sheffield David Shapiro April 1997 - May 1998 Psychological Therapies Research Centre Leeds University Leeds Michael West Aston Business School Aston University Birmingham January 1997 - December 1999

David Woods January 1998 - June 1999 Institute of Work Psychology Sheffield University

______________________________________________

Summary
______________________________________________ ________
A primary prescription that policy makers and practitioners have offered for meeting the challenges facing the National Health Service is the development of multidisciplinary team working. The importance of team working in health care has been emphasised in numerous reports and policy documents on the National Health Service. One particularly emphasised the importance of team working if health and social care for people are to be of the highest quality and efficiency:

"The best and most cost-effective outcomes for patients and clients are achieved when professionals work together, learn together, engage in clinical audit of outcomes together, and generate innovation to ensure progress in practice and service."

Over the last thirty years this has proved very difficult to achieve in practice because of the barriers between professional groupings such as doctors and nurses. Other factors such as gender issues also influence team working. For example, G.P.s are predominantly men while the rest of the primary care service population is predominantly women; community mental health psychiatrists are predominantly men, whereas the rest of the population of community mental health teams is predominantly women, and in hospital settings the ranks of consultants continue to be largely made up of men. Other factors which impede the creation of effective

multidisciplinary teams include multiple lines of management, perceived status differentials between different professional groups, and lack of organisational systems and structures for supporting and managing teams.

The Health Care Team Effectiveness Project was commissioned by the Department of Health. The overall aim of the research described here was to determine whether and how multidisciplinary team working contributes to quality, efficiency and innovation in health care in the NHS.

The objectives of the research were to establish:

which team member characteristics such as age, gender, occupational group, experience, qualifications, and team size, influence how well the teams work together;

how team working processes, such as participation, reflexivity, communication, decision-making and leadership contribute to the effectiveness of teams, particularly the quality of health care and the development of innovative practice;

The research programme was carried out over a three year period by a team of researchers based at the universities of Aston, Edinburgh, Glasgow, Leeds and Sheffield. During the course of the study information on team working was gathered from some 400 health care teams. This involved consulting over 7,000 NHS

personnel and a large number of NHS clients. Five national workshops were held with key representatives from primary health care and community health care. A wide range of research methods was used, including questionnaire surveys, telephone interviews, in-depth interviews, observation, focus groups and video and audio tape recordings of meetings

The research was carried out in two stages: quantitative data collection from 100 primary health care teams (PHCTs), 113 community health care teams (CMHTs) and 193 secondary health care teams (SHCTs), and in-depth work with a sub-sample of teams.

Key findings

Effectiveness Quality of teamworking is powerfully related to effectiveness of health care teams: The clearer the team's objectives The higher the level of participation in the team The higher the level of commitment to quality The higher the level of support of innovation . the more effective are health care teams across virtually all domains of functioning

Innovation Quality of teamworking is powerfully related to innovation of health care teams:

The clearer the team's objectives The higher the level of participation in the team The higher the level of commitment to quality The higher the level of support of innovation

.. the more innovative are health care teams across virtually all domains of functioning

Mental Health Those working in teams have much better mental health than those working in looser groups or working individually. The benefits appear to be due to:

Greater role clarity Better peer support

Those working in teams are also buffered from the negative effects of organizational climate and conflict.

The better the functioning of team with respect to Clarity of objectives Levels of participation Commitment to quality Support for innovation

the better the mental health of team members across all domains of health care.

Organisational performance There is a significant and negative relationship between the percentage of staff working in teams and the mortality in these hospitals, taking account of both local health needs and hospital size. Where more employees work in teams the death rate is significantly lower (calculated on the basis of the Sunday Times Mortality

Index, Dr Foster; deaths within 30 days of emergency surgery and deaths after admission for hip fracture)1. Retention and turnover Within health care, those working in well functioning teams are more likely to stay working in their settings than those working in poorly functioning teams. Leadership In Community Mental Health and Primary Health Care, where there is no clear leader/co-ordinator or where there is conflict over leadership team objectives are unclear, and there are. Low levels of participation Low commitment to quality Low support for innovation Poor team member mental health Low levels of effectiveness and innovation Communication Communication, integration and regular meetings in PHC and CMC health care teams are associated with higher levels of effectiveness and innovation, yet the quality of meetings (particularly in Primary Health Care) is often poor. Professional diversity Diversity of professional groups in Primary Health Care is clearly linked to levels of team innovation. In newly formed Community Mental Health Teams, this relationship does not appear. The same findings emerged from research carried out with 85 breast cancer care teams2.

This finding is based on research recently completed by the research team at the Aston Centre for Health Services Organisation Research (further details available from West or Borrill). This finding is based on research recently completed by the research team at the Aston Centre for Health Services Organisational Research (further details available from West or Borrill).

Conclusions

Systematic and revolutionary organizational change is necessary if the positive results of this research are to be implemented in practice.

NHS organizations have to developed as team-based, rather than hierarchical.

Structure, culture, work design, HRM and management have to accommodate and enable rather than impede team-based working.

NHS employees should be trained in the KSAs for working in teams.

NHS managers should be trained to manage team-based organizations.

Chapter 1
Teamwork, Communication and Effectiveness in Health Care: A Review
The challenges of organising health care in the modern United Kingdom context are considerable. There are continual improvements in medical technologies, greater levels of knowledge and awareness amongst patient populations and increasing demands for the variety of sources of health care available within the National Health Service. The provision of free health care at the point of delivery to the population has become one of the most important issues in the national political agenda in the early part of the twenty-first century. At the same time the National Health Service has become a massively complex institution characterised by large organisations, repeated restructurings, and subject to a wide range of political and economic pressures. The response of the government has been to promise a huge increase in spending on the NHS; a key question to be answered in relation to this political agenda is how can we organise health care and achieve good, fair and cost effective services for the whole population. This report focuses on determining whether, and if so, how teamworking can help.

In this first chapter we review the research evidence about the potential benefits of teamworking and the factors that influence the effectiveness of teams, focusing particularly upon their use in health care settings. We draw on empirical evidence from research conducted in the United Kingdom, mainland Europe, North America and Australia. The literature on team composition and the processes which influence team performance is briefly reviewed with particular emphasis on communication, decision-making and problem-solving. We then explore the influences of

organisational context and leadership, before presenting the theoretical model which guided the research programme described in this report.

First we consider what a team means. The activity of a group of people working co-operatively to achieve shared goals is basic to our species (Baumeister & Leary, 1995). The current enthusiasm for teamworking in health care reflects a deeper,

perhaps unconscious, recognition that this way of working offers the promise of greater progress than can be achieved through individual endeavour. Mohrman, Cohen, and Mohrman (1995) define a team as:

a group of individuals who work together to produce products or deliver services for which they are mutually accountable. Team members share goals and are mutually held accountable for meeting them, they are interdependent in their accomplishment, and they affect the results through their interactions with one another. Because the team is held collectively accountable, the work of integrating with one another is included among the responsibilities of each member". Benefits of teamwork The belief that teamwork is the most effective way of delivering products and services has gained increasing ascendancy within diverse organisational settings (Guzzo & Shea, 1992; West, 1996). As organisations have grown in size and become structurally more complex, the need for teams of people to work together in co-ordinated ways to achieve objectives that contribute to the overall aims of organisations has become increasingly urgent. Mohrman et al. (1995) offer ten reasons for implementing team-based working in organisations: Teams are the best way to enact the strategy of organisations, because of the need for consistency between organisational environment, strategy and design (Galbraith, Lawler, & Associates, 1993). Teams enable organisations to speedily develop and deliver services cost effectively, while retaining high quality. Teams enable organisations to learn (and retain learning) more effectively (Senge, 1990). Cross-functional teams promote improved quality of services (Deming, 1986; Juran, 1989). Cross-functional teams can undertake effective process re-engineering

(Davenport, 1993). Time is saved if activities, formerly performed sequentially by individuals, can be performed concurrently by people working in teams (Myer, 1993). Innovation is promoted within team-based organisations because of crossfertilisation of ideas (Senge, 1990; West & Pillinger, 1995).

 Flat organisations can be monitored, co-ordinated and directed more effectively if the functional unit is the team rather than the individual (Galbraith, 1993, 1994). As organisations have grown more complex, so too have their information processing requirements; teams can integrate and link in ways individuals cannot (Lawrence and Lorsch, 1969, Galbraith, 1993, 1994).

This approach to the delivery of services and products is not simply a managerial fad, since there is substantial empirical evidence that the introduction of teamwork can lead to increased effectiveness in the delivery of both quantity and quality of goods or services (Guzzo & Shea, 1992; Weldon & Weingart, 1993).

Macy and lzumi (1993) conducted an analysis of 131 organisational change studies in order to determine their effectiveness. Those interventions with the greatest

effects on organisational performance and 'the bottom-line' were team-related interventions. They also reduced turnover and absenteeism more than did other interventions, showing that team oriented practices can have broad positive effects in organisations. Other research by Kahleberg & Moody (1994), who studied over 700 work establishments, found that those in which teamwork was developed were more effective in their performance than those in which teams were not used. Finally, Applebaum and Batt (1994) offer similar evidence. They reviewed the results of a dozen surveys of organisational practices, as well as 185 case studies of innovative management practices. They too found compelling evidence that teams contribute

to improved organisational effectiveness, particularly increasing efficiency and quality. Teamwork in health care The importance of teamworking in health care has been emphasised in numerous reports and policy documents on the National Health Service (NHS). One (NHSME, 1993) particularly emphasised the importance of teamworking if health and social care for people were to be of the highest quality and efficiency:

"The best and most cost-effective outcomes for patients and clients are achieved when professionals work together, learn together, engage in clinical audit of outcomes together, and generate innovation to ensure progress in practice and service."

Some limited research has suggested the positive effects of multidisciplinary teamworking in health care. However, there are many difficulties inherent in

comparing evaluation studies, which include teams having different objectives and organisation patterns, studies variously controlling for other concurrent changes in local services and the pre-existing variations in services and cultures (Jackson, Gater, Goldberg, Tantam, Loftus & Taylor, 1993).

In terms of the delivery of care, teams have been reported to reduce hospitalisation time and costs, improve service provision, enhance patient satisfaction, staff motivation and team innovation. outcomes below. We review the literature relevant to each of these

Reduced hospitalisation and costs Sommers and colleagues (2000) compared primary health care teams with physician care across 18 private practices, and concluded that primary health care teams lowered hospitalisation rates and reduced physician visits while maintaining function for elderly patients with chronic illness and functional deficits. Significant cost savings were born from reduced hospitalisation, which more than accounted for the costs of setting up the team and making regular home visits. Jones (1992) also reported that families who received primary health team care had fewer hospitalisations, fewer operations, less physician visits for illness and more physician visits for health supervision than control families. A similar pattern emerged for terminally ill patients, where their increased utilisation of home care services more than offset savings in hospital costs, such that there were average savings of 18% in hospital costs (Hughes, Cummings, Weaver, Manheim, Brown & Conrad, 1992).

In another study in the U.S., Eggert and colleagues (1991) concluded that a team focussed case management system generated similar benefits for elderly, chronically ill patients. The team approach reduced total health care expenditures by 13.6%, when compared to an individualised case management system. The team combined earlier discharge, more timely nursing home placement and better-organised home support and care, to reduce patient hospitalisation by 26%. Similarly, the cost increases in ambulatory and nursing home care were offset by fewer and shorter stay hospital admissions and reduced home care utilisation. For patients with dementia in this study, the team model of case management reduced overall costs even further, by 41% (Zimmer, Eggert & Chiverton, 1990). At the end of the 27-month study, there were more team than control patients living at home and fewer in nursing homes. An

audit of the case managers' records highlighted more intense management activity in the team group, where patients were referred more frequently for medical evaluation, respite and day care. Team case managers had smaller caseloads, made more home visits and had more case conferences. Teams were more familiar with local community resources and were reported as being more responsive to patient crises. The team approach was reported to offer greater intensity of case management, which resulted in more efficient care provision in hospitals and home health services.

Improved service provision Primary care teams appear to produce better detection, treatment, follow-up and outcome in hypertension (Adorian, Silverberg, Tomer & Wamosher, 1990). Specifically, nurses in England reported that working together in primary health care teams reduced duplication, streamlined patient care and enabled specialist skills to be used more cost-effectively (Ross, Rink & Furne, 2000).

Jansson, Isacsson and Lindholm (1992) analysed the records of general practitioners and district carers over 6 years in Sweden. Care teams (GP, district nurse, assistant nurse) were introduced into one region but were absent in another comparative region. The care teams reported a large rise in the overall number of patient contacts and in the proportion of the population who accessed the district nurse. Concurrently, there was a reduction in emergency visits, which they attributed to better accessibility and continuity of care in the teams.

Jackson and colleagues (1993) reported a similar pattern twelve months after the introduction of a community mental health team in England. They reported a threefold increase in the rate of inception to care, a doubling in the prevalence of treated psychiatric disorder and a reduction in demand on the hospitals outpatient services. It was suggested that the team was making specialist care more available to patients with severe mental illness who would not have previously received care from mental health services. The team also provided care in a timelier manner that was accessible and continuous.

Enhanced patient satisfaction Hughes and colleagues (1992) compared the provision of hospital-based team home care and customary care for 171 terminally ill patients in a large U.S. Department of Veterans Affairs hospital. They noted increased access to home care services and improved patient and carer satisfaction with hospital-based team home care. Both

patients and caregivers of the team expressed significantly higher levels of satisfaction with continuous and comprehensive care at one month, and they continued to express higher levels of satisfaction at six months. The team program maintained patients at home for significantly more days than the control group, who were kept in hospital in general wards for longer. Patients of the team received almost twice as many home visits as the control group and visited the clinic significantly fewer times.

Increased satisfaction by patients who had access to a primary health care team was reported to include a higher mean number of social activities, fewer symptoms and slightly improved overall health. These differences were noted in comparison to patients who only had access to a physician (Sommers et. al., 2000).

Staff motivation Primary care teamworking has been reported to improve staff motivation (Wood, Farrow, & Elliott, 1994). In a study in Spain, Peiro, Gouzalez-Roma & Romos (1992) showed relationships between work team processes, role clarity, job satisfaction and leader behaviours. Effectiveness of teamwork was also related to job satisfaction and mental health of team members. Sommers and colleagues (2000) suggested that lower rates of hospitalisation for patients of primary health care teams were more likely to be found in teams where individual members were most satisfied with their working relationships.

Innovation Teamwork is reputed to promote innovation in organisations including those in the health care sector. In order to promote organisational innovation, policy makers and practitioners are increasingly asking for clarification of the factors that determine innovation in teams. Many input and process variables have been demonstrated to predict innovation in teams.

In relation to inputs, there is some evidence that heterogeneity of team composition is related to team innovation (Hoffman & Maier, 1961; McGrath, 1984; Jackson, 1996). West and Anderson (1996) carried out a longitudinal study of the functioning of top management teams in 27 hospitals and examined relationships between team and organisational factors and team innovation. Their results suggested that team processes best predicted the overall level of team innovation, while the proportion of innovative team members predicted the rated radicalness of innovations introduced.

Specifically, West and Wallace (1991) found that team collaboration, commitment to the team and tolerance of diversity were positively related to team innovativeness.

By what means are these various benefits of teamworking in health care realised? Partly at least through their composition and through effective team processes such as communication, decision-making and problem-solving. We therefore briefly

review research in these areas before turning to consider the influence of the organisations within which teams function.

Team composition and Processes There is considerable agreement that heterogeneity of skills in teams performing complex tasks is good for effectiveness (e.g., Campion et. al., 1994; Guzzo & Dickson, 1996; Jackson, 1996; Millikan & Martins, 1996; Maznevski 1994). Heterogeneity of skills and knowledge automatically implies that each team member will bring a different knowledge perspective to the problem, a necessary ingredient for creative solutions (Sternberg & Lubart, 1990; West, 1997).

However, teams that are diverse in task-related attributes are often diverse in individual attributes. Variation in individual characteristics can trigger stereotypes and prejudice (Jackson, 1996) which, via interteam conflict (Tajfel, 1978; Tajfel & Turner, 1979; Hogg & Abrams, 1988), can affect team processes and outcomes. As an example, Alexander, Lichtenstein and DAunno (1996) found that individuals in multidisciplinary treatment teams in U.S. Department of Veterans Affairs hospitals, who were members of larger and more heterogeneous teams, reported poor team functioning. Physicians and social workers assessed team functioning more positively than did nurses. The greater the diversity of individual characteristics of team tenure, age and occupation within teams, the more negatively did team members assessed team functioning.

Gender Gender is an important influence on communication within teams. Not only are men consistently more assertive in public situations and confrontations (Kimble, Marsh & Kiska, 1984; Mathison & Tucker, 1982), but also communication expectations differ for men and women. Sex-role stereotypes prescribe passive, submissive and expressive communication for women while men are expected to be active, controlling and less expressive communicators (LaFrance & Mayo, 1978). Punishment for violation of expectations (Jussim, 1986; Jussim, Coleman & Lerch,

1987; Jackson, Sullivan & Lodge, 1993) may influence both the perceptions of women in teams and their willingness to participate in team communication. Such considerations are vitally important in health care teams where women dominate in number, but men predominate in the highest status positions (in the present research, GPs and psychiatrists, for example).

In support, Alexander, Lichtenstein and DAunno (1996) reported that the greater the gender diversity, the more positive were team members assessment of how cohesively and harmoniously teams operated. Their research suggested that mixed gender teams included different orientations to work, namely a female focus on workplace processes and relationships and a male focus on tasks and outcomes.

Team roles It is important that teams have the appropriate mix of clearly defined team roles. Jansson, Isacsson and Lindholm (1992) analysed the records of general practitioners and district carers over a six-year period across 2 districts in Sweden following the introduction of care teams into one region. They found that through the independent roles of nurses and doctors were retained in the primary health care teams, all team members interacted with the population in various situations, including home visits and complemented each other across different competencies.

Team affective tone Another important, but more controversial approach to understanding work team processes and effectiveness, is offered by research on team affective tone. George (1990) suggests that if members of a team experience similar kinds of affective states at work (either negative or positive), then affect is meaningful not only in terms of their individual experiences, but also at a team level. A number of studies have demonstrated a significant relationship between team affective tone and behaviour such as absenteeism (George, 1989, 1990, 1995). George proposes that teams that are interested, strong, excited, enthusiastic, proud, alert, inspired, determined, attentive and active, enable cognitive flexibility, creativity and effectiveness (George, 1996). However, she argues that team affective tone may not exist for all teams, so it cannot be assumed a priori that it is a relevant construct for every team. George (1996) sees team affective tone and team mental models as having a reciprocal influence. So in a team with a negative affective tone, members would have different cognitive processes from those in a team with a positive affective tone, which then may influence team effectiveness.

There is some evidence that team mental models play an important role in team decision-making (Klimoski & Mohammed, 1994), impacting on aspects of team decision-making such as problem definition, speed and flexibility, alternative evaluation and implementation (Walsh & Fahey, 1986: Walsh, Henderson & Deighton, 1988). A team that has a high negative affective tone may tend to be more rigid when making decisions. The nature and outcomes of team decision-making are therefore likely to be affected by the interaction between team affective tone and team mental models.

Communication The study of communication in teams has a long history in social psychology, but recent reviews by Guzzo & Dickenson (1996) and Guzzo and Shea (1992) reveal the paucity of thorough industrial and organisational research in this area. Blakar (1985) proposes five pre-conditions for effective communication in teams. Team members must have shared social reality within which the exchange of messages can take place, including a shared language base and perception. Team members must be able to decentre, to take the perspective of others into account in relation to both their affective and cognitive position (Redmond 1989, 1992). Team members must be motivated to communicate. There must be negotiated and endorsed contracts of behaviour (i.e. agreement among team members about how interactions take place). Finally, the team must attribute communication difficulties appropriately, so if one of the other preconditions is not being met, the team is able to correctly identify the problem and develop a solution.

Several

research

studies

in

England

have

highlighted

interprofessional

communication problems within primary health care teams.

West and Field (1995)

and Field and West (1995) interviewed 96 members of primary health care teams and described factors that impacted upon teamworking and communication in health care. Structured time for decision-making, team cohesiveness and team-building all influenced communication within teams. They highlighted the failure of health care teams to set aside time for regular meetings to define objectives, clarify roles, apportion tasks, encourage participation and handle change. Other reasons for poor communication included differences in status, power, educational background, assertiveness of members of the team, and the assumption that the doctors would be the leaders (see also West & Pillinger, 1995; West & Slater, 1996).

Communication difficulties between different professional groups have been highlighted particularly. Bond, et. al., (1985) surveyed 161 pairs of General

Practitioners (GPs) and health visitors, and 148 pairs of GPs and district nurses who had patients in common. They reported low levels of communication and collaboration between GPs and community nursing staff and suggested that GPs had a very poor understanding of the health visitor's role. Similarly, McClure (1984)

describes low levels of communication in a survey of 48 health visitors and 45 district nurses attached to general practices. Community nurses reported that

communication with practice staff was usually only about specific immediate patient issues rather than team objectives, strategies, processes and performance review. Health visitors were noted to be similarly unenthusiastic about progress in teamwork. Ross, Rink and Furne (2000) found that health visitors perceived teams as less effective. They suggested that health visitors were comparatively more defensive about the benefits of changing role boundaries and considered themselves less able to contribute to the teams as currently constituted. Cant and Killoran (1993) reached similar conclusions, based on their research study with 928 practice nurses, 682 health visitors and 679 district nurses. They argued that joint professional training and the instigation of regular team meetings were necessary to promote good communication.

Cott (1997) used a social network analysis of 93 health care workers across 3 multidisciplinary long-term care teams to explore communication processes within teams. She concluded that higher status multi-professional members communicated most openly and worked fairly autonomously across loosely structured tasks, with low levels of authority. In contrast, hierarchical nursing sub-teams did not report high levels of information sharing.

West and Slater (1996) reported that much of the potential benefit of teamwork was not being realised, with less than one in four health care teams building effective communication and teamworking practices (see also West & Poulton, 1997). In a similar vein, the Audit Commission report in 1992 drew attention to a major gap between the rhetoric and reality:

"Separate lines of control, different payment systems leading to suspicion over motives, diverse objectives, professional barriers and perceived inequalities in status, all play a part in limiting the potential of multiprofessional, multi-agency teamwork. . . for those working under such

circumstances efficient teamwork remains elusive" (Audit Commission, 1992).

A number of researchers in different countries have highlighted the impact of communication problems on patients across different types of teams. Nievaard (1987) interviewed 112 nurses and 298 patients across 6 medical and surgical wards of 2 general hospitals in the Netherlands. The study demonstrated the phenomenon of problem shifting, where communication problems within the team were transferred onto patients. It was reported that for hospital teams with a good communication climate, nurses perceived patients as more attractive and interesting and less dependent. However, if nurses viewed relationships with doctors, managers and nurses in the team as problematic, their images of patients tended to be more negative (unattractive, non-cooperative, dependent) and they did not want to increase their contacts with patients.

Yeatts and Seward (2000) reported similar findings when they compared 3 selfmanaged work teams in a medium size U.S. rural nursing home. They concluded that enhanced communication between team members positively affected the service to residents. Observations of a high performing teams meetings showed that team members had a high level of respect for each other, listened to each other, and were not afraid to disagree when they held different views. Team members sought and valued approval from each other, and they assisted each other to complete tasks.

Several studies have demonstrated how individual perceptions about teamwork and roles can influence communication in teams. Dreachslin, Hunt & Sprainer (2000) developed a grounded theory of the role that race plays in the self-perceived communication effectiveness of nursing care teams in the U.S. They concluded that racially diverse team members evaluated team communication according to different perspectives and alternative realities.

When team members develop belief systems that are consistent with their perspective and incongruent with other vantage points, differences in perspective can result in alternative realities. Alternative realities encourage participants to attribute causality differently which in turn fuels team conflict and miscommunication by diminishing the teams ability to reach a common understanding of both the source of the conflict and the optimal path to its resolution through effective communication (p. 1408).

Black participants were more likely to suggest that race exacerbated team conflict and miscommunication, whereas white participants attributed problems to role and status in the team. Further, different emphases and responsibility for communication were acknowledged amongst the diversity of races, ethnicities, ages and genders. Social isolation, selective perception and stereotypes also served to reinforce these differences and deepen communication problems. Fewer occasions for social interaction reduced opportunities to develop shared beliefs and a common social reality across racial groups. The researchers therefore suggested that team

members be encouraged to understand different perspectives and appreciate alternative realities, in order to lessen social isolation and reduce selective perceptions and stereotyping behaviours.

Freeman, Miller and Ross (2000) also developed a grounded theory about collaborative practice at the levels of the organisation, group and individual. They conducted case studies of 6 teams working in a variety of specialist healthcare services (diabetes, medical ward, primary healthcare, neuro-rehabilitation unit, child development assessment, community mental health) and concluded that the meanings different professionals ascribed to teamwork shaped how they communicated and what they communicated about. When there was a lack of congruence about aspects of teamwork, communication could potentially be compromised. Individual perceptions determined the level of role understanding considered necessary, and the value assigned to others contributions. Differences in the understanding and valuing of team roles and levels of team learning exacerbated underlying resentments, undermined professional esteem and created conflict. Individual perceptions also influenced communication regarding tasks and about sharing professional knowledge and ideas.

Decision making Effective decision-making processes are central to team performance. Several

studies have reported the positive benefits of participative decision making in health care teams. Yeatts and Seward (2000) compared 3 self-managed work teams in a medium size U.S. rural nursing home. Team members of highly performing teams reported that their ability to participate in work related decisions greatly increased their job satisfaction and desire to come to work. These team members adopted a consensus model of decision making, in which they clarified the problem, considered alternatives, weighed the strengths and weaknesses of each alternative, and

selected the best option. Following their participation in making decisions, team members reported an enhanced self-image and self-confidence, and they described more positive interactions amongst themselves and with residents.

In contrast, Cott (1997) suggested that team members may not be equally empowered to participate in decision making. Using a social network analysis of 93 health care workers across 3 multidisciplinary long-term care teams, she reported that the highest status nurses and the core multidisciplinary professionals participated most in decision making and problem solving activities. In comparison, the lower status nursing sub-team primarily planned and assisted each other with their more mechanistic tasks.

Problem solving Team problem solving improves when members examine their definitions of a situation to ensure they are solving the "right" problem (see for example, Bottger & Yetton, 1987; Hirokawa, 1990; Landsberger, 1955; Maier, 1970; Schwenk, 1988). In contrast, teams that detect problems too slowly or misdiagnose them often are ineffective. Attributing problems to the wrong causes, or not communicating about potential consequences, often undermine team effectiveness, especially when team members fail to reflect on the possibility of error (Schwenk, 1984; Staw & Ross, 1989).

Teams that engage in more extensive scanning and discussion of their environments perform better than those which do not identify problems (Ancona & Caldwell, 1988; Main, 1989; Billings, Milburn & Schaalman, 1980). Tjosvold (1985; 1990) linked the open exploration of opposing opinions within teams with effectiveness. Maier and colleagues also suggested that cognitive stimulation produced novel ideas, and that team effectiveness could be improved if teams were encouraged to be "problem minded" rather than "solution minded" (Maier & Solem, 1962; see also Maier, 1950, 1970). Effectiveness was improved when teams questioned current approaches or considered other aspects of problems (Maier, 1952). Similarly, Hackman & Morris (1975) found that additional process discussions facilitated the quality of team performance. The judged creativity of team decisions was related to the number of comments made about performance strategy. When teams produced alternative solutions to a problem, or separated and recombined problem solving strategies, enhanced productivity was reported (Maier, 1970).

Teams that have to make complex decisions report that planning enhances their performance (Hackman, Brousseau & Weiss, 1976; Smith, Locke & Barry, 1990). However, when the environment becomes more uncertain, problem identification is more difficult (Hedburg, Nystrom & Starbuck, 1976; Kiesler & Sproull, 1982). Ineffective teams tend to deny, distort or hide problems (Stein, 1996). In some

teams, the identification of problems is discouraged as problems are regarded as threats to morale, or a source of conflict (Janis, 1982; Miceli & Near, 1985; Smircich, 1983).

Thus far we have reviewed the benefits (and potential difficulties) of teamworking in health care organisations - but the fact that teamworking takes place within organisations is often ignored in the zeal to promote team effectiveness. Accordingly, we now turn to address what is currently known about the influence of their organisations upon teams. Organisational context Recent research suggests the broader context within which teams work has an influence on their performance. Indeed the major change in emphasis in research on teams in the last 15 years has been the shift from discussion of intrateam processes to the impact of organisational context on teams. The organisation within which a health care team functions can influence team effectiveness in a variety of powerful ways. Researchers, such as Hackman (1990) and Tannenbaum, Beard and Salas (1992) have suggested that the following are among the contextual factors that influence team effectiveness:

Team and organisational rewards Team objectives and performance feedback Training and technical assistance Physical work conditions Organisational climate Inter-team relationships Contracts and management structures Team size

These factors will be discussed further, in turn.

Team and organisational rewards It has long been known in the social sciences that rewards are important for improving performance. Reward systems, such as public recognition, preferred work assignments and money enhance motivation and performance, particularly when the rewards are contingent upon task achievement (Hackman, 1990; Sundstrom et al., 1990; Vroom, 1964). However, team performance is most effective when rewards are administered to the team as a whole and not to individuals, and when they provide incentives for collaboration and communication rather than individualised work (Hackman, 1990). This reinforces individuals working together as a team. Gladstein (1984) found that in sales teams, pay and recognition affected the leaders behaviour and the way the team structured itself. Yet, NHS management directly undermines teamwork in primary health care when they provide bonus systems to GPs as independent contractors, despite the whole team contributing to the final outcome.

Clear team objectives and performance feedback In healthcare environments, team members need information about local health needs and services, and national policies and guidelines, in order to set objectives and target their activities appropriately. Further, feedback on team performance is important for setting realistic goals and fostering high team commitment (Lathom, Erez & Locke, 1988). Job satisfaction requires accurate feedback from both the task and other team members (Drory & Shamir, 1988). However, team feedback can be difficult to provide to teams with either long cycles of work or one-off projects (Sundstrom et. al., 1990).

Training and technical assistance Hackman (1990) argued that training and technical assistance is required for teams to function successfully. Knowledge and training about team functioning is needed to supplement team members own technical and medical skills and knowledge (Poulton & West, 1993; Poulton & West, 1994a, 1994b; Poulton & West, 1997). Limited empirical evidence suggests training is correlated with both self-reported effectiveness (Gladstein, 1984) and managers judgements of effectiveness (Campion et. al., 1993) in teams.

Physical Work Conditions Physical conditions are another situational constraint that affect the relationship between performance dimensions and team effectiveness. For example, a health

care team whose members are dispersed across sites, will find decision making more difficult and ineffective than a team whose members share the same physical location.

Organisational Climate The climate of the organisation - how it is perceived and experienced by those who work within it - will also influence the effectiveness of teams (Allen, 1996). Where the climate is one characterised by high control, low autonomy for employees, lack of concern for employee welfare and limited commitment to training, it is unlikely teamworking will thrive (Markiewicz & West, 1997).

The extra commitment and effort demanded in team-based organisations requires organisational commitment to the skill development, well-being and support of employees (Mohrman, Cohen & Mohrman, 1995). Competition and intrigue can

further undermine team based working in health care, since teamwork depends on shared objectives, participative safety, constructive controversy and support (West, 1990; West & Anderson, 1996). Ross, Rink and Furne (2000) reported that team members willingness to work in teams was limited by the lack of a common set of values about the benefits of teamwork. They recommended the need for clear

objectives, leadership, commitment and wide organisational ownership as precursors for working in teams.

Professional subcultures also influence team effectiveness. Kinnunen (1990) used an anthropological approach to distinguish different subcultures between medical, nursing and management staff in a large primary health care organisation in Finland. These three professional groups described different relationships to formal power structures, which influenced their group behaviour, leadership style, administrative orientation, decision-making preferences and patient interactions. In general, doctors and managers shared basic assumptions about work that were paternalistic, proactive, dominant and emphasised loyalty to authorities. In contrast, nurses

stressed participation, delegation, traditions and symbiotic harmony in work relations.

Inter-team relationships In a comprehensive study of team-based organisations involving both questionnaire and case study methods, Mohrman et. al. (1995) demonstrated that inter-team competition is a major threat for team-based working. Teams that compete may develop greater commitment to the teams success than the organisations success.

Thus the health care team may focus on increasing the financial benefits to their team at the expense of the wider National Health Service. Teams competing against, rather than supporting each other may withhold vital information or fail to offer valuable support in the process of trying to achieve team goals, without reference to the wider goals of the organisation. Thus, health care teams may fail to pass on information about former patients to other teams, focusing their efforts on their own teams immediate demands.

Ross, Rink and Furne (2000) reported a lack of focus on patient care in their evaluation of primary care nursing teams in England. Nurses perceived that current organisational change promoting teamwork was concerned with structure,

professional and organisational issues rather than with patient care. Some nurses were concerned that moves towards integrated nursing were primarily motivated to cut costs.

Contracts and Management Structures Other relevant aspects of the organisational environment in health include the independent contractor status of GPs and different management structures. There are very few organisations where one or more senior team members work as independent contractors and the rest of the team work within a variety of organisations. Even the most sophisticated management practices, in environments such as the oil and gas industry, are struggling with notions of how to operate joint venture systems - whereas health care teams must deal with these issues constantly but without the training and support given to teams in these other sectors.

Team size The size of the team is also important, since bigger teams experience much greater strains on effective communication. In most other sectors, teams tend to be divided once they reach 12 or 13 members. But primary and secondary health care teams (for example) can be 20, 30, 40 or more members in size. These teams would be more correctly termed organisations. In and of itself, this would not be a problem, if those who run such organisations are adequately trained to manage large operations. They require knowledge of the management of culture, power, conflict, spans of control, strategies, innovation and above all, people. Yet primary health care team leaders are rarely given such training (West, 1994). It is to the topic of leadership that we now turn.

Leadership There is considerable research evidence that leaders affect team performance (e.g. Brewer, Wilson & Beck 1994; Komaki, Desselles & Bowman, 1989) and evidence of the relationship between leadership style and team effectiveness. Eden (1990)

examined the effects of platoon leaders expectations on team performance. His work with the Israeli Defence Forces showed that those platoons which trained under leaders with high expectations, performed better on physical and cognitive tests. Podsakoff and Todor (1985) investigated the relationship between team members perceptions of leader reward and punishment behaviours and team cohesiveness, drive and productivity. Results showed that both leader contingent reward and

punishment were positively related to team drive and productivity. Leader contingent reward was also related to cohesiveness, while leader noncontingent punishment behaviour was negatively related to team drive. Jacob and Singell (1993) examined the effects of managers on the won-lost record of professional baseball teams over two decades and found that leaders did influence team performance by exercising tactical skills and improving the performance of team members. George and

Bettenhausen (1990) studied teams of sales associates reporting to a store manager and found that the favourability of leaders moods was negatively related to related to employee turnover.

Primary health care team members in England rated their effectiveness more highly when they had strong leadership and high involvement of all team members (Ross, Rink & Furne, 2000). In nursing care teams, Dreachslin, Hunt and Sprainer (2000) concluded that leadership mitigated the influence of race in self-perceived communication effectiveness. Participants comments supported the theme that

team leaders who encouraged discussion about differences enhanced perceived team effectiveness. They suggested that leaders provided a unifying force through validating the alternative realities and appreciating the different perspectives of team members, thus moderating the potentially negative effects of racial diversity on team processes. Developing Teams in Organisations To what extent is it possible to develop team working to ensure higher levels of effectiveness? Tannenbaum, Salas, & Cannon-Bowers (1996) have reviewed research in this area and related results to a comprehensive model of team which

integrates interventions (Tannenbaum, Beard and Salas, 1992). They describe a number of intervention types include team member selection and teambuilding: Team member selection Although organisations tend to use quite sophisticated methods for selecting employees for individual jobs, they rarely use systematic methods for selecting for teams. But systematic selection methods can help identify people with greater skill levels. There is strong evidence that a team composed of skilled and motivated people will be more effective than other teams (Tziner, 1988). Selection interventions could improve team effectiveness by increasing the professional or skill diversity of health care team members, thereby increasing the range of competencies in the team. Teambuilding Some teambuilding interventions focus on role clarification, some on interpersonal relationships or conflict resolution issues, while others take more of a general problem-solving approach (Tannenbaum, Salas & Cannon-Bowers, 1996). Team norms, attitudes, climate and power Many team

distribution can be affected by teambuilding approaches.

processes, including communication, decision-making and mutual role understanding, are often direct targets of team building interventions.

Weldon and Weingart (1993) describe the importance of planning in teams for achieving team goals, and suggest that team members are characteristically slow to respond to changes in their tasks or their environments that make their strategies ineffective or their goals obsolete. They propose five ways of supporting team work. Goals should be set for all dimensions of performance that contribute to the overall effectiveness of the team; feedback should be provided on the team's progress towards its goal; the physical environment of the team should remove barriers to effective interaction (consider the difficulties faced by members of a dispersed health care team); team members should be encouraged to plan carefully how their contributions can be identified and co-ordinated to achieve the team goal; and team members should be helped to manage failure, which can damage the subsequent effectiveness of the team.

Pritchard, Jones, Roth, Stuebing and Ekeberg (1988) tested some of these ideas by measuring the effects of team feedback, goal setting and incentives on productivity. Five organisational units in the military were studied. One, a maintenance section, repaired a variety of electronic equipment used for aircraft communications. The other four sections together made up a material storage and distribution branch. Productivity baselines were established before each team received new "treatments" (i.e., performance feedback eight months after the study began, goal setting five months later, and incentives a further five months later) to determine the incremental effects of these "treatments". First, the level of performance of the teams was

measured over a period of eight months and then information on their performance was given to each unit for five months. The teams next set clear targets in addition to the performance feedback, and their performance was measured for another five months. Feedback was in the form of computer-generated reports, given monthly to the personnel of each unit. Finally, incentives were offered for high performance, in the form of time off from work. Using these approaches, the average increase over baseline productivity was 50% for feedback, 75% for goal setting and 76% for incentives. The results showed a major increase in productivity among the teams, though the unique contribution of each component of the intervention is difficult to estimate accurately. performance. Transition of organisations to teamworking One of the most exciting developments in the field is the new emphasis upon the development of team-based working in organisations (Mohrman, et. al., 1995; Markiewicz & West, 1996, 2001). This reflects a concern amongst practitioners with how team-based working can be effectively introduced into organisations. Mohrman et. al., studied 25 teams in four companies using a grounded research methodology, involving managers and internal customers. In the second phase of their research they surveyed 178 teams across seven corporations, involving team members, managers and customers. In this way, they developed a five stage design sequence for the transition to a team-based organisation: 1. Identifying work teams and the nature of the task This involves process analysis to determine essential work activities that have to be conducted and integrated to produce products or services; deliberations analysis which identifies dialogues about issues that have to be repeatedly resolved in order Both goal setting and feedback had powerful effects on

to provide shared direction and enable people to complete their tasks; and task interdependence analysis which determines where and to what extent individuals and teams have to rely on each other to complete their tasks. 2. Specifying integration needs In order to integrate across multiple teams and components of business units, Mohrman et al recommend management teams, representative integrating teams (where an overall co-ordinating team had representatives from each of those teams collectively involved in producing a product or service), individual integrating roles, and improvement teams. 3. Clarifying management structure and roles This stage involves putting as much self-management responsibility into the teams as possible; involving team members in determining how leadership tasks will be performed and by whom; using lateral mechanisms for cross-team and organisationwide integration so that teams participate in that integration; and creating management roles which link teams to the organisational strategy and ensure they are responsive to the organisational and wider environmental context. 4. Designing integration processes The research evidence suggests that team-based organisations should set clear directions in the organisation, (for example by defining, communicating and operationalising a strategy at all levels, aligning goals, assigning rewards in accordance with organisational goals, and planning collectively); managing information distribution and communication; and developing an appropriate decision making strategy (by clarifying decision making authority, and appropriately involving organisational contributors). 5. Developing performance management processes Finally, the model suggests the need to manage performance - defining, rewarding and reviewing performance and involving internal and external customers, and team members. Mohrman et. al., report that the more people were rewarded for individual performance, the worse team performance was. The more people were rewarded for team performance, the better was the team and the business units performance and the more process improvements the team and the business unit instituted.

Conclusions and Research Mode

A review of the literature reveals that progress has been made in understanding the factors that influence the ability of people to work effectively together in teams. However, in the health care domain progress is still patchy and only a few studies are constructed on firm theoretical bases. Progress is further inhibited by the added difficulty of operationalising the concept of effectiveness. In the research described in this report we attempted to build our research on a well-accepted theoretical base and to engage a large number of health care teams in the research endeavour. Moreover, we were charged with grasping the nettle of effectiveness in health care and developing robust and sufficiently broad measures of this difficult concept. Finally, the research team, drawn from a wide range of epistemological backgrounds and theoretical orientations, determined to employ diverse, powerful and innovative research methods to answer the question of what factors influence the effectiveness of health care teams. The starting point for the research was a model of the factors influencing team effectiveness and which distinguishes between at least three major domains of effectiveness. Theoretical approaches to understanding teams at work have been dominated by the input-process-output structure, mainly because of its categorical simplicity and utility (see Figure 1 below) (West, Borrill, & Unsworth, 1998). This is the model used to guide the research described in this report. Figure 1: Input, process, output model of team effectiveness

INPUTS Domain Health Care Environment Organisational context Team task Team composition

GROUP PROCESSES Leadership Clarity of objectives Participation Task orientation Support for innovation Reflexivity Decision making Communication/ integration

OUTPUTS Effectiveness - self and externally rated Clinical outcomes/quality of health care Innovation - self and externally rated Cost effectiveness Team member mental health Team member turnover

Inputs Teams work within a domain such as primary care, secondary care or community mental health. They also work in a health care environment that may be more or less deprived. The team works for and within an organisation; thus it will be affected by the interaction with the surrounding organisational context. A team has a task that potentially impacts upon team processes and effectiveness (the management of immunisation for children under five years; intensive care nursing; or care of the elderly with mental health problems). The team consists of a collection of individuals - who represent the groups composition varying in professional background, gender, age, personality etc. Finally, the team exists within a wider society that will affect the teams fundamental beliefs and value systems, i.e., the cultural context. Processes Processes within teams enable them to achieve their goals. A fundamental requirement for effectiveness is that teams have clear objectives to which their members are committed. Other processes include participation in decision-making, emphases on quality, and support for innovation. Another fundamental process is the extent of coordination and integration of team members work (Worchel, Wood, & Simpson, 1992). And of course, leadership and communication are likely to be

important to team effectiveness. Another potentially important process variable is reflexivity or the extent to which team members collectively reflect on the objectives, strategies, processes and environment of the team and make changes appropriately and accordingly. Outputs Six principle outputs can be distinguished: overall effectiveness, clinical outcomes, team member mental health, innovation, team member turnover, and cost effectiveness. In the research programme described in this report we explore the relationships between inputs and processes; inputs and outputs; and processes and outputs in 390 UK NHS teams, during the course of which we consulted with over 7,000 NHS personnel and with a large number of NHS clients.

Using this model, we determined to explore the extent to which team working was associated with better quality health care for patients and to identify the factors associated with effective teamwork.

Chapter 2
Primary Health Care Team Research Methods and Sample Details
The research with Primary Health Care Teams had two stages. The first was a questionnaire survey, and related data collection methods, involving large numbers of teams conducted in order to gather data on team inputs, processes and outcomes. The second stage involved intensive examination of a sub-set of teams to explore in more depth targeted team processes and outputs. Quantitative Methods

An overview of the methods used is given in Figure 2.1.

Figure 2.1: Details of the three samples and research methods

Sample size

Survey data

Additional questionnaires/ Telephone interviews

External ratings

100 teams Team composition 1156 respondents Team functioning Team effectiveness Team innovation Member stress

Team effectiveness Team innovation Team composition Team meetings Team management Decision making

The Sample The research design required data to be gathered from 100 Primary Health Care teams (PHCTs) varying across a number of dimensions, including size (number of team members, number of GPs, list size); Jarman index; location (urban, rural, inner city), and geographical location. Databases of GP practices were accessed from 19 Health Authorities and 300 teams were randomly selected.

Letters explaining the objectives of the research and inviting teams to participate in the research, together with an information sheet were sent to the senior GP partner,

senior health visitor and practice manager/senior receptionist in each practice. A reply slip was included, which also solicited additional information about the team (fund holding status, frequency and type of meetings, Jarman index, number of GPs, list size etc.)3

The initial letter was followed up with a telephone call to the practice manager/senior receptionist at all 300 practices. If teams had already indicated a willingness to participate, practical arrangements for questionnaire distribution were made. Researchers requested the name of a contact person in the team to enable continued effective liaison. The contact person was telephoned at a later date to determine whether the team was willing to take part in the research. Teams that did not return a reply slip were also telephoned and provided with additional information.

Further follow-up telephone calls were made until the team made a decision about participation in the research (some PHCTs were contacted six or seven times before a decision was made). When teams agreed to collaborate in the research, questionnaires were sent to the contact person for distribution to team members.

After three months the response rate from 10 teams was below 30% and 23 had not returned any questionnaires. These teams were dropped from the sample and

replaced with 7 teams based in an inner city area, and 7 from a rural location, resulting in a final sample of 100 teams. The total response rate was 55.8%.

Response rates for teams ranged from 21.4% to 100%, with a mean of 57.6%.

Data Collection Methods Data on team functioning and effectiveness were collected using three methods: self report questionnaires completed by individual team members; self report and telephone interview surveys with the team contacts; and external ratings from primary health care representatives and health authority staff.

1156 respondents from 100 PHCTs completed questionnaires on their perceptions of team functioning and team effectiveness. Of these, 85% were female; 15% were GPs; 14.2 % practice nurses; 23% trust nurses (health visitors, district nurses,

Copies of interview schedules, questionnaires and all data collection instruments are available from the first author of this report.

midwives); 33.5% administrative staff, 7.3% managers and 4.3% professions allied to medicine (PAMs).

Team contacts from 77 PHCTs provided information on team context, team composition and team processes in a self-report questionnaire survey, and 100 provided information via a telephone interview. This enabled a reliability check on the data for 77 of the teams.

Questionnaires completed by individual team members This questionnaire was in four sections (a copy of the primary health care questionnaire is included in Appendix I). Section 1: Team working This contained seven measures of team working. Four of these were drawn from the Team Climate Inventory (Anderson & West, 1994,1998) that is based on a welldeveloped theoretical model of team functioning (West, 1990). The four measures assess levels of: team participation clarity of and commitment to team objectives emphasis on quality support for innovation.

Three other measures were included: reflexivity the extent to which team members reflect upon their team objectives, strategies and processes and make changes accordingly (West, 1996; Swift & West, 1998). team innovation the extent to which the team has introduced innovations in objectives, work strategies, processes and relationships

Respondents were also asked to write descriptions of the major changes or innovations introduced by the team in their work in the previous 12 months. Section 2: Effectiveness This included 21 measures of primary health care team effectiveness adapted from Poulton and West (1999). There are three underlying dimensions:

team working patient orientation organisational efficiency

Section 3: Team member stress This included a measure of psychological stress, the GHQ-12 (Goldberg, 1972; Goldberg & Williams, 1991). The GHQ-12 is widely used as a screening tool for detecting minor psychiatric disorder in the general population, and in occupational mental health research. It covers feelings of strain, depression, inability to cope, anxiety based on insomnia, lack of confidence and other psychological problems. Within a Department of Health-funded study of the mental health of the NHS workforce, the GHQ-12 showed good validity against a psychiatric interview (Hardy, Shapiro, Haynes, & Rick, 1999). Section 4: Biographical information This section included questions on biographical and team characteristics (e.g. age, gender, ethnic origin, job title, employer, team composition, team leader). Additional Practice Information Survey This was completed by the contact person in the PHCT (usually the practice manager). It included questions on: team context (relationships with external

agencies such as health authorities and trusts); type of primary health care practice (fundholding, non-fundholding, dispensing) quality of premises; team composition number in each occupational group, grade, hours worked, time working in the team); staff development; and team processes (communication and decision making in meetings). Additional Practice Information - Telephone Interview Schedule The contact person in the team (usually the practice manager) responded to the telephone interviews. The focus of the questions was on decision-making and communication in the team: specifically who was involved in making operational, strategic and clinical decisions in the team, how these decisions were communicated in the team and what mechanisms were in place within the team to promote communication (memo systems, message books, informal meetings, email). Information was also gathered on the services and clinics provided by the team.

External ratings team effectiveness and innovation Health Authorities employ staff to provide support to primary health care teams, a role involving working closely with a wide range of teams. Contact was made with staff in this role at each of the Health Authorities where the teams in the sample were located. They were asked to provide ratings of effectiveness and innovation for all of the teams from their area that were participating in the research. External ratings of effectiveness were obtained for 84 teams. They were rated on the same 21

effectiveness dimensions included in the primary health care team questionnaire.

Examples of the dimensions include: The extent to which teams made efficient use of the practice budget The extent to which teams previewed and adjusted skills in line with the identified health care needs of the practice population External ratings changes introduced by teams Three representatives from Primary Care rated the changes or innovations introduced by the teams (reported in the questionnaire for individual team members). One of the raters was employed by a Local Health Authority and had responsibility for developing team working in primary care. Another was employed by a community trust, also in a role which supported primary care teams. The third rater was a part time general practitioner who had been involved in development and research projects in primary care. They rated teams on four dimensions (West & Anderson, 1996): magnitude - how great would be the consequences of changes introduced radicalness - to what extent the status quo would change as a consequence novelty - how new in general were the changes impact - to what extent changes would improve PHCT effectiveness.

Using the ICC (2) (Shrief & Fleiss, 1979) the inter rater agreement was calculated for each dimension: Magnitude - 0.663, Radicalness - 0.630, Novelty - 0.539, Impact 0.779. Sample Details In this section we describe characteristics of the primary health care team sample that participated in the questionnaire and interview component of the research

programme. Following the model used to guide this research (see page 1) we describe four categories of inputs: team task - Indicated by the size of the practice population and its

fundholding status team composition - team size, ratio of part-time team members, gender mix, number of occupational groups represented in the team health care environment - the Jarman Index (an index of social deprivation), its location (city, urban, urban/rural or rural) organisational context - the NHS Region within which the team is located.

We also describe the relationships between these four domains of inputs.

The

reader will also find an account of the frequency and content of meetings held in the teams, and of the team members perceptions of leadership in the team in subsequent chapters.

Team Task Practice population or list size The practice population or list size ranged from 1500 to 21,850. The mean size was 6,902 patients with a standard deviation of 4,692 (see Figure 2.2). Figure 2.2: Percentages of primary health care teams with patient populations or list sizes of various sizes
List Size
30% 25% 20% 15% 10% 5% >20000 1000115001<2500 10000 2501500175010% 15000 20000 5000 7500 List Size

Fundholding status Forty one percent of teams in the sample had fundholding status.

Location Four location categories were used: city, urban, urban/rural and rural. Sixty-five percent of teams were located in urban areas other than cities, 20% in cities, 6% in areas described as both urban and rural, and the remaining 8% were in rural locations.

Health care environment Jarman index The Jarman index is a measure of social deprivation. The higher the score the greater the health needs of the practice populations served by the team. Scores in the sample included in this research programme ranged from 0% to 100. The mean Jarman score was 15.52 and the standard deviation 22.72. Figure 2.3 shows the percentage of teams with each category of Jarman score.

Figure 2.3: Team Location

70% 60% 50% 40% 30% 20% 10% 0% City Urban Urban/Rural Rural

Figure 2.4: Jarman index

25 20 15 Frequency 10 5 0% 0 1 - 10%

Whole time equivalents Using hours worked to calculate team size enabled an estimate to be made of the number of whole time equivalents. This statistic shows that team size varied from 1.49 to 31.9 members. The mean size was 9.35 with a standard deviation of 6.75. The size of teams working in so-called single handed GP practices, ranged from 1.88 to 16.13, with a mean of 7.48 and a standard deviation of 3.74. Number of GPs The number of GPs in the teams ranged from 1 to 11. The mean number of GPs was 3.7 and the standard deviation was 2.4 (Figure 2.5). Whole time equivalent GPs

The range of whole time equivalent GPs was from one to ten. The mean was 3.16 and the standard deviation 2.0.
Figure 2.5: Number of GPs in the primary health care teams

Team Size (no of GP's)

30% 25% 20% 15% 10% 5% 0%

1 2 3 4 to 5 6 to 8 9 to11

Gender The majority of team members (85.5%) of the Primary Health Care team sample were female. The break down for gender by occupational group is shown in Figure 2.7 and this reveals that the only imbalance in favour of men is in the highest status group GPs. Otherwise, primary health care is a domain in which women form the vast majority of the workforce. Primary health care is largely in the hands of women in the UK. And of course, this has important implications for our understanding of its

functioning, issues of team processes and, given the anomalous preponderance of male GPs, of leadership issues. Fig 2.7: Distribution of gender by occupational group in primary health care teams GPs Practice Nurses Male Female 99 75 5 160 Trust Nurses 16 249 Admin staff 10 378 24 61 8 42 1 4 Managers PAMs Other

Fig 2.8: The distribution of mean age across the occupational groups Mean Age GPs Practice Nurses Trust Nurses Admin/clerical Managers PAMs Others 41.2 42.1 42.2 43.7 44.3 41.6 40.8 Standard Deviation 7.6 7.6 8.3 10.2 8.4 11.0 8.2

Age The distribution of age across the occupational group is shown in Figure 2.8.

Occupational Groups

The majority of teams comprised: GPs, practice nurses, administrative staff, district, health visitors and one or more practice managers. Less than 15% of teams did not include trust nurses, and only 6% of teams had no manager. Twenty four percent of teams included other types of staff (counsellors, community psychiatric nurses, physiotherapist etc).
Proportion of occupational groups The proportion of each of the main occupational groups in the teams is shown in Figure 2.9.

Figure 2.9: Proportion of occupational groups in the sample

PAM's Others Managers

GPs Practice Nurses

Admin/cleri cal

Trust Nurses

A different picture of team composition emerges when hours worked is taken into account and team membership calculated using 'whole time equivalent' figures. This shows that largest grouping is GPs, with only 10.7% of the input to the team being provided by trust nurses.

2.10: Proportion of occupational groups in the sample

Managers

PAM's Others GPs

Admin/ clerical Trust Nurses

Practice Nurses

Organisational Context

Figure 2.11 Regional variations

Location South East and London Midlands East Anglia Nottinghamshire South Yorkshire West Yorkshire

Number of teams 18 4 4 18 43 13

Teams were distributed across six regions with the distribution shown in Figure 2.11.
Relationships between Input Factors In this section we describe the relationships between aspects of team task, team composition, health care environment and organisational context. As we might predict, there are some important and significant relationships between them.

The number of team members and the number of patients on the teams list were positively correlated (0.85), with an average of 291 patients on a teams list per member of staff. This ratio did not vary across location (city, urban, urban/rural, rural), Jarman index, or average of number of hours worked by team members.
There were no significant differences in the composition of the teams between different types of locations (city, urban, urban/rural, rural). Nor were there differences in the composition of fundholding and non-fundholding practices.

There was a significant relationship between Jarman score and number of managers; teams with fewer managers had a higher Jarman score.

Teams with a higher Jarman index also had significantly more other types of staff in the team. This may reflect the fact that the range of services required is much greater in socially deprived than socially enriched areas.

There was a higher proportion of Professions Allied to Medicine (PAMs) in teams with a larger list size.

Jarman index / location 73 teams provided a Jarman index score of more than 0%. The mean score was 15.52 and the standard deviation 22.72. These were distributed across locations as follows: City 18 Urban 39 Urban / rural - 5 Rural 6 Unclassified - 5

The Jarman index for city practices was significantly higher (mean = 32.6%) than for urban practices (mean = 11.9%)4

Qualitative Research Methods Research in the second stage of the research programme explored in depth, and using a variety of consultation and qualitative research methods, all issues of team functioning and effectiveness. The methods used are shown in Fig 2.12.

The data collected as part of this research can be subjected to much further analysis and information extraction. The researchers are committed to working with others to ensure the maximum exploitation of this hard won data set. If there are analyses readers wish to conduct the researchers would urge them to contact the first author of this report.

Figure 2.12: Consultation and Qualitative Research Methods used for Primary Health Care Teams Analysing PHC team processes Audio and video recording of two meetings for each of twelve teams Development of PHC objectives Series of four national workshops with domain relevant experts from primary care. Work with 12 teams to validate objectives. Development of PHC effectiveness measures In-depth work with two PHCTs to develop measures. Training and dissemination to ten PHCTs. Training team facilitators.

Video and audio recordings of team processes All teams involved in the questionnaire and interview components of the research programme were invited to participate in the next stage of the research. This involved analysis via video and audio recording of two of their team meetings. Teams were selected randomly for this element of the research. Twelve teams volunteered.

We selected meetings that were multidisciplinary in composition and that involved decision-making (as opposed to information dissemination only). This is because understanding team working in this context demands that we observe professionals from different backgrounds working together dynamically, and integrating their different perspectives to initiate action and change. Multi-disciplinary meetings were those in which a range of disciplines (doctors, nurses, health visitors, practice managers, etc.) was represented and participated. For the most part, primary health care teams allowed us to observe the practice business meetings, in which the dayto-day running of the practice was discussed. In one team, the GP partners made all decisions affecting the practice. In this case we recorded the partners meeting. Wherever possible, we recorded two meetings of the same type for each team. Dates for meeting recordings were at the discretion of the practice, so the two meetings recorded were not always in sequence. Researchers requested that recorded meetings should be held in their usual locations, with their usual meeting protocols (agendas, minutes, chairing procedures, etc.), and that attendance should be the same as if the meeting was not being recorded. The researcher who managed the recording equipment made herself as unobtrusive as possible. Meeting size ranged from three people to twenty-five.

Audio recording was done with two omni-directional PZM tabletop microphones linked to different channels of a high quality audiotape recorder; the microphones were set up so as to maximise channel differentiation but to be unobtrusive enough that participants would not move them. A single static video camera on a tripod was trained to record the gross movements of as many of the participants as possible; this record was used only to aid speaker identification during transcription. Before each meeting was opened all participants introduced themselves and their occupation and upon the basis of this each was allocated a speaker number. Therefore the first person to introduce him or herself became speaker 1, the second speaker 2 and so on.

Meetings were transcribed from the audiotapes by an audio typist who had not attended the meeting. Audio typists transcribed complete contributions in order,

according to when they began, labelling each contribution by speaker number, but did not code finer timing information. Speaker identification was facilitated both by the video recording and by a seating plan drawn up during the meeting by the person recording the meeting. A contribution was defined as a period of speech from one individual in which the only major pauses coincided with silence from the other speakers, so that the pause was likely to be caused by the speaker thinking and not by the speaker listening to someone else's contribution. Under this definition, speakers cannot follow themselves in the speaking order. Overlapped speech was transcribed, with the extent of the overlap roughly marked. Infrequently, parts of the meetings were omitted because they were so badly overlapped that we could not track individual contributions. After transcription, the transcripts were completely

anonymised taking out all staff, patient, place names, place and local authority names or possible team or person identifiers.

An example transcription excerpt is given in Figure 2.13. Transcription proceeds one contribution per row. Column one contains the speaker number. Column two

contains the words said, with coding information in a different font, and column three contains any notes which the transcriber wished to make (for instance, about people entering or leaving the room).5
5

In previous work using these methods on four to twelve person meetings, transcribers were able to agree very reliably who made any one contribution; using the kappa statistic, K=.93, k = 2, N = 230, with an average of 2% and a maximum of 6% non-backchannel contributions left as unidentified.

Figure 2.13: An example of the layout of the transcription format used for PHCT meeting transcription

1 4 3

Shall I open /4 the meeting Yep, lets get on with it. My apologies I am going to have to leave before the end. I have an appointment in Place 1. Are you skiving off? Group laughter

Because one of the factors of interest in our study is how well teams communicate across disciplines, our analysis relies on a classification of meeting participants by occupation. For ease of reference, categories are identified by colour as well as number. For primary health care teams, Figure 2.14. shows the categories used.

Figure 2.14: Categories used for PHCT meeting participants

1 2 3 4

GPs practice managers practice nursing staff, including nurse practitioners attached staff (mostly health visitors, midwives and district nurses) administrative staff (mostly secretaries and receptionists)

miscellaneous (visitors, resident caretakers, medical students)

Development of Performance Measures for PHCTs In the broader organisational literatures on team effectiveness, a widely adopted approach is the Productivity Measurement and Enhancement System (ProMES) based on research by Naylor, Pritchard & Ilgen (1980) (see also Pritchard, 1995). Effectiveness criteria are established in group discussions with team members and managers. The variables are then psychologically scaled to a common effectiveness scale. Based on group consensus about expected levels of effectiveness, which are given a zero value, maximum effectiveness levels (set at +100), and minimum levels (-100) are set. Each variable is also weighted in terms of

its perceived contribution to the overall effectiveness of the team or organisation. The system is then used to set objectives, develop indicators, monitor and improve performance and give feedback to the team (Pritchard, 1995). The approach has been spectacularly successful in many settings (Pritchard, 1995) and is promising for primary health care, because of the sophistication of the approach, its theoretical robustness and practical utility in complex contexts.

The ProMES was implemented in three main stages: 1. Core objectives for primary health care teams were developed using the constituency approach and ProMES in four national workshops with representatives from Primary Care. 2. Usable ProMES effectiveness measures were developed and applied within primary health care teams. 3. Primary health care team members and trust representatives were trained to develop and implement effectiveness measures using ProMES in primary health care teams.

An initial stakeholder analysis identified 13 stakeholders in primary health care. These included: GPs Health Visitors District Nurses Practice Nurses Midwives Administrative staff Department of Health NHS Executives Patients Health Authority Researchers PAMs CPN Advice was sought from contacts in primary health care about key experts who could represent the views of each stakeholder group, and about whether the initial list of

stakeholders was sufficiently comprehensive. The experts suggested by the contacts were sent information about the research programme, invited to attend the four oneday workshops, and asked to suggest additional or alternative key experts who could also make a contribution. In addition, professionals who were currently engaged in clinical practice in primary health care teams were invited. The majority of those contacted were able to commit themselves to attending two or three of the workshops. A full list of those attending and their institutional affiliations is given in Appendix II.

During the workshops focus group methods were used. Delegates were divided into three working groups. These were designed so that (a) a range of stakeholder views was represented, and (b) one or two or group members had attended most or all of the workshops and so could share with new members the learning and experience from previous workshops. Each group worked with a trained facilitator, and a note taker recorded the group discussion and the decisions made.

Workshop 1 Objective: to develop objectives for primary health care.

The delegates were presented with a set of objectives for primary health care developed by the researchers (based on the work of Poulton & West, 1994) and then worked together to discard, add or refine objectives. The revised objectives were discussed with members of four primary health care teams (who endorsed their relevance and value), and combined into a single list.

Workshops 2 and 3 Objective: to develop measures of effectiveness in relation to the primary health care.

Delegates were presented with the refined and agreed objectives for primary health care. Each group worked on developing effectiveness measures for objectives.

Workshop 4 Objective: to plan the implementation of effectiveness measures in primary health care.

In the fourth Workshop, participants critically appraised the objectives and measures developed, and considered how they could be applied in practice by PHC teams and others. This session was used to plan the implementation and evaluation of the effectiveness measures in practising teams. Seven core objectives, with associated sub-objectives, were identified and agreed by the professionals attending the workshops and 19 effectiveness measures were developed (see Appendix II).

Work implementing performance measures was carried out in two phases. In the first phase, we carried out in-depth work with two primary health care teams to develop performance measures, based on the objectives and measures developed in the four national workshops, and used these to provide feedback on team performance. Details of this work are provided in Appendix III. We worked with one team over a period of 15 months, and with the second for a period of 8 months. A design team was established in each PHCT that included at least one representative from each of the occupational groups in the team. In one-hour workshops held every fortnight, ProMES was used to develop performance measures specific and appropriate to each team. Teams carried out further development work between meetings such as gathering data and consulting colleagues.

In the second phase, PHCT representatives and trust employees attended a ProMES in Primary Health Care training programme. etters inviting representatives from PHCTs to attend the ProMES training were sent to 60 PHCTs that had participated in the first stage of the research (all teams with a response rate of 50% and above). Letters were sent to the practice manager, senior health visitor and senior partner. Follow-up phone calls were made to the teams, but representatives from only two attended the training. The other participants were service

representatives, employed by community trusts, to support and develop primary care team working.

The training programme included an overview of the ProMES approach; the development of performance measures; guidance on running ProMES workshops in PHCTs; training in how to collect and use performance information. The programme for the training is outlined in Appendix IV. After the training, three follow-up

workshops were held with participants. The purpose of these was to provide support to those using ProMES with primary health care teams, and to critically review the measures developed in the training. Feedback from the participants about the success of their interventions in teams has been positive and suggests there is real

benefit to all aspects of primary health care team functioning from employing this approach. At the same time, it is a demanding exercise that requires commitment by team members to implement.

The results from stage 1 and 2 of the research programme are described in the following two chapters. Details of the findings from the primary health care team

surveys and external ratings are outlined in Chapter 3, and the objectives and effectiveness measures developed for primary care, outlines in Chapter 4.

____________________________________________________________________________

.Chapter 3

Primary Health Care Team Results from Survey and External Ratings

Summary of Findings

Large PHC teams are rated as more effective and innovative by external raters.

The greater the number of professional groups represented in the primary health care teams, the more highly rated is the innovativeness of the team.

The better the team processes and reflexivity, the more innovative they are rated by external raters.

The greater the number of team meetings, the higher the level of innovation in primary health care teams.

PHC teams with clear leaders have good team processes.

Conflict over leadership leads to poor quality team working. However, teams where leadership roles are shared are more innovative.

Introduction The data analysis explored two main questions

Is there an association between the composition of a primary health care team and team processes? Is there an association between the composition and processes of the primary health care team and the effectiveness of the team?

The team characteristics, team processes and measures of team effectiveness are summarised in Figure 3.1.
Figure 3.1: Team Inputs, Processes and Effectiveness Variables Characteristics Individual Age Gender Ethnicity Grade Time in job Time in team Team Occupational groups Size (number of members) Number of GP's Hours worked Grade Mix Training Team context List size Location (urban, rural, city) Jarman index Fundholding status Dispensing Purchasing Quality of premises External contacts Relationships with HA Processes Team processes Participation Innovation Objectives Emphasis on quality Reflexivity Number meetings Types of meetings Frequency of meetings Decision making Leadership Integration and communication in the group Effectiveness Team ratings Organisation Team working Patient focus Innovation External ratings (innovation) Magnitude Radicalness Novelty Impact External ratings (effectiveness) Organisation Team working Patient focus Types Innovations Quality of care External collaboration Responsibility of health Use of resources Professional development Team satisfaction Responsiveness Stress (GHQ 12)

Team inputs Information about the team members ages, gender, ethnicity, grade, professional group, employer, tenure and team leadership was collected from each team member.

Information was also gathered on team size, hours worked, qualifications, training, list size, practice location (urban/rural/city), Jarman index, fundholding status, and whether the practice was purchasing and/or dispensing. Team Processes Individual team members rated team processes on six dimensions: participation; support for innovation; clarity of team objectives; emphasis on quality; reflexivity; and integration. The variables participation, support for innovation, clarity of team objectives and emphasis on quality were very highly correlated and were combined to form one variable describing team processes. Information about decision-making processes, communication, number and types of meetings, who attended meetings, and how the team was managed was collected from practice managers. The

information on team meetings was categorised according to who contributed to operational, strategic and clinical decisions. In addition, a new variable

interdependence was developed which assessed the extent to which there were mechanisms within the team to encourage interdisciplinary communication. Team effectiveness This was assessed using information from a variety of sources. Team members rated their teams effectiveness on three dimensions: team working, organisational efficiency and patient orientation. Team members also rated their teams

innovativeness and described the innovations implemented by the team in the previous year. These reports were categorised to determine the types of innovations implemented. External raters assessed the innovations reported by the teams on four dimensions: magnitude; radicalness; novelty and impact on team effectiveness. External ratings of team effectiveness were provided by Health Authority representatives on two dimensions clinical and organisational. Individual team members also completed the GHQ-12 (a measure of mental health or psychological stress). The measures of interest for this report are overall effectiveness, effectiveness of patient-centred care (both externally rated and self-rated), overall innovation (both externally rated and self-rated), number of innovations to do with healthcare, and mental health measured by GHQ-12. Results The main method of analysis was multiple regression. For each dependent variable, possible predictors were split into groups according to type of variable (e.g. occupational group, team context), and stepwise regression was used to identify

those which might ultimately predict the dependent variable. The second stage of each analysis involved entering all those identified into a further stepwise regression, to find out which variables had significant effects independent of other predictors. This way, process variables were predicted by team characteristics, and effectiveness, innovation and mental health were predicted by team characteristics and team processes.

Since there was often evidence of relationships between size and other variables, this was always dealt with first. Where relationships were apparent, later analysis revealed whether this was due to team size per se or another feature of having a larger team. Question 1 Is there an association between the composition of a primary health care team and team processes?

There was no evidence that PHC team size had an association with any team process except frequency of meetings. Here we see that teams of 20 or less have, on average, 2.6 meetings a month; teams of 20-30 members have 6.1 meetings a month, and teams of over 30 have 6.5 meetings a month.

Other predictors of team processes (after the second stage of analysis) are shown in Figure 3.2.

Figure 3.2: Relationships between Team Composition and Team Processes Dependent variable Team processes Predictor variables Proportion of managers Proportion of other staff No. of GPs (WTE)1 None Proportion of managers No. of practice nurses Proportion of other staff No. of managers -0.256 0.418 0.366 -0.290 0.035 0.001 0.002 0.012 0.065 0.175 0.237 0.305 0.253 -0.244 p 0.010 0.032 0.036 R2

0.192

Reflexivity Integration Number of meetings Consensus on leadership

1

WTE = whole time equivalents

Patterns emerging here are mainly to do with the representation of managers and other staff types in the teams (anyone other than GPs, nurses, admin/clerical staff,

managers and PAMs). Having a larger proportion of other staff in the team has a positive effect on team processes and consensus on leadership. A larger number of managers also has a positive effect on team processes, but has the opposite effect on integration. There is a negative association between a larger number of managers and agreement about who leads the team.

Question 2 What affects the effectiveness and innovation of a primary health care team?

Team size was positively associated with a number of dimensions of effectiveness and innovation, as shown in Figure 3.3. Generally, larger teams were rated as more effective by external raters and introduced more innovations overall, and specifically in relation to patient care. Figure 3.3: Relationship between team size and ratings of effectiveness, innovation and mental health Variable Overall effectiveness (external) Overall effectiveness (self-rated) Effectiveness of patient care (external) Effectiveness of patient care (self-rated) Innovation (external) Innovation (self-rated) Number of innovations re: patient care Mental health Correlation 0.284 0.086 0.255 0.125 0.403 0.123 0.255 0.056 p 0.012 0.401 0.002 0.222 < 0.001 0.226 0.013 0.585

Further analysis revealed that the relationships between team size and all the innovation variables was curvilinear, with teams of sizes around 40 being the most innovative. Notice that there were no relationships between team size and self-rated effectiveness, innovation or mental health.

It is also interesting to note the associations between team size and the individual items of the external effectiveness ratings, to see what aspects of effectiveness are most related to team size. These are shown in Figure 3.4. Larger teams appear to be more responsive to patients and are more likely to conduct clinical audit.

Figure 3.4: Relationship between team size and individual externally rated effectiveness items Effectiveness item Provision of information about services Implementing procedures for dealing with patients comments, suggestions and complaints Maintaining clinical competence in line with patient needs Auditing clinical practice Setting protocols Commitment to professional and personal development Understanding and valuing roles of all members Implementing a clear strategy for communication Profiling health needs and targeted interventions Reviewing and adjusting skill mix Collaborating with other agencies Making effective use of budget Implementing recommendations of the PHC Charter Concentration on achievement of The Health of the Nation targets Correlation 0.131 0.269 0.186 0.263 0.373 0.273 0.003 0.132 0.160 0.209 0.094 0.126 0.311 0.258 p 0.236 0.015 0.083 0.017 0.001 0.012 0.981 0.203 0.165 0.068 0.389 0.265 0.004 0.024

The main reasons for larger teams being more effective appear to be their effectiveness in setting protocols and implementing recommendations of the PHC charter.

Results of the stepwise regression analyses of effectiveness on team characteristics and processes are shown in Figure 3.5.

Figure 3.5: Relationships between Team Composition and Processes, and Ratings of Effectiveness Dependent variable Overall effectiveness (external) Overall effectiveness (selfrated) Effectiveness of patient care (external) Effectiveness of patient care Predictor variables No. of admin. staff Proportion of GPs None Team size Team processes 0.357 0.632 0.010 <0.001 0.127 0.400 0.400 -0.279 P 0.003 0.035 R2 0.199

(self-rated) The relationship between self-rated effectiveness of patient care and team processes is not entirely surprising, given that both variables were constructed from the individual questionnaires sent out and hence this analysis is prone to common method variance. The relationship between team size and externally rated effectiveness of patient care is shown in Figure 3.6. This relationship is more reliable and suggests that better patient care is delivered in larger primary health care teams sizes, up to 30 to 40 members. Figure 3.6: Relationship between team size and effectiveness of patient care (externally rated)
6

Effectiveness of patient care

1 0 10 20 30 40 50 60 70

Team size

Figure 3.7: Relationships between Team Composition and Ratings of Innovation Dependent variable Innovation overall Predictor variables Professional diversity Reflexivity Team size Professional diversity No. of practice nurses Team processes Lack of clear leadership Reflexivity Team processes 0.308 0.318 0.290 0.263 0.299 0.342 0.274 0.384 0.315 P 0.002 0.001 0.003 0.024 0.011 0.005 0.023 0.018 0.050 R2

0.363

No. of healthcare innovations

0.298 0.454

Innovation (self-rated)

It seems that professional diversity in the team, reflexivity (taking time out to review objectives, strategies and processes) and team processes all have positive effects on innovation. Some of these relationships are illustrated in figures 3.7 to 3.11. Figure 3.8: Healthcare innovations and professional diversity

Mean number of innovations re: quality of healthcare

6 5 4 3 2 1 0 <7 7 8 9 10 11 12 > 12

Number of professions represented in team

Figure 3.9: Overall innovation and professional diversity

Overall innovation - external rating

5 4.5 4 3.5 3 2.5 2 1.5 1 6 or fewer 7 8 9 10 11 12 or more

Number of professions represented in team

Figure 3.10: Relationship between reflexivity and overall innovation

2.0

1.5

1.0

.5

0.0

Overall innovation

-.5

-1.0

-1.5 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0

Reflexivity

LFigure 3.11: Relationship between team processes and number of innovations in healthcare
10

Innovations in healthcare

-2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 4.2

Team climate

Bearing in mind that a larger score represents poorer mental health, results show that teams which are composed of members who are relatively similar in terms of age, have a larger proportion of managers, and avoid conflict over who leads the team, are characterised by better mental health for their members. This is shown in Figure 3.12. Figure 3.12: Relationships between Team Composition, Processes and Team Members Mental Health Dependent variable Mental health Predictor variables Proportion of managers Age diversity Conflict over leadership -0.420 0.363 0.253 p <0.001 0.001 0.017 0.375 R2

Frequency of team meetings Frequency of meetings was also examined as an explanatory variable. Figure 3.13 shows that frequency of meetings in primary health care team predicted external ratings of innovation. Coming together to discuss objectives, exchange information and make decisions is likely to lead to the generation of ideas for new and improved services and ways of working. Figure 3.1: Frequency of PHC team meetings as a predictor of innovation

Dependent variable Innovation overall (external) Number of innovations re: healthcare

0.242 0.198

p 0.026 0.072

R2 0.059 0.039

This result is illustrated in Figure 3.14.

Figure 3.14: External Ratings of Overall Innovation and Number of Meetings

4 3.8 3.6

Overall innovation

3.4 3.2 3 2.8 2.6 2.4 2.2 2 1 or less 1 to 4 4 to 6 6 to 10 More than 10

PHCT Meetings per month (average)

It was also shown that this effect is independent of both self-rated processes and team size. Leadership Research evidence suggests that leadership is an important factor contributing to team effectiveness. We therefore explored the contribution of leadership to team effectiveness and innovation in primary care teams separately. We explored the extent to which there was a clear leader in the PHC teams, and who was regarded as the leader. Only a third of PHC teams reported having a single clear leader. Nearly half reported having a number of people lead the team, which, in most contexts, is likely to cause considerable confusion. The most frequently named leader of PHC teams is the Practice Manager. Only a third of team members nominated a GP. Clarity of leadership was examined as an explanatory variable. Figure 3.15 shows that team processes were poorer where there was no clear leadership, (from either one individual or several people), or where there was conflict over leadership.

Figure 3.15: Clarity of Leadership in the PHC team predicting processes Dependent variable TCI mean score Reflexivity Integration Predictor variables Lack of clear leadership Conflict over leadership Lack of clear leadership Conflict over leadership All having leadership roles -0.311 -0.294 -0.366 -0.250 0.214 p 0.001 0.002 <0.001 0.008 0.035 R2 0.180 0.193 0.046

Figure 3.16 shows that lack of clarity about leadership in the teams also predicted lower levels of effectiveness as rated by the team. Figure 3.16 also shows that innovation, rated externally, was higher in teams with shared leadership, or in teams where more people reported that everyone had leadership roles. Of course, it may be that shared leadership is itself an innovation with primary health care. Teams which had no clear leadership or conflict over leadership were less likely than others to reviews and modify their objectives, strategies and team processes.
Figure 3.16: Clarity of Leadership in the PHC team predicting externally rated effectiveness and innovation Dependent variable Overall effectiveness (external) Overall effectiveness (selfrated) Effectiveness of patient care (external) Effectiveness of patient care (self-rated) Innovation (external) Innovation (self-rated) No. healthcare innovations Mental health Predictor variables Lack of clear leadership Lack of clear leadership None Lack of clear leadership Shared leadership All having leadership roles Lack of clear leadership None Conflict over leadership 0.294 0.003 0.086 -0.215 0.237 0.216 -0.365 0.034 0.020 0.033 <0.001 0.046 -0.249 -0.299 p 0.029 0.003 R2 0.062 0.089

0.087 0.133

These effects, and those for externally rated effectiveness, are all entirely mediated by group processes suggesting that the mechanism by which leadership influences effectiveness is through developing good team processes, such as shared objectives, participation, emphasis on quality and support for innovation.

We also find that there is less clarity of leadership in teams which have a greater proportion of part time workers (r = 0.309, p = 0.016), and there is less likely to be a single clear leader in teams with greater professional diversity (r = 0.309, p = 0.016). Both of these support the finding in Figure 3.2 that there is less consensus on who the team leader is in teams with a larger proportion of other staff types.

Overall, the findings from this stage of the research reveal a very clear picture of the factors predicting the effectiveness and innovations of primary health care teams: size, clear leadership, professional diversity and integration through regular meetings are key factors in predicting PHC team performance. Of course, it could be that teams that innovate and are effective have the confidence to recruit members from diverse professional backgrounds, and are required to meet more often because of the innovations they introduce. Clear leadership may emerge as a consequence of innovation and as a consequence of the cohesiveness arising from effective performance. Such interpretations are feasible and need to be explored empirically. However, the interpretation implied in our presentation of results (inputs and process predict performance) is consistent with research into effectiveness of teams from across a range of sectors and countries (Cohen & Bailey, 1998, West, Borrill & Unsworth, 1999).

Chapter 4
Qualitative Research: Developing Objectives and Effectiveness Measures for Primary Health Care Teams

Summary of Findings

Using the constituency approach seven core objectives were developed which were judged relevant and covered all the main aspects of primary health care activity.

Measures which could be used to measure performance on each of the objectives were developed in workshops with a range of primary health care stakeholders.

Primary health care teams used the ProMES approach to develop measures which could be used to measure their performance against the objectives for primary care.

Primary health care teams were able to use the measures developed to get feedback on their performance, and use this information to introduce improvements in patient care.

Measuring Effectiveness in Health Care

There is little agreement in primary health care about what constitutes effectiveness. One reason for this is that primary health care comprises a wide range of stakeholders (health care professionals, trusts, health authorities, patients, carers, voluntary groups) each with their own aims, objectives and priorities which influence how effectiveness is conceptualised. In addition, there is considerable variation in philosophies of care among the professionals groups within primary care (Toon, 1994), and different approaches and perspectives on what is judged to be high quality of care (Maxwell, 1992). One consequence of this is that health care will be judged as more or less effective depending upon the criteria adopted by the particular stakeholder, or on the philosophy or care espoused by a professional group.

To enable these differing priorities and perspectives within health care to be taken into account the qualitative research carried out by the research team used the constituency approach (Connolly 1990) to develop objectives, and the Productivity Measurement and Enhancement System (ProMES) developed by Naylor, Pritchard and Ilgen (1980) to develop effectiveness measures. There were two main stages to the work: developing objectives and effectiveness measures in national workshops; developing effectiveness measures with primary health care teams. Stage 1 - National Workshops: Developing Objectives and Effectiveness Measures for Primary Health Care The aim of this stage of the qualitative research was:

To develop a set of objectives for primary health care which was acceptable to all perspectives in primary health care

To develop effectiveness measures which were acceptable to all perspectives in primary health care.

The constituency approach was used to develop objectives for primary health care in four national workshops with representatives from primary health care (see chapter 2). These objectives were then validated in workshops with representatives from 12 primary health care teams. The objectives and sub-objectives developed as a result

of the workshops and consultations with primary health care team representatives are shown in Fig 4.1. Figure 4.1: Core Objectives for Primary Health Care teams Promote, maintain and improve health Provide high quality health care Accurate identification of individual and population health care needs Review and improve the effectiveness of health care provision Manage illness, injury and disease taking account of agreed standards and evidence based practice Enable patients/clients to make informed decisions about their own health. Proactively encourage positive health behaviour Implementation of health education and preventative care programmes Human resources skills, knowledge, expertise, time Physical resources budgets, equipment, premises Individual annual training plans which take account of the plans of the PHCT Equal access to training/development resources Team working Mechanisms for reviewing and acting upon staff dissatisfactions, conflicts and complaints Gather information and feedback from clients/community stakeholders/opinion leaders Build external relationships with clear objectives and high levels of participation, interaction and trust

Enable personal and community responsibility for individual health

Efficient use of resources Continuous personal and professional development High team member commitment, stress and satisfaction Responsiveness to clients and community Collaboration and partnership with other relevant organisations

The first aim of the national workshops was to get agreement on the objectives for primary health care, and to develop a set of objectives that cover all aspects of team activity. The work carried out by workshop participants, and the subsequent

amendments made as a result of the rating and discussions with PHCT representatives, enabled this main objective to be achieved. Given the diversity of

views, agendas and perspectives in primary health care it was a major achievement that by the end of the four workshops agreement had been reached.

The second aim of the workshops was to develop effectiveness measures for primary health care. A preliminary set of effectiveness measures was developed during the workshops that reflect the range of stakeholder perspectives. developed further and used by primary health care teams. These can be

The research team

carried out additional work on some of the preliminary measures, developing indicators of team effectiveness that could be used to measure performance. Effectiveness Measures Developed in the Workshops Objective 1 - Promote, maintain and improve health

Quality of care

Patient Charter taken into account health promotion activities carried out appropriate skill mix in the team to meet patient needs measure - % of appropriate consultations as % of total consultations measure - appropriate immunisation rates (without adverse incidents) measure - effective management and knowledge of chronic diseases (epilepsy, diabetes, asthma) measure - quality of patient consultations measure - appropriate admissions to hospital

Accessibility of service

appropriate number of surgeries offered and times (also flexibility) appropriate length of consultation (also flexibility) waiting times. Time taken to get routine and emergency appointments (with any member of the PHCT) availability of non face to face contact i.e. telephone access clients seen consistent with the severity of their needs (e.g. emergencies seen quickly)

Chronic disease management

effective management and knowledge of incidence of critical diseases: E.g. Epilepsy, Diabetes, Asthma In terms of Diagnosis Registers Protocols Interviews

Referral/use of other services

appropriate referrals to other services. Such a measure could indicate a lack of skills in the team or illness in the community number of effective or appropriate contacts with agencies such as palliative care, social services etc. i.e. good network of services level of appropriate access to the right services. Quality of partnerships and alliances in referrals is important here. This indicator might also be linked with the range of skills in the team

identification and reduction of health and social care grey areas e.g., when health care professionals do social care activities appropriate waiting times for admission to hospital i.e., for treatment from other agencies rates of emergency admissions/self referrals

Treatment

use of evidence based treatment and prescribing protocols appropriate intra-team referral. The group felt this was possibly more important than referral to other agencies low adverse complications incidence care delivery derived from plan of care. Having action plans helps evaluation of goals the team produces R&D strategy (based on consensus) the team produces clinical audit and clinical supervision action plans (based on consensus)

care packages/episodes of care (rather than just number of contacts) progress towards Health of the Nation targets

Identification of health needs

identification of health needs and the mechanisms to adjust efforts to match these needs utilisation of external bodies to identify service plans and needs met

Data Collection: Practice level

measure - types of information collected demographics/diseases/conditions/ activity levels) measure - accessibility of data collected to PHCT measure - PHCT contributing to data compare with national/regional data

Data collection: Local, regional and national sources

Assess completeness of data set

Use of data for: daily planning longer term planning - strategy/direction identifying gaps in provision and skill mix measure - number of action taken/changes made - up-take of training - modification of skill mix - review process - formal service plans

budget allocation consistent with priorities Identifying and utilising opinion leaders in the community and community networks

Accountability

meeting NHS care standards meeting NHS reporting requirements meeting requirements of other appropriate external agencies progress towards Health of the Nation targets

Objective 2 - Enable personal and community responsibility for individual health effective health education and preventative health care programs appropriate immunisation rates (without critical adverse incidents). appropriate will vary in accordance with local needs) information to patients and health education - includes information and knowledge and explanation for patients - making it personalised - so patient is recognised as an individual increasing knowledge about health in the population - i.e. with employers, teachers etc. patients educated to make appropriate self-referrals to members of the PHCT. Where is the locus of control, within the team or with the patient? number of health problems revealed by screening provision and take up of preventative health care programmes What is

Objective 3 - Efficient use of resources

monitor appointment management - DNAs protocols: new, renewed, rejected use of accommodation/equipment develop skills inventory and monitor use of skills measure - input costs: GP: practice size measure - initiatives developed to use time effectively and review process measure - balance between outputs and resources/monitor over time measure - % of time with patients review duplications of roles/effort

existence of evidence based prescribing protocols for practice (and review of these). There is a need to close the loop between cost effectiveness and clinical effectiveness of prescribing

existence of evidence based treatment protocol (including shared protocols and reviews) use of clinical guidelines (not just medical - so incorporates everyone in team) planned clinical audit degree to which safety standards were being complied with (Baseline could be minimum standards set by Health and Safety Executive) how effectively the PHCT computer systems are being used

Objective 4 - Continuous personal and professional development

Development of skills

Regular development and learning needs to be considered at the level of the individual, the team, the national governing bodies, and the professional bodies that monitor health care professionals

strategy plan for training and development - long term and short term - individual skills/job description - match health needs/individual needs - match to organisational objectives - who contributes to developing the plan - take account of each individuals understanding of development

measure - commitment to development of skills in the team equal access to/management of training budget research activities carried out - how funded, quality and quantity. utilisation of a full range of training methods (e.g., on the job, networking) opportunities for job exchange skill sharing - opportunities and time spent mechanism in place to evaluate the effectiveness of training and development that is done, including publication of the existence of training opportunities, time available, equipment, instructors, etc.

job performance of staff should be assessed on a regular basis using an agreed upon procedure

Checklist of activities which contribute to the development of skills:

availability of peer support/mentoring (inside and outside)/advocacy individual career development plans reviews of development plans IIP in place staff appraisal - linked to short term and long term goals identification of training needs - and review of these training equally available across team feedback on training attended/portfolio maintained/accreditation protected time available for professional development access to resources to support training and development indicators for learning and training

Team member mental health/stress

procedures to resolve conflict between patient/practitioner needs procedures for taking account of personal needs/family commitments staff allowed to be off sick

Objective 5 - High team member commitment, stress and satisfaction

measure - staff turnover/absence measure - how valued staff feel/commitment/grievances mechanism in place for reporting satisfaction levels back to team members and dealing with dissatisfaction when it becomes a problem

Team development

team participate in team development activities? availability of social budget shared understanding of roles and values the team contribute to the annual report/business plan? the team has regular meetings procedure for coordinating sub-groups and whole team procedures for communication

procedures to integrate roles/expectations across professional groups processes for critically evaluating and improving decision making equal opportunity for participation in decision making

Learning organisations

measure - support for innovation measure - resulting changes

Objective 6 - Responsiveness to clients and community

a commitment to client satisfaction within the team complaints procedures in place accessible information produced for patients patient choice re health care take account of patient perceptions of improved health and stress use questionnaires/surveys to assess patient satisfaction user involvement in decisions about their own health actions taken in response to patient suggestions for improvement not the same as complaints - giving patients the opportunity to make comments without feeling as if they are complaining.

carefully listening to the client giving clients the information to make informed choices getting inputs on client needs from clients, community and opinion leaders, groups representing clients

Objective 7 - Collaboration with other organisations

measure - staff use of skills and resources available effective contacts with related agencies and groups outside the PHCT

Development Work Carried Out by the Research Team

The research team carried out additional work after the national workshops refining some of the measures of effectiveness identified in the workshops. described below. These are

Objective 1 - Promote, maintain and improve health

The PHCT would have a monthly (or more frequent) staff meeting where a sample of cases was reviewed. This review would include the appropriateness of who saw the client, what procedures used, and whether that client was handled appropriately in all aspects. The measure would be the percentage of cases which were considered as being managed appropriately. This would also be the basis for discussion of what improvements need to be made for those specific clients and for clients in general.

The task of developing a health needs analysis can be broken down into definable steps, e.g. get information on how to do such an analysis, decide on a plan for doing the analysis for that particular PHCT, gather the information, put the information together into a form that the PHCT can use to make decisions. Each of these steps would be given a time for completion. The indicator would be the percentage of the analysis completed compared to the anticipated time for completion.

Survey on client perceptions of health improvement after treatment.

For

example, each client is given a questionnaire or a sample of clients are called by phone and asked about improvements. Measure is the percentage of clients improving. For the various specific targets given by agencies outside the PHCT such as immunisation rates, develop a scoring system whereby each level of meeting the objective gets a certain number of points, e.g. if the target immunisation rate was 80%, actually doing 80% would give 100 points, 60% immunised would be 20 points, 70% 80 points, 90% 130 points, etc. The

number of points would be based in the importance of that target. The index would be the percentage of actual points earned compared to the maximum possible points received if all targets were met.

The percentage of required reports completed on time

The number of required reports returned by agencies requesting corrections or additional information. (This would be an index of the quality of the reports.)

Objective 3 - Effective use of Resources

Number of new initiatives developed that are designed to help team members use their time better. These initiatives should also be reviewed on a regular basis to ensure they are still effective.

Percent client contact time as a percentage of total time.

This measure gets at

how much time is devoted to clients. It does not measure how well that time is being spent. Other indicators are needed to address this issue. (Note that this indicator is one where there is probably an optimal level between the extremes. Too little time with clients may suggest too much administration time. Too much time with clients may suggest too little administration time.)

Percentage of staff turnover over time. High staff turnover leads to inefficient resource utilisation because it takes time to teach procedures to new staff and work is lost as a departing staff member leaves. This measure would also be an indicator for the satisfaction of team members.

Percentage of appointments which are unfilled or where the client did not come.

Objective 4 - Continuous personal and professional development

Training and development. A list of training and development experiences for each person on the team would be developed each year. For example,

attendance at a certain type of conference, training on a piece of office equipment, learning a new procedure, etc. This list would be the development plan for that person for that year. There would be two measures for training and development. The first would be the percentage of team members who had the written plan. The second measure would be the percentage of the development plan items actually completed.

Which team members are reviewed, given feedback, and have a formal, jointly developed action plan for making improvements.

Objective 5 - High team member commitment, stress and satisfaction

Measure overall satisfaction on a monthly or bi-weekly basis with a very brief questionnaire that would take no more than 2 minutes to compete. Measure would be the percentage of staff indicating Satisfied or Very Satisfied with their jobs.

Staff turnover is also a satisfaction measure. Note this measure under Effective Management of Resources.

Objective 6 - Responsiveness to clients and community

Establish a formal procedure where clients can make complaints including a process for following up on these complaints. Measure is the number of such complaints which were not concluded to the clients satisfaction within one week.

Stage 2 - Effectiveness Measures Developed by Primary Health Care Teams

The ProMES approach is based on a theory of motivation which proposes that effort is maximised when there is a clear link between effort and outcomes, there is agreement about what are valued outcomes, feedback is provided on performance and the evaluation of performance is judged to be fair (Pritchard, Jones, Roth, Stuebing & Ekeberg (1988). Research evidence shows that involving individuals in the process of agreeing the valued outcomes from their work and developing methods for assessing their performance has a greater impact on performance than when these are imposed (Pritchard, 1995). The research team therefore carried out ProMES work with primary health care teams so they had the opportunity to develop their own effectiveness measures.

There were two main aims for this work: To demonstrate that primary health care teams could develop effectiveness measures using the ProMES approach To demonstrate that primary health care teams could use the measures developed to get feedback on their performance.

Qualitative work using ProMES was carried out with two PHCTs (see Chapter 2).

There were four distinct stages to the work:

1. Establishing a 'design team', these were representatives from the team who were primarily responsible for developing the measurement and feedback system.

2. Reaching agreement that the objectives developed in the constituency workshops, were relevant and related to all the main activities of the organisation/team.

3. Develop measures that could be used to assess the extent to which these objectives are being achieved.

4. Using measures to gather information about how well the team was performing.

The researchers worked with the primary health care design teams over a period of eighteen months, meeting for one hour once a fortnight. As a result of this work the primary health care teams successfully developed effectiveness measures that they could use to assess performance on all of the objectives for primary health care. A major issue in primary health care is the considerable work pressures and demands made on all members of the team. This is a major constraint on the time team members have available to engage in activities which do not directly contribute to the delivery primary health care team services. It was therefore a significant

achievement that teams were able to develop measures, and demonstrates what can be achieved as result of a relatively small investment of time.

The measures developed by the teams are listed below. Information on how to use the measures is provided in Appendix III.

Objective 1: Promote, maintain and improve health Measure 1 - Review of quality in case management Percentage of cases judged to be managed appropriately on the most relevant quality dimensions.

Measure 2 - Young Peoples Sexual Health Percentage unwanted teenage pregnancies in a 6-month period Percentage of teenagers prescribed the morning after pill in a 6-month period Percentage of teenagers requesting pregnancy tests in a 6-month period

Measure 3 - Young Peoples Health - Alcohol and Drug Misuse Number of teenagers attending A & E after drug overdose in a 3-month period Number of teenagers attending A & E after excessive alcohol consumption in a 3month period. Measure 4 - Patient access to consultations with a GP The number of days that patients wait to see a GP of their choice Measure 5 - Patient access to a quality consultation with GPs Percentage of patients whose appointment with a GP is minutes duration in a 3-month period. Measure 6 - Use of out of hours services by patient

Percentage reduction in the use of private out of hours services by patients in a 6-month period.

Measure 7 - Patients have access to an appropriate health professional Percentage of patients, in a 6 month period, who have contact with a health professional from the team at a time and location most appropriate to them and to the professional. Measure 8 - Patients have access to a home visit from an appropriate health professional. Percentage of patients in a 6 month period who have a home visit from the health professional judged by the patient and the health professional to be most appropriate.

Objective 2: Enable personal and community responsibility for individual health Measure 9 - Patients understand the role and function of the PHCT.

Number of patient requests, use health professionals time and PHCT services which are inappropriate in a 3 month period. Objective 3: Efficient Use of Resources

Measure 10 - Patients able to manage minor illness Percentage of patients seen by health professionals in the team who had a minor illness which could have been managed themselves.

Measure 11 - Patients/clients who do not attend for an appointment Average percentage of total patients' appointments not kept in a week (calculated over a 3-month period).

Measure 12 - Efficient use of administrative systems Percentage of patients not attending appointments with health professionals in the team which result from errors in the administrative system. Measure 13 - Efficient use of GP resources in the team Average number of patients seen by a GPs in a week

Objective 4: Continuous personal and professional development

Measure 14 - Team member access to training Percentage of who are satisfied with the extent to which their training needs are assessed and met in the previous year. Objective 5: High team member commitment, stress and satisfaction

Measure 15 - Team member commitment and satisfaction Percentage of staff in the team who feel committed and satisfied

Measure 16 - Team members use each other's skills, knowledge and expertise appropriately

Percentage of team members who report that skills, knowledge and expertise within the team are used appropriately in 3-month period.
Measure 17 - Effective team working Percentage of requests for help and information and referrals from other team members which are inappropriate in a 3 month period. Objective 6: Responsiveness to client and community

Measure 18 - Patients Experiences of the PHCT service (1) Percentage of patients who report that their experiences of the PHCT services match the range and standard agreed by the PHCT.

Measure 19 - Patients experiences of the PHCT services (2) (Using the existing measure) Percentage of patients whose experiences of the PHCT services meet the standard set by the team. Measuring Performance

The second aim of the qualitative work with primary health care teams was to demonstrate that it was possible for primary health care teams to use effectiveness measures to obtain feedback on performance. Both of the primary health care teams were able to used effectiveness measures to gather feedback information. Below we detail the procedure used by one of the teams to develop a measure of patient satisfaction, gather feedback from patients and then make changes on the basis of this feedback

Measure 19 = Percentage of patients who report that their experiences of the PHCT services match the standard agreed by the PHCT.

The measure was developed by the design team as follows.

The team listed all of the services they provided (e.g. consultations with a health care

professional, clinics, district nursing and health visiting services) and also considered features of the delivery of services which they believed would be associated with patients satisfaction (e.g. short waiting times, prompt repeat prescriptions, phone answered quickly, access to advice).

A questionnaire was developed which enabled patients to report their experiences of the services and the features associated with satisfaction (see Appendix III, p..). The areas covered by the questionnaire were: waiting time to see a GP; waiting time for the phone to be answered; waiting time for a repeat prescription; waiting time to see a practice nurse; awareness of health visitor services, waiting times at health visitor clinics; and waiting time for district nurse visits. Patients were also asked to

provide comments on how different services they had experienced could be improved.

Patients were asked factual questions about their experiences, not for opinions.

For example:

The last time you wanted an appointment with any of the GPs, how soon did you get one? Same day Next day After 2 days [ ] [ ] [ ]

Longer_________

The last time you asked for a repeat prescription, how long did you have to wait to get it? 1st time ___________ days. Not Applicable [ ] 2nd time___________ days

The team identified additional patient information that would help to understand the information collected on patients experiences (age, gender, number of visits to the surgery in the previous month).

Before distributing the questionnaires the team determined the standards they wanted to achieve. For each question they decided what would be acceptable and unacceptable responses, and the standard they would like to achieve. They

determined the percentage of patients they would expect to experience the service in a particular way, the percentage that was unacceptable, and the percentage that would be an ideal. For example, the team decided what percentage of patient they would expect to see a GP on the same day, the next day, after 2 days, or after a longer period of time, the percentages for each which was unacceptable and the percentages that they would like to achieve.

Over a one week period all patients (or for children, their carers) attending the surgery were asked to complete a questionnaire. 100 questionnaires were sent to home addresses, and an additional 100 distributed via district nurses and health visitors. The information from patients was collated and a mean score calculated for each item on the survey.

The score for each item was then compared with the expected standard, unacceptable standard and ideal standard, and the differences between the actual mean and these percentages calculated. This provided the team with feedback

about the extent to which the experiences of patients matched the standards the team were trying to achieve, where experiences fell below standards, and where they were achieving the ideal standard. Results from the patient satisfaction survey

Responses were received from 320 patients which provided a valuable source of feedback on the services provided by the primary health care team. On many aspects the reported experiences of patients matched or exceed that of the standards set by the team. Where the reported experiences fell below the team's standards the reasons for this were explored by the design team and changes made to the provision of this service. For example, the survey revealed that 50% of

patients had waited for between 10 and 15 minutes to see the practice nurse after their appointment with the GP. waiting a maximum of 5 minutes. This was below the target set, 90% of patients

Two main reasons were identified for the longer waiting time: patients were not clear about the procedure for seeing the practice nurse after their GP consultation; and there were insufficient consultation rooms to accommodate the patients who needed to see these nurses. Two changes were proposed to reduce the waiting time.

Firstly, an information slip explaining the procedure for seeing the practice nurse was

produced which GPs could give to patients when they referred them to the nurse. Secondly, the use of consultation rooms was reviewed. An antenatal clinic, which used two consulting rooms, was held at the same time as the morning surgery. It was proposed that this clinic was run at a different time thus providing two additional rooms the practice nurses could use for patient consultations. Discussion

The aims of the qualitative research were to develop agreement among primary health care professional about the objectives for primary health care, to develop measures that would provide feedback on the extent to which effectiveness was being achieved, and to demonstrate that primary health care teams could develop and use effectiveness measures.

The national workshops brought together a wide range of primary care stakeholders; representatives from district nursing, health visiting, general practice, practice nursing, midwifery, mental health, professions allied to medicine, social services, health authorities, the Department of Health, NHS Executive, NHS trusts, patient actions groups and academia. During the course of the workshops these

stakeholders, who had differing aims, objectives, priorities and philosophies of care were able to reach agreement about the objectives for primary health care. The practitioners at the workshops judged these objectives relevant and useful, as did members of primary health care teams who were consulted during the development process.

Using the objectives developed in the workshops, the ProMES approach was used with two primary health care teams to develop measures. This stage of the research also involved working with multidisciplinary groups of health professionals who developed a set of effectiveness measures that could be used to assess the team's performance. These measures were used by the teams to get feedback on how effectively they functioned, and, as illustrated in the example discussed above, this feedback was used to improve the quality of care to patients and to use the resources available to the team more efficiently.

The second stage of the research demonstrated that it is possible for primary health care teams to develop and use effectiveness measures. Further work is required to improve and refine the measures developed by the primary health care teams, and to

test their generalisability for primary health care teams in a range of settings.

The qualitative research has demonstrated that a combination of the constituency approach and ProMES provides a practical method that can be used to help primary health care teams clarify their objectives and to obtain feedback on the effectiveness of the services provided. This will help health professionals to prioritise resources and to deliver high quality, cost-effective health care.

Chapter 5
Community Mental Health Teams Research Methods and Sample Details
Introduction NHS secondary mental health care is delivered primarily through multidisciplinary community mental health teams (CMHTs). These face many challenges. They are tasked with complex statutory and professional responsibilities (Peck & Parker, 1998). The demands of a primary care-led NHS often conflict with the policy

imperatives of the sensitive area of risk management relating to severe mental health problems (Onyett, 1995). In addition, the voice of service users gains strength,

adding to workload and pressures. Team members are employed within two very different bureaucracies; those of health and social care, and come from diverse professional backgrounds. However, the development of joint commissioning

approaches between health and local authority social services requires them to function as integrated teams (Hannegan, 1999). Their constituent professions may jibe at the adjustments this requires (Mistral & Velleman, 1997), for which their training may not prepare them well.

The current policy agenda is increasingly outcomes-focussed. Accordingly, CMHTs are required to monitor their performance (Bhugra, Bridges, & Thompson, 1995) and effectiveness, as a strong commitment to monitoring and evaluation is considered essential for adequate management of CMHT services (Carter Evans, Crosby, Prendeergast & De Sousa Butterworth, 1997). The competition for resources amongst elements of health and social care provision requires that each provide data to demonstrate the value of its contribution. More positively, effectiveness measures may also bring some clarity to teams' efforts to chart their own progress towards meeting diverse expectations.

The organisation of CMHTs is central to their functioning (Bhugra, et. al., 1995; Onyett, 1997). Their core rationale is to bring together a range of professions in order to deliver more effective care co-ordination than could be achieved without an integrated, multidisciplinary team. Achieving that integration is by definition an

organisational task (Onyett, 1995; Pincu, Zarin & West, 1996), requiring that the team

be more than the sum of its diverse constituent members acting individually. This task is rendered considerably more challenging by the need for multi-agency working across the health-social services divide (Department of Health, 1995).

User and carer perspectives are increasingly important. The National Health Service Patients' Charter for Mental Health Services (Department of Health, 1997) sets out rights and expected standards of service for users and potential users of these services. It aims to ensure that the NHS 'listens and acts upon people's views and needs'. A continuing push for users and carers to be involved in decisions relating

to mental health care (Faulkner, 1997), and also to be included at the level of planning and developing services, presents a further challenge to teams which deliver integrated care within the Care Programme Approach (Department of Health, 1990).

There is a growing international literature on CMHTs as a mode of delivery of mental health care. Within the UK, a notable source of this has been the Sainsbury Centre for Mental Health. The key issues with emerge from the research literature include the following, for each of which a representative citation is provided: The many managerial, professional and clinical barriers to effective multidisciplinary teamwork (Peck & Norman, 1999). The importance of integrated operational management of CMHTs (Onyett, 1997). Leadership, integration and agency as key precursors of effectiveness (Grusky, 1995). The threats to effectiveness arising when resource constraints lead teams to over-emphasise control and efficiency at the expense of creative thinking and innovation (Drolen, 1990). The mismatch between current training arrangements and current and future service needs (Sainsbury Centre for Mental Health, 1997). The specific leadership skills required by CMHTs, in which training is necessary (Reed, 1995; Sluyter, 1995. The highly demanding nature of CMHT work (Prosser, Johnson, Kuipers, Szmukler, Bebbington & Thornicroft, 1996). Detriments to morale and effectiveness from excessive workload (King, LeBas & Spooner, 2000).

As the Health Team Effectiveness research programme was nearing completion, the National Service Framework for Mental Health (NSF; Department of Health, 1999) was published. This seeks to establish national standards for mental health care. Within the NSF, national support for local action includes workforce planning, education and training. This aims to enable mental health services to ensure that their workforce is sufficient and skilled, well led and supported, to deliver high quality mental health care. A Workforce Action Team (WAT) has been established to

provide national leadership in developing and taking forward the workforce action plan. We have identified within the WAT interim report (dated April 2000) several themes that a study of mental health team-working can usefully address: Education and training: What are the training requirements for effective teamworking and how might these be met? Recruitment and retention: What are the salient features of team composition? What factors are associated with staff turnover, and how might retention be improved? Leadership: How does this impact on quality of care? How can it be best developed? Primary care: What characteristics of primary health care teams are conducive to high-quality mental health care? Professionally non-affiliated staff: What can the contributions of support workers tell us about the potential for further development of non-affiliated staff?

Before offering answers to some of these questions, we describe the methods used in our research.

The research with Community Mental Health Teams (CMHTs) had two stages. The first was a questionnaire survey, and related data collection methods, involving large numbers of teams to gather data on team inputs, processes and outputs. The

second stage involved intensive examination of a sub-set of teams to explore in more depth targeted team processes and outputs. An overviews of the methodology for stage 1 is given in Figure 5.1.

Figure 5.1: Details of CMHT research methods stage 1

Additional questionnaires/

Sample size 113 teams 1443 respondents

Survey data Team composition Team functioning Team effectiveness Team innovation Stress

Telephone interviews Team composition Team meetings Team management Decision making Clinical systems management

External ratings Team effectiveness Team innovation

Summary of Research Methods

A. National workshop to derive CMHT effectiveness criteria

B. 113 Community Mental Health Teams Survey of all team members Questionnaires or telephone interviews with team leaders External ratings of team effectiveness External ratings of innovations introduced by the teams

C. 10 Community Mental Health Teams

Videotaping and analysis of team meetings Caseload analysis and client selection Interviews with practitioners on two occasions, 6 months apart Use of HoNOS to record client outcomes User and Carer Service Satisfaction Questionnaires

Quantitative methods The Sample The research design required data to be gathered from 100 CMHTs. Initially, chief executives of 101 community mental health trusts in 4 regions, Northern and Yorkshire, North West, Trent, and North Thames, were approached, to inform them of the study and to encourage participation of all CMHTs managed by that trust. The aim was to limit the geographical spread while accessing representative CMHTs, in terms of different socio-economic locations, skill mix and client base. Three months after the first mailshot, follow-up letters were sent to all trusts not responding. Of the 101 approached, 81 responded: 11 had no community adult mental health services; 12 declined to participate and the remaining 58 provided names and contacts for all

CMHTs managed.

The main reasons for not participating were either that (a)

caseloads were such that teams were too busy (3 trusts); or (b) the Trust was in the process of reorganisation (7 trusts); or (c) the teams were already taking part in other research (2 trusts).

With the CMHT names provided we made direct contact with 162 CMHTs, inviting participation in the study after consensus to participate had been achieved within each team. The final number of participating teams was 113 from 45 trusts. Details of the sample are shown in Table 1. At different stages of the access procedure, it was open to Trusts or CMHTs to refuse to participate; the sample was therefore made up of volunteering CMHTs. We performed a post hoc check on socio-

economic representativeness, which indicated that the whole range of deprivation scores was represented (Mental Illness Needs Index (MINI) range 91.3 (low need) to 118.5 (high need), mean 103.3). Data collection Methods Data on team functioning and effectiveness were collected using three methods: self report questionnaires completed by individual team members; self report or telephone interviews with team leaders; and external ratings from community health care representatives, social services and health authority staff.

The named contact for each of the 113 participating CMHTs provided a comprehensive list of all team members, which included all personnel attending regular team meetings. Survey questionnaires were sent to 1925 named individuals, with returns from 1450 (75%). The return rates for professional groups were:

administrative staff 57%; community psychiatric nurses 82%; occupational therapists 83%; psychiatrists 55%; clinical psychologists 90%; social workers 53%; and support workers 68%. Overall, 925 women (64%) were included in the sample, and the mean age was 40 (SD 8.37).

Team leaders from 91 CMHTs provided information on team context, team composition, team processes, and clinical management in a self-report questionnaire survey. Questionnaires completed by individual team members This questionnaire was in four sections (the CMHT survey is included in Appendix I).

Section 1: Team working This contained six measures of team working. Four of these were drawn from the Team Climate Inventory (Anderson and West, 1994; 1998) that is based on a welldeveloped theoretical model of team functioning (West, 1990). The four measures assess levels of team participation clarity and commitment to team objectives emphasis on quality support for innovation

Two other measures were included:

reflexivity, the extent to which team members reflect upon their team objectives, strategies and processes and make changes accordingly (West, 1996; Swift & West, 2000)

team innovation, the extent to which the team has introduced innovations in objectives, work strategies, processes and relationships

Respondents were also asked to describe the major changes or innovations introduced by the team in their work in the previous 12 months. Section 2: Effectiveness These included 27 measures of community mental health team effectiveness derived at a stakeholder workshop (Rees, Stride, Shapiro, Richards & Borrill, in press; Richards & Rees, 1998). Three underlying dimensions were evident: team working patient/client orientation organisational efficiency

Section 3: Team member stress This included a measure of psychological stress, the GHQ-12 (Goldberg, 1972). The GHQ-12 is widely used as a screening tool for detecting minor psychiatric disorder in the general population, and in occupational mental health research. It covers

feelings of strain, depression, inability to cope, anxiety based on insomnia, lack of confidence and other psychological problems. Within a Department of Health funded study of the mental health of the NHS workforce, the GHQ-12 showed good validity against a psychiatric interview (Hardy, Shapiro, Haynes & Rick, 1999).

Section 4: Biographical information This section included questions on biographical and team characteristics (e.g. age, gender, ethnic origin, job title, employer, team composition, team leadership). Additional Team Information Survey This was completed by the team leader or co-ordinator in 92 CMHTs, and combined the PHCT data collection via survey and telephone interviews. Besides the

information on decision-making and communication systems, data were collected relating to the clinical systems the CMHT implemented for dealing with referrals, both emergency and routine, and for accessing inpatient beds. External ratings team effectiveness Each of the 113 CMHTs in the survey sample was approached to nominate three professionals external to the team, within the local Trust, Social Services, Health Authority, or GP practices in their catchment area, in order to collect corroborative data on team effectiveness. Thirty-three teams nominated up to 4 external judges each. Judges ratings were made using the same 27 effectiveness dimensions that team members had used to rate their teams effectiveness. External ratings team innovation Two experts known to the research team rated the descriptions of changes or innovations introduced in each CMHT over the previous 12 months, and which team members had described in their questionnaire responses. The changes were rated on the following dimensions (West & Anderson, 1996): magnitude, how great would be the consequences of changes introduced radicalness, to what extent the status quo would change novelty, how new in general were the changes impact, to what extent changes would improve CMHT effectiveness

Sample Details The aim of the research programme is to determine which team characteristics are associated with good team functioning and team effectiveness. In this section, we describe characteristics of the CMHT sample that participated in the survey components of the research programme. Following the model used to guide this research see Chapter 1) we describe four categories of inputs:

team characteristics indicated by size, mean age, mean tenure, gender mix, ratio of full-time members, length of time the team had been in existence. team composition indicated by the distribution of occupational groups within the team. team task indicated by the MINI (high or low deprivation scores), how quickly the CMHT saw emergency referrals, pooling of referrals, the use of a single integrated set of client case notes, whether waiting lists were in operation for client assessment.

team environment indicated by how the CMHT was commissioned and the English NHS region within which the team was located.

We also describe the relationships between these four domains of inputs.

The

reader will also find an account of team members perceptions of leadership in the team. Team Characteristics Number of team members In terms of the number of individuals employed within each team, this ranged from 6 to 51. The mean size was 17.04 members, SD 7.99. Distribution of sample team size is shown in Figure 5.2 below. Age and gender distribution Overall, 925 women (67%) were included in the sample. Figure 5.3 shows the

percentage of women in CMHTs. The mean age was 40, SD 8.37. Across the 113 CMHTs, only one CMHT was made up of only women. The age distribution appears normal, but it is noteworthy that there are very few CMHT workers below 30 or above 50 years old. This age profile resembles that of qualified nurses.

Figure 5.2: Distribution of team size (number of team members) across the sample

35 30

Number of members

25 20 15 10 5 0 up to 10 11 to 15 16 to 20 21 to 25 > 25

Number of Teams

Figure 5.3: Percentage of women in CMHTs in the sample

35 30 25
% of women

20 15 10 5 0 80 to 100% 70 to 79% 60 to 69% 50 to 59% 25 to 49% % of teams

Tenure All team members indicated how long they had been in the CMHT. Mean tenure across teams was 37 months, SD 19 months. This relates to the short life of one third of teams in the sample, which had been in existence for less than 2 years.

Whole time equivalents

Taking account of hours worked to calculate team size in terms of whole time equivalents shows that team size varied from 5.5 to 48.5. The mean size was 15.81 WTEs, SD 7.53. The small size of the mean difference between numbers of

members and WTEs suggests that the great majority of team members were employed on a full-time basis. Twelve per cent of CMHTs in the sample were

comprised solely of full-time workers. Across the sample, mean percentage of fulltime workers was 77.49, SD 13.74. Length of time CMHT in existence CMHTs were formally introduced on a national basis in 1990, to provide integrated care in the community for mentally ill people. In this sample, the length of time the teams had been existence varied from 6 months to 7 years. For subsequent

analysis, these were categorised as less than 2 years (n = 31); from 2 to 5 years (n = 36); and 5 or more years (n = 25).

Team composition Figure 5.4 gives the breakdown of the sample by professional group and by gender (n = 1363).

Figure 5.4: CMHT occupational groups by gender WOMEN ADMIN CPN OT PSYCHIATRY COUNSELLING/PSYCHOLOGY SOCIAL WORK SUPPORT WORK OTHER TOTAL 181 316 92 27 43 129 68 62 918 MEN 7 214 14 55 17 83 22 32 444

As expected, the largest occupational group was nursing, making up 39% of the sample. The next largest groups were social work (16%) and administrative staff (14%). Occupational therapy (8%), support work (7%), psychiatry (6%) and

psychology/counselling (4%) were the smaller occupational groups. As compared with the overall preponderance of women, who formed two-thirds of the respondents, administrative staff were, unsurprisingly, even more predominantly female. Almost 90% of occupational therapists were women. In contrast, two-thirds of psychiatrists were men. The gender mix of nursing and social work showed a modest

preponderance of women.

At the team level, multidisciplinary mix was as shown in Figure 5.5. Noteworthy here is the fact that just 12% of teams included members from all disciplines (psychiatry, social work, psychiatric nursing, clinical psychology, occupational therapy).

Figure 5.5: Multidisciplinary mix within CMHTs

All disciplines All disciplines except psychiatry All disciplines except psychology/OT All disciplines, no SW CPN plus SW CPN plus 2 other Health CPN Psychiatry and SW CPN plus one other Health

10

15

20

25

30

35

Number of CMHTs

Team Task Mental Illness Needs Index (MINI) MINI scores for the Health Authority areas within which CMHTs populations were based ranged from 91.3 (low need) to 118.5 (high deprivation), mean 103.34, SD 6.91. Figure 5.6 shows the distribution of CMHTs across Health Authorities with low (30%), medium (40%) and high (30%) deprivation scores. Figure 5.6: Teams in areas of high of high, medium and low need as indicated by the MINI

high need

low need

medium need

CMHT response to emergency referrals CMHTs indicated how quickly, on average, emergency referrals were seen. Elapsed time before emergencies were seen ranged from within the hour to within two weeks. Across all teams, the mean wait for emergencies was 26 hours, SD = 44 hours. Within-CMHT pooling of referrals CMHTs provided information on how referrals, other than emergencies, were dealt with. Five CMHTs did not pool referrals; 20 CMHTs pooled some referrals; and 63 CMHTs pooled all referrals. Use of integrated case notes In 40 CMHTs, each discipline kept client case notes separately; in 12 CMHTs, notes were separate but available for reference by other disciplines; and in 34 CMHTs, each client had one integrated set of case notes. Assessment waiting list implementation Fifty-three CMHTs indicated that they did not operate a waiting list prior to assessment, and 36 CMHTs indicated that they did operate a waiting list.

CMHT Organisational context NHS Region

CMHTs were sampled from four NHS regions in England. The participating regions contributed 32, 26, 32 and 23 respectively. To safeguard the anonymity of the

participating teams, these regions are not identified in this report.

Local commissioning arrangements There were three models of commissioning for these teams: 25 CMHTs were commissioned by Health Services only; 39 by Health and Social Services jointly; and 33 by Health and Social Services separately. CMHTs are constituted on a multiagency basis between health and social services. Most CMHTs necessarily combine staff working within the management structures of each of the two agencies, tasked with meeting the objectives of both. However, each CMHT must function as a

coherent entity working towards mutually agreed objectives and following mutually

understood

and

functionally

interdependent

practices.

Local

commissioning

arrangements may impact on integrated team functioning. CMHT constructed process and outcome variables Apart from scales derived from survey items, for example, those from the Team Processes Inventory, CMHTECQ and GHQ, others were computed to measure the clarity of team leadership, within-team variation in relations to the clarity of team leadership, the teams efforts to communicate other than in a formal meeting environment, and the teams turnover. Clarity of CMHT leadership All team members provided information about the clarity of leadership in the CMHT. Team means were aggregated from the single survey item: Does the team have a single clear leader or co-ordinator? where yes scored 1 and no entry was scored 0. Members of 13 CMHTs in the sample were unanimous in reporting that

their team had a single clear leader, while members of six CMHTs were unanimous in declaring that the team had no single clear leader or co-ordinator. The aggregated measure was treated as a process variable. Within-team variation in relation to the clarity of team leadership Blaus index of variation was used to calculate the extent of disagreement within the team about the clarity of leadership. Five variables with values ranging from 0 to 1 were constructed for the proportion of each team giving each of the five possible responses. The resulting variable was treated as a process variable. Internal Communication In the Additional Team Information survey, team leaders were asked two questions to indicate (a) how much team members had access to information other than that conveyed in meetings, for example, with the use of memos, whiteboards, newsletters; and (b) how much social activity team members participated in together. Responses were combined to give a measure of the CMHTs intent to communicate both informally and socially, on a scale of 1 (poor) to 5 (high quality effort to communicate). Figure 5.7 shows how teams varied on this dimension, which was treated as a team process variable.

Figure 5.7: CMHTs' Intent to Communicate

50

40

30

20

number of teams

10

0 1 to 2 2 to 3 3 to 4 4 and above

interdisciplinary communication

Turnover In the Additional Team Information survey, team leaders were asked to indicate how many staff had left the team in the previous 12 months. Turnover was computed as the percentage of staff in the team (size) who had left, and was treated as an outcome variable.

Relationships between input dimensions Team size Figure 5.8: Scatterplot of CMHT climate against team size
4.5

4.0

self report: combined TCI scales

3.5

3.0

2.5

2.0 0 10 20 30 40 50 60

team size going by team members declared

Three CMHTs were larger than all others, and scatterplots (e.g. Figure 5.8) showed that on some dimensions of team functioning these were outliners. They were removed from all analysis involving team size. Team size was clearly associated with the diversity of disciplines within the team: larger teams were more likely to have psychiatrists (r = .20, p < .05) and psychologists (r = .25, p < .01). Larger teams also had a lower percentage of full-time staff (r = -.22, p < .05). Team size was associated with only one aspect of the teams task or organisational environment: larger CMHTs were more likely to operate a waiting list for assessment, r = .29, p < .01.
Team average age

The average age of the CMHT was likely to be greater if social workers (r = .30, p < .01) and psychologists (r = .25, p < .01) were members. However, it was likely to be lower if the MINI score of deprivation was high (r = -.36, p < .001).

Team average tenure Average job tenure was lower if there were social workers in the team (r = -.20, p < .05). Predictably, average job tenure was strongly associated with the length of time the team had been in existence (r = .51, p < .001). Length of time CMHT in existence

Beyond the relationship with average job tenure in the team, the life of the team was associated with whether it operated at separate sites (r = -.23, p < .05), that is, the longer the team had been in existence, the less likely was it that staff were based at different locations. Percentage of full-time workers We saw above that larger teams were likely to have a higher percentage of part-time workers. Such part-time practitioners were likely to be psychiatrists (r = -.26, p < .01), psychologists (r = -.31, p < .01) and occupational therapists (r = -.28, p < .05). Although they described themselves as being part-time, these disciplines may well have divided their time between the CMHT in question and other responsibilities.
Percentage of women in the team

A lower percentage of women in the team was associated with a higher deprivation rating, R = .21, p < .05.
Further effects of team composition When psychiatrists were in the team, it was less likely that the team would implement a single, integrated set of case notes for each client (R = -.32, p < .01), but if an occupational therapist was in the team, the opposite was the case (R = .31, p < .01). Again, psychiatrists were more likely to be in the team if the deprivation rating was high (R = .23, p < .01). If there were social workers in the team, emergencies were likely to be seen more quickly (R = -.27, p < .05). The team was more likely to operate a waiting list for assessment if there were social workers (R = .28, p = .01), psychologists (R = .23, p < .05) or occupational therapists (R = .23, p < .05) in the team. Further effects of clinical system implementation Where the team pooled referrals at a central point, they were also likely to operate a single, integrated set of case notes for each client (R = .25, p < .05), although a single referral point was also associated with lower deprivation scores (R = -.29, p <

.01). Where the deprivation score was higher, waiting lists for assessment were less likely to be in place (R = -.26, p < .05). Intensive analysis Selection procedure

Teams rating themselves as highly effective or as highly ineffective, in comparison to the self-ratings of the full sample of 113 teams, were selected. This recruitment strategy was designed to maximise the power of betweenteam analyses to detect associations between intensive analysis variables and effectiveness. We computed aggregate team scores on the CMHT teamworking questionnaire scales. These comprised the Team Climate Inventory (TCI) participation, support for innovation, and task orientation scales; the user orientation, use of resources, and internal process scales from the service delivery effectiveness (SDE) items; and the 12-item GHQ. The standardised team scores on the TCI and the SDE were summed, and the standard team score on the GHQ-12 subtracted from this total. This algorithm yielded 14 teams above the 80th percentile and 14 below the 20th percentile. To achieve a target sample size of 16 teams in this phase of the study, all 28 were invited to participate after the team had gained consensus amongst members. Ten volunteered, distributed across the 4 NHS regions as follows: A, 3; B, 2; C, 1; and D, 4. MINI scores for the 10 teams covered a wide range, 91.3 to 110.0, with a mean of 101.4. Three of the 10 had rated their activity as effective, leaving 7 who rated their team as ineffective. We followed up the 4 teams rated as effective which had not already responded, but failed to increase the number. To a considerable degree, therefore, this was a self-selected sample. The 10 teams comprised three self-rated as effective and seven self-rated as ineffective.
Representativeness of sub-sample CMHTs Independent t-tests were used to compare group means on appropriate dimensions, together with Levenes test for equality of variances. This process included comparisons for:

Team characteristics and composition: size; age; tenure; percentage of full-time workers in the team; percentage of men in the team; length of time the team had been in existence; professional mix.

Task environment: MINI; number of GPs linked to the CMHT; ; whether the team held a single, integrated set of case notes for each client; whether referrals were pooled or not; use of a waiting list for emergency referrals.

Team processes: team processes scale; perceived clarity of leadership; intent to communicate informally; amount of meeting time available; perceived quality of relationship with GPs, Health Trust, and Social Services.

Team effectiveness: levels of stress; self-report effectiveness; external judges ratings of performance; external judges ratings of team innovations; self-report innovativeness; turnover.

For the most part, group means were similar, with t values ranging from .07 to 1.65. Exceptions are shown in Figure 5.9. Levenes tests for equality of variances did not reach significance for any dimension. Figure 5.9: Significant differences between survey sample and sub-sample Survey sample mean (SD) 78.48 (13.27) 38.66 (18.89) 0.34 (0.48) 0.96 (0.16) 3.59 (0.38) Sub-sample mean (SD) 67.28 (15.09) 24.62 (20.39) 0.80 (0.42) 1.09 (0.20) 3.25 (0.47) T value probability

Percentage of full-time staff Average tenure of staff (months) Psychologist in the CMHT CMHT stress level (GHQ) External ratings of CMHT effectiveness*
*

2.52 2.23 3.26 2.20 2.07

.01 .03 .007 .03 .05

The Ns on this dimension were 25 in the survey group, 8 in the sub-sample group

These comparisons showed that, in relation to most team characteristics, and most aspects of task environment, team process and effectiveness, the sub-sample group did appear to be reasonably representative of the full survey sample. However, subsample teams had a lower percentage of full-time staff, their members were likely to have been in the team for a shorter time, were more likely to have a psychologist or counsellor in the team, their members experienced a higher level of stress, and teams were rated lower on effectiveness by external judges. This last finding is not surprising, given that more self-ratedly ineffective than effective teams had volunteered to take part in this phase of the research. Chi-square tests showed that the sub-sample teams were representatively distributed throughout the 4 NHS regions, but were not representatively distributed in terms of local commissioning

arrangements, chi square = 6.33, p < .05, with 7 teams being jointly commissioned by Health and Social Services, and 3 by the Health Service only. Descriptives on the 10 CMHTs in the intensive phase Figure 5.10 shows the characteristics of the ten teams. All had CPNs, all were multidisciplinary, although only 3 of the 10 teams had access to psychiatry input within the team. Figure 5.10: Team characteristics of sub-sample CMHTs
Team A 16 44.9 (11.70) 13.58 Less than 2 years Team B 12 37.1 (7.43) 23.56 2 to 5 years Team C 15 34.6 (7.94) 21.07 2 to 5 years Team D 15 41.2 (7.22) 35.73 Over 5 years Team E 22 38.2 (7.44) 10.55 Less than 2 years no yes no yes 36 64 96.6 H&S jointly C Team F 12 40.4 (8.05) 12.00 Less than 2 years no no no yes 62 100 94.8 H&S jointly B Team G 18 45.2 (6.45) 9.80 Less than 2 years No Yes Yes Yes 27 60 97.4 24 H&S jointly B Team H 17 41.9 (8.02) 3.85 Less than 2 years No No Yes Yes 23 77 102.6 35 H&S jointly C Team I 46 43.3 (9.85) 47.11 2 to 5 years Team J 12 41.0 (8.43) 69.00 Over 5 years

Size Mean age Mean tenure (months) Length of life of CMHT

Psychiatry in OT in Psychology in SW in % men % full-time MINI Linked GPs Commissioning NHS region

No Yes Yes Yes 33 83 97.4 34 H&S jointly B

no yes yes no 38 56 110.0 38 H only A

no yes yes no 15 54 110.0 26 H only A

Yes Yes Yes Yes 27 55 91.3 30 H&S jointly B

yes yes yes yes 37 67 120 H&S jointly D

yes yes yes yes 42 58 104.2 35 H only A

It is apparent that there were differences between the teams on all structural dimensions. For example, team 1 stood out as a large team compared with others, and teams B and C were in Health Authorities which had a higher MINI score than others, indicating localities with higher deprivation. All teams had a mix of disciplines, although team F had only social work input, over and above the CPNs who were present in all teams. Percentages of male and fulltime workers in these teams varied widely. Half of the teams had been in existence for less than 2 years, and this partly accounted for shorter tenure of staff, although there is commonly high turnover of staff in the CMHT environment.
Audio and video recording of Community Mental Health Team meetings The ten teams volunteering for the intensive stage research also agreed that their meetings could be recorded; meeting sizes ranged from six to twenty five members. The meetings were multi-disciplinary business meetings in which decisions were

made about the running of the team. This meeting was chosen by the teams as the one to record because it is the main forum, outside clinical meetings, that are multidisciplinary meetings. Wherever possible, we recorded two meetings of the same type for each team. Dates for meeting recording were at the discretion of the CMHT. For details of recording procedures, equipment and transcription see Chapter 2. Longitudinal data collection: clinical outcomes; use of resources; patient and carer satisfaction Two site visits six months apart were arranged with each participating CMHT. Before the first, every practitioner in the team completed a caseload audit summary to define as precisely as possible the population the CMHT was serving (Manchester Audit Tool, recommended by the Kings Fund mental health team). This required a breakdown by 19 classifications of the care workers entire current caseload in terms of diagnosis, severity, complexity and chronicity. Simple guidelines for completing the audit questionnaire were included. For the purposes of patient selection, an individual breakdown was also completed by each practitioner, using either codes or names. Stratified sampling was carried out by the research team using SPSS Stratification ensured that users selected were

random number generation.

representative of (a) the individual practitioners caseload; and (b) the teams entire caseload profile. Stratification was based on scores provided by practitioners for each client for severity, chronicity and complexity. Scores were then summed. For each of the 10 teams, 40 users were selected.

If practitioners indicated that clients were unable to complete a questionnaire, because they were in an acute episode, or could not read, or did not read English well, they were replaced by others on the key workers caseload matched for stratification. Packs of information sheets, consent forms and return envelopes were sent to each key worker for each of their selected clients (and carers where appropriate). Practitioners were requested to discuss the research with clients and hand them information and consent form, and for the purposes of confidentiality, to ask them to send signed consent direct to the research team. Once the research team had received signed consent from client or carer, Patient/Carer Service Satisfaction questionnaires were sent out, to assess the effectiveness and acceptability of the service received, and users quality of life. The questionnaire was that developed by the Tameside & Glossop Rehabilitation team, who gave permission for its use.

First site visit Two researchers visited each team for a day to interview all practitioners with a caseload. At the interview, practitioners answered questions about each of their selected clients individually. Biographical data were collected: gender, age ethnic origin, type of housing, and whether they had a formal carer. Questions relating to a period of the previous 6 months covered use of resources, for example, types and dosage of psychoactive medication, day hospital attendance, number of contacts with CMHT professionals, time spent as an in-patient. Other questions covered

referral, diagnosis, CPA level, care plan targets, whether or not the client was on section, and projected clinical outcomes for 6 months ahead. Practitioners were provided with 6 HoNOS forms for each client, to be completed by the practitioner at each contact over the following 6 months, or up to discharge. Second site visit, after an interval of 6 months Self-report key worker schedules were designed, following the model of the practitioner interviews implemented at the first site visit. The schedules were sent two weeks in advance of the visit to the site to collect completed schedules and HoNOS questionnaires, and respond to queries. All team members completed the survey questionnaire for a second time, so that change over time could be measured.

The costs of use of resources were computed by combining estimates of quantity and cost per unit. Medication costs were based on the 1997 BNF. Contact costs were based on Unit Costs of Health and Social Care (Netten & Dennett, 1997). Elements included in the costs were medication, inpatient days, day hospital attendance, respite care, day centre use, drop-in use, occupational therapy groups, outpatient attendance and contacts with CMHT staff.

Data analytic strategy Psychometric Considerations The self-report measures of community mental health team process and effectiveness used in this research were complex and novel. Accordingly, prior to analysis to address the substantive questions listed above, preliminary psychometric analysis considered the intercorrelations among these measures at the team level. Specifically, the extent of specificity vs. redundancy in these measures had not previously been assessed. To be useful in testing theories relating team processes to effectiveness, the measures would have to demonstrate sufficient specificity that the relationships among them not be most parsimoniously explained as reflecting a

single evaluative or morale factor. As shown in Figure 5.11, all intercorrelations were significant and substantial, ranging from .53 to .89. Figure 5.11: Team-level correlations among self-report measures of CMHT process and effectiveness
TCI: participation TCI: TCI: support clarity of reflexivity innovation for objectives innovation CMHTEQ: CMHTEQ: external internal requirements processes

TCI: support for innovation TCI: support for innovation TCI: clarity of objectives TCI: clarity of objectives Reflexivity Reflexivity innovation innovation SDE: external requirements SDE: external requirements SDE: internal processes SDE: internal processes SDE: monitoring/evidence SDE: monitoring/evidence

.858 .858 .633 .633 .706 .706 .531 .531 .642 .642 .818 .818 .555 .555

.771 .771 .816 .816 .743 .743 .701 .701 .889 .889 .676 .676

.706 .706 .662 .662 .652 .652 .791 .791 .620 .620

.732 .732 .566 .566 .744 .744 .578 .578

.526 .526 .671 .671 .551 .551

.842 .842 .878 .878

.777 .777

All correlations have N = 113, p < .01.

We therefore considered whether the effectiveness of a team as reported by its members could be distinguished from its climate, also as reported by those same team members. We entered team means on the four Team Processes Inventory (TCI) scales and the three SDE scales in a factor analysis. This indicated that a

single factor was the most efficient way to describe the differences among the teams; attempts to force a two-factor solution did not support a distinction between team processes and self-reported effectiveness, as the scales with highest loadings on each factor comprised a mixture of both TCI and SDE scales.

External ratings of effectiveness were available for 33 teams. For this subsample, we considered whether external ratings of effectiveness were any more highly correlated with self-reported effectiveness than with the TCI. There was no such difference. External ratings of effectiveness were as highly correlated with team members ratings of team processes, r = .64, as with their reports of team effectiveness, r = .60.

The quality of innovations described by the members of all 113 teams was rated by external judges. These ratings were no more strongly correlated with team members TCI ratings, r = .44, than with their reports of the teams effectiveness, r = .39. Considering only the most relevant TCI scale, support for innovation was

correlated at r = .48 (indistinguishable from the overall TCI correlation of .44) with external ratings of innovation. The fact that the TCI correlates rather more highly, at r = .64, with external ratings of team effectiveness than with external ratings of innovations, r = .44, is further evidence against specificity. Meanwhile, self-reported innovation (considered an effectiveness dimension) correlated very similarly at r = .55 with external ratings of innovation.

These analyses rather suggest that the self-report measures should be most parsimoniously considered to reflect a single evaluative or morale factor, rather than to tap specific aspects of team process or effectiveness. The measures may not, therefore, be sufficiently precise to reveal subtle relationships between CMHT process and effectiveness. Sequencing of Multivariate Analyses Analysis proceeded in two steps. First, the questions identified above were

addressed in sequence. For each dependent variable in turn, potential predictor variables were entered in groups. For example, for each team process variable in

turn, team characteristics were entered as one set of predictors, before moving on to a second analysis looking at team composition factors as predictors, a third analysis with team task factors as predictors, and a fourth with organisational context factors as predictors.

Secondly, for each class of dependent variables in turn, variables that had emerged from the foregoing analyses as showing predictive relationships independent of the other variables in their respective group were entered into new analyses including all such variables across the groups. Analyses at this second step identified predictors that were independent in their effects of other significant predictors across all classes of predictor. Such predictors warrant closer attention; accordingly, this account of our findings will emphasise this second phase of the analysis.

In the next chapter we describe the results of these analyses.

Chapter 6
Community Mental Health Teams Results from Survey and External Ratings

Summary of Findings
Teams whose members were more positive about team processes are rated as more effective by external stakeholders

Teams whose members agree as to how clearly the team leadership role is defined are rated by external stakeholders as more effective

Teams whose members describe their team processes positively perceive their teams as more effective

Teams whose members report clarity as to the leadership role perceive their teams as more effective

Teams that have been in existence for a relatively long time tend to describe their team as more effective

Larger CMH teams are rated as more innovative by external judges

Teams who perceive their performance as highly reflexive are rated as more innovative by external judges

Teams who perceive their team processes and reflexive behaviour as positive also see themselves as more innovative

Teams who perceive their team processes levels of stress

as poor experience higher

Teams with older members enjoy more stable membership, as do teams including social workers

Teams using integrated client case notes, and teams not including psychiatrists, are clearer about the leadership role

Introduction Our analysis was informed by the input-process-outcome model presented in Chapter 1. This entailed predicting process variables from input variables, and

predicting outcome variables from both input and process variables.

As for the analysis of PHCTs, we addressed two main questions:

Is there an association between the composition of a community mental health team and team processes?

Is there an association between the composition and processes of the community mental health team and the effectiveness of the team?

The team characteristics, team processes and measures of team effectiveness are summarised in Figure 6.1.

Figure 6.1: Team Inputs, Processes and Effectiveness Variables

Characteristics
Individual Age Gender Time in job Time in team Time in team Team Occupational groups Size (number of members) Number of GP's Hours worked Multidisciplinary mix Gender mix Team context Commissioning type MINI index Use of integrated case notes Response time for emergencies Waiting list in operation NHS Region Relationship with Social Services Relationships with GP's Relationship with Trust Team processes

Processes
Team processes Participation Innovation Objectives Emphasis on quality Reflexivity Number of meetings Types of meetings Frequency of meetings Potential time for different disciplines to meet Decision making Leadership Integration and communication in the group

Effectiveness
Team ratings Organisation Team working Patient focus Innovation External ratings (innovation) Magnitude Radicalness Novelty Impact External ratings (effectiveness) Organisation Team working User/carer focus Types of Innovations Quality of Care External collaboration Responsibility for health Use of resources Professional development Team satisfaction Responsiveness Stress (GHQ 12) Turnover

Individual team members rated team processes on six dimensions: participation; innovation; team-objectives; emphasis on quality; reflexivity; and interdependence. Information about decision-making processes, communication, number and types of meetings, who attended these meetings, and how the team was managed was collected from Practice Managers. The information on team meetings was

categorised according to who contributed to operational, strategic and clinical decisions. In addition a new variable integration was developed which assessed the extent to which there were mechanisms within the team to encourage inter disciplinary communication and working.

Team Inputs Information about the team members ages, gender, ethnicity, professional group, tenure and team leadership were collected from each team member. Information was also gathered on team size, hours worked, commissioning type, MINI index, NHS Region, whether the CMHT made use of a single integrated set of client case notes, how long the team took to respond to emergency referrals, whether they operated a waiting list for assessment, and the teams relationships with GPs, Trust and Social Services.

Team Processes Individual team members rated team processes on six dimensions: participation; support for innovation; clarity of team objectives; emphasis on quality; reflexivity; and integration. The variables participation, support for innovation, clarity of team objectives and emphasis on quality were very highly correlated and were combined to form one variable describing team processes. Information about decision-making processes, communication, number and types of meetings, and who attended meetings was collected from team leaders. An index of the amount of time the various disciplines in the team could potentially meet was computed. In addition, a new variable integration was developed which assessed the extent to which there were mechanisms within the team to encourage interdisciplinary communication. Team Effectiveness As with PHCTs, this was assessed using information from a variety of sources. Team members rated their teams effectiveness on the three dimensions of the Community Mental Health Team Effectiveness Questionnaire (CMHTEQ; Rees, Stride, Shapiro, Richards & Borrill, in press), developed within this project: team working; organisational efficiency; and patient orientation. Team members also rated their teams innovativeness and described the innovations implemented by the team in the previous year. External raters assessed the innovations reported by the teams on four dimensions: magnitude; radicalness; novelty and impact on team effectiveness. External ratings of team effectiveness on the CMHTEQ were provided by external experts nominated by the team and based in local GP practices, Social Services, the Trust or Health Authority. Individual team members also completed the GHQ-12 (a measure of mental health or psychological stress). The measures of interest for this report are overall effectiveness of the team in delivery of services, user-centred care and dealing with the demands of the parent organisation (both externally rated and self rated), overall innovation (both externally rated and self

rated),

team

turnover,

and

mental

health

measured

by

the

GHQ-12.

Results The method of analysis was similar to that described for the PHCT data in Chapter 3. Stepwise multiple regressions were carried out, with possible predictors of each dependent variable being split into groups according to type of variable, to identify those which might ultimately predict the dependent variable. As with the PHCT analysis of Chapter 3, to reduce the complexity of the data set and to guard against Type 1 errors arising from multiple statistical tests, we focused on a second-level analysis combining predictors across the groups. Process variables were predicted by team characteristics or inputs; and effectiveness, innovation, turnover and

mental health were conceptualised as outcomes predicted by team characteristics (inputs) and by team processes. In a final stage of the analysis, both team characteristics (inputs) and team processes were considered together as predictors of the outcome variables (effectiveness, innovation, turnover and mental health).

We found that the CMHT data called for a different approach to the issue of team size than we adopted for the PHCT data. It transpired that the relationship between team size and other variables was largely due to 3 outlying teams (with more than 36 members). Accordingly, rather than considering team size first in all analyses as we had done with the PHCT data, we excluded these teams from analyses including the size variable, and thereafter treated team size in the same way as other team characteristics.

Question 1 Is there an association between the composition of a CMHT and team processes? Figure 6.2: Relationships Between Team Composition and Team Processes Dependent variable Consensus on leadership Reflexivity Integration Team processes Potential to meet Number of meetings Predictor variables Team size Tenure Presence of psychiatrist(s) none none none Tenure -.250 -.215 -.225 p .017 .039 .018 R2 .097 .042

.212

.046

.045

Figure 6.2 presents the significant team composition predictors of each team process variable. Larger teams, and teams whose members had been longer

in their jobs in the team, showed less consensus on the clarity of leadership in the team. Teams including psychiatrists as members were less reflexive. The tabled result was obtained after excluding the responses of psychiatrists themselves, which had inflated the observed relationship. This, our preferred analysis, is more conservative because it excludes the effect of the tendency of psychiatrists themselves to rate the team as less reflexive. It may therefore be interpreted as showing an association, albeit modest, between the reflexivity ratings of non-psychiatrist team members and the presence of psychiatrists within the team.

Question 2 What Affects the Effectiveness and Innovativeness of a Community Mental Health Team?

Figure 6.3: Relationships Between Team Composition and Processes, and Ratings of Effectiveness
Dependent variable Overall effectiveness (external) Predictor variables Team processes Consensus on leadership Team processes Consensus on leadership Length of time the team had been in existence .643 -.311 .861 .163 .121 p <.001 .046 <.001 .003 .020 R2 .413 .487 .742 .766 .780

Overall effectiveness (selfrated)

As shown in Figure 6.3, two variables predicted, independently of one another, the ratings of CMHT effectiveness given by their local stakeholders: team processes and lack of consensus relating to team leadership. In other words, teams whose

members rated their team processes favourably, and teams whose members agreed amongst themselves as to how clearly the leadership role was defined within the team, were rated by external stakeholders as more effective. Figures 6.4 and 6.5 below show these predictive relationships graphically.

Figure 6.4: Impact of team processes on externally-rated team effectiveness (data from 32 CMHTs)
5.0

4.5

4.0

3.5

3.0

2.5

2.0 2.0 2.5 3.0 3.5 4.0 4.5 5.0

Team Processes

Figure 6.5: Impact of clarity of leadership on externally-rated team effectiveness (data from 32 CMHTs)
5.0

4.5

external rating of team effectiveness

4.0

3.5

3.0

2.5

2.0 -.2 0.0 .2 .4 .6 .8 1.0 1.2

clarity of team leadership

Figure 6.3 also presents the three factors predicting self-reported team effectiveness on the 27-item Community Mental Health Team Effectiveness Questionnaire (CMHTEQ), each independently of the other two: Overall score on the Team

processes Inventory, as shown in Figure 6.6; clarity of leadership, as shown in Figure 6.7; and the length of time the team had been in existence, as shown in Figure 6.8. In other words, members of teams whose members described their processes positively, members of teams whose members reported clarity as to the leadership role, and members of teams that had been in existence for a relatively long time, all tended to describe their team as more effective.

Figure 6.6: Impact of team processes on selfreported effectiveness (data from 113 CMHTs)
4.5

4.0

3.5

3.0

2.5

2.0 2.0 2.5 3.0 3.5 4.0 4.5

team processes

Figure 6.7: Impact of clarity of leadership on self-reported effectiveness (data from 113 CMHTs)

4.5

4.0

3.5

3.0

2.5

2.0 -.2 0.0 .2 .4 .6 .8 1.0 1.2

clarity of team leadership

Figure 6.8: Impact of length of time team in existence on self-reported effectiveness (data from 113 CMHTs)
4.0 3.9 3.8 3.7 3.6 3.5 3.4 3.3 3.2 3.1 3.0 2 years or less 2 to 5 years 5 or more years 3.3 3.5

3.4

Length of time team in existence

Relationships between Team Composition and Processes, and Ratings of Innovation

As shown in Figures 6.9 and 6.10, two variables predicted, independently of one another, the quality of the innovations reported by team members as rated by independent, expert judges: Reflexivity, beta = .51, t = 6.72, p < .001; and team size, beta = .38, t = 5.00, p < .001. In other words, teams whose members rated their teams as highly reflexive, as well as larger teams, described innovations that were judged to be of higher quality.

Figure 6.9: Impact of reflexivity on expert ratings of innovation quality (data from 113 CMHTs)
5 4

external rating of team innovations

0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0

reflexivity

Figure 6.10: Impact of team size on expert ratings of innovation quality (data from 113 CMHTs)

external rating of team innovation

0 0 10 20 30 40

team size

Three factors, acting independently of one another, were associated with the innovativeness reported by team members: self-reported reflexivity, beta = .45, t = 3.75, p < .001, as shown in Figure 6.11; overall score on the Team Processes Inventory, beta = .32, t = 2.65, p = .009, shown in Figure 6.12; and freedom of interaction (scheduled co-presence of the different disciplines at meetings), beta = .15, t = -2.20, p = .03. In other words, teams whose members reported a high level of reflexivity, teams whose members reported team processes as positive, as well as teams scheduling relatively little cross-disciplinary interaction in formal meetings, all tended to be described by their members as relatively strong with respect to innovation.

Figure 6.11: Impact of team reflexivity on self-reported innovativeness (data from 113 CMHTs)
4.5

4.0

3.5

innovativeness

3.0 2.5

2.0

1.5 1.0 2.5

3.0

3.5

4.0

4.5

5.0

5.5

6.0

team reflexivity

Figure 6.12: Impact of team processes on selfreported innovativeness (data from 113 CMHTs)
4.5

4.0

3.5

innovativeness

3.0

2.5

2.0

1.5 1.0 2.0 2.5 3.0 3.5 4.0 4.5

team processes

Relationships between Team Composition and Processes, and Stress Levels in CMHTs Three variables predicted, independently of one another, the level of stress reported by team members on the General Health Questionnaire: Team processes (overall score on the Team Processes Inventory), beta = -.50, t = -5.22, p < .001; informal communication (social events, message boards, etc.), beta = .21, t = 2.20, p = .03; and freedom of interaction (scheduled co-presence of the different disciplines at

meetings), beta = -.20, t = -2.20, p = .04. These findings indicate that the following team characteristics are associated with relatively high stress (or low stress levels) amongst staff: a positive team processes; plentiful opportunities for informal interaction among members; and relatively little provision for scheduled, formal, cross-disciplinary encounters at team meetings. Figure 6.13 shows the impact of team processes on stress.

Figure 6.13: Impact of team processes on stress

1.6

1.4

1.2

1.0

.8

.6 2.0 2.5 3.0 3.5 4.0 4.5

team processes

GHQ = General Health Questionnaire, 12 item version

Relationships between Team Composition and Processes, and Turnover in the team
As shown in Figures 6.14 and 6.15, two compositional variables predicted staff turnover, independently of one another: mean age of team members, beta = - .25, t = -2.21, p = .03; and the presence of social workers in the sample of respondents, beta = -.23, t = -2.02, p = .05. These findings suggest that teams with older members

enjoyed more stable membership, as did teams including social workers.

Figure 6.14: Impact of mean age of team members on turnover (data from 113 CMHTs)
1.6

1.4

1.2

1.0

.8

.6 2.0 2.5 3.0 3.5 4.0 4.5

team processes

Figure 6.15: Staff turnover in CMHTs with and without social workers (data from 92 teams)
20

15

15

10

8 5

0 no yes

social workers in the team

Relationships between Team Composition and Leadership

Two factors, acting independently of one another, were associated with the clarity of leadership described by team members: the use of a single, integrated set of case notes for each client, beta = .29, t = 2.85, p = .006, as shown in Figure 6.16; and the presence of one or more psychiatrists amongst team members completing the questionnaire, beta = -.27, t = -2.63, p = .01, as shown in Figure 6.17. It should be noted that, before computing the mean clarity of leadership for each team, responses from psychiatrists themselves were removed from the latter analysis. In other words, teams using integrated case notes, and teams not including psychiatrists, were clearer about the leadership role.

Figure 6.16: Integrated case notes and clarity of leadership (data from 92 CMHTs)
.8

.8 .7 .7

.6

.5

.5

.4 No partial access Yes

Use of one integrated set of case notes

Figure 6.17: Psychiatric membership of the team and clarity of leadership (data from 113 CMHTs, with responses of psychiatrists themselves removed from the analysis)

.8

.7 .7

.6

.5

.5

.4 no yes

psychiatrists in the team

Discussion As for primary health care teams, the findings from this stage of the research reveal a clear message for CMHT policy and practice, in relation to the factors predicting the effectiveness and innovations of community mental health teams: positive team processes, and clarity as to the leadership role within the team, make for a more effective team, as judged by external stakeholders as well as the team members themselves; requisite size makes for a higher quality of innovation; reflexive processes aids innovation; longer-established teams are rated more innovative by external judges and see themselves as more effective.

We have considered policy implications of the findings reported in this chapter in relation to the Workforce Action Team issues identified in Chapter 1. In terms of education and training, it is clear that teamworking skills are key to the effective delivery of mental health care in the community. trainable. These skills are specific and

They are not acquired implicitly through professional socialisation into

such disciplines as nursing, medicine, clinical psychology, or social work. Nor are they attained through unfocussed, unsustained team-building exercises of the kind that are widely marketed into the NHS and other large organisations. Rather, they comprise key types of knowledge, skill and ability required for effective teamworking (Stephens & Campion, 1994). As depicted in Figure 6.18, these fall into 5 domains: conflict resolution; collaborative problem-solving; communication; goal-setting and performance management; planning and task co-ordination. Both initial professional training and continuing professional development for CMHT members should incorporate systematic training in these. To prepare and enable members of the

constituent professions to function effectively within CMHTs, the NHS should require

that the majority of such training should be delivered in multiprofessional rather than uniprofessional learning environments. The requirement for such delivery may

create difficulties for HEIs but the NHS should work with them to ensure that training funded by the NHS meets the NHSs pressing requirements for effective multiprofessional teamwork.

Figure 6.18: Knowledge, Skill and Ability Required for Effective Team-working (Stephens & Campion, 1994)
Conflict resolution: Recognise and encourage desirable but discourage undesirable team conflict Recognise type and source of conflict confronting the team and implement appropriate resolution strategy Employ integrative (win-win) negotiation strategy rather than traditional distributive (win-lose) strategy Collaborative problem-solving: Identify situations requiring participative group problem solving and utilise proper degree and type of participation Recognise obstacles to collaborative group problem solving and implement appropriate corrective actions Communication: Understand networks and utilise decentralised networks to enhance communication where possible Communicate openly and supportively, sending messages which are (1) behaviour - or event-oriented; (2) congruent; (3) validating; (4) conjunctive; and (5) owned Listen non-evaluatively and appropriately use active listening techniques Maximize consonance between nonverbal and verbal messages, recognise and interpret the nonverbal messages of others Engage in ritual greetings and small talk, and recognition of their importance

Goal-setting and performance management: Help establish specific, challenging and accepted team goals Monitor, evaluate, and provide feedback on both overall team performance and individual team member performance Planning and task co-ordination: Co-ordinate and synchronise activities, information and task interdependencies between team members Help establish task and role expectations of individual team members and ensure proper balancing of workload in the team

Stephens, M.J., & Campion, M.A. (1994). The knowledge, skill and ability rquirements for teamwork: Implications for human resource management. Journal of Management, 20, 503-530.

In relation to recruitment and retention, we note the disturbing fact that only 12% of our CMHTs included all 5 key disciplines: nursing, psychiatry, social work, occupational therapy, and clinical psychology. This presents a challenge to effective delivery of the full spectrum of mental health care, and confirms the importance of the Workforce Action Teams agenda. In that context, we draw attention to certain

features of the demographics of CMHT staff identified by this research. Sixty-seven per cent of CMHT staff were women; the mean age of the staff was 40, with a standard deviation of 8.4 years, and most workers aged between 30 and 50. This profile highlights the importance of flexible working to accommodate family demands and thereby retain staff. It also confirms the importance of retaining CMHT staff beyond the age of 50. Within the somewhat restricted age range we observed, teams with older members experienced less turnover. Turnover was greater among smaller teams, suggesting that teams should be large enough to provide sufficient support. Our findings implicate poor team processes in CMHT staff stress, which is likely to be inimical to staff retention. Longer-established teams rating themselves as more effective suggests that stability of the team itself may yield greater job satisfaction through the experience of effectiveness.

In relation to leadership, we see this as a key and integral feature of team functioning and hence, as shown by our data linking team processes to effectiveness, vital to the delivery of effective mental health care. We found that clarity in relation to the

leadership role was reflected in external ratings of CMHT effectiveness, as well as in team members own ratings of their teams effectiveness. Clear and effective

leadership will be essential to delivery of the National Service Framework, and development of the required leadership skills, which are learnable irrespective of professional discipline, will require training resources. Training for CMHT leadership must relate to the complex multi-agency environment, and combine clarity with flexibility and innovativeness. Such training needs to be evidence-based, locally

available, and ongoing rather than occasional or intermittent.

In relation to primary care delivery of mental health care, we draw attention to findings from Chapter 3 highlighting the benefits of larger PHCTs, and of a wide spectrum of professions being members of the PHCT, alongside the value of clear PHCT leadership. In the course of our PHCT research we obtained ample anecdotal evidence that PHCT members are highly aware of the challenges presented to them by their growing responsibilities in this area.

In relation to the Workforce Action Teams interest in developing the role of professionally non-affiliated staff in mental health care delivery, we have some indicative findings on support workers within CMHTs. They comprised 7% of respondents to our survey, and 75% of them were female. Their ratings of their teams were very favourable, and interviews with them during the intensive analysis reported in Chapter 7 were highly positive in tone. The intensive analysis also

revealed the considerable dependence of CMHTs on support workers for the timeintensive, practical aspects of care of patients with severe and enduring mental health problems. Accordingly, we strongly endorse the development of

professionally non-affiliated staff as a resource within CMHTs.

Chapter 7
Community Mental Health Teams Results from Qualitative Research
Summary of findings Basic minimum standards of staffing and hence care are not yet universally fulfilled by NHS mental health care

The costs of mental health care vary across teams providing it, over and above the apparent clinical requirements of the caseload as reflected in diagnosis and severity

CMHTs face conflicting demands from primary care and from the needs of patients with severe and enduring mental health problems

Venturing beyond our immediate data, we suggest that such key issues are likely to have important effects on the morale, stress and effectiveness of CMHT staff and on their capacity to initiate and maintain the innovative, collaborative and flexible patterns of teamworking whose importance is highlighted by our survey findings

Sub-sample team characteristics In this section, we report an analysis of the sub-sample of CMHTs taking part in the second phase of the project. In this qualitative work we looked in greater depth at CMHT effectiveness by taking account of clinical and economic performance. Its objectives were:

To develop methods to assess the economic costs and clinical outcomes for a representative patients seen by a CMHT To examine the relationships between task context (severity of caseload), inputs (time, costs), team processes, and outputs (clinical improvement, goal attainment and CMHT practitioner stress)

Process and effectiveness dimensions of the 10 teams

Figure 7.1 gives the aggregated team means for the primary process and effectiveness variables.

Teams were selected for the second phase of the research on the basis of combined standardised team processes scores, self report effectiveness score, and stress levels (see Chapter 5). Figure 7.1 shows that the 10 teams varied on the

performance dimensions as effective or ineffective: Teams B, G and J perceived their teams as performing effectively, but this was only partly supported by external judgements of team performance, in which innovativeness, but not effectiveness, was rated highly for those teams.

Figure 7.1: Process and effectiveness dimensions of the 10 CMHTs

Team A Overall Team processes Max = 5 Number of meetings Max = 11 Multidisciplinary communication Max = 5 Team Mean stress level: GHQ item mean Max = 3 Clarity of team leadership Max = 1 Team relationship with GPs Max = 5 Team relationship with Trust Max = 5 Team relationship with SS Max = 5 Assessment WL implemented Integrated case notes Referrals pooled Self report effectiveness Max = 5 External rating effectiveness Max = 5 External rating of innovations Max = 5 2.95

Team B 3.85

Team C 3.20

Team D 2.96

Team E 3.06

Team F 3.46

Team G 4.24

Team H 3.30

Team I 2.92

Team J 3.94

2.75

3.50

3.50

3.00

2.00

3.50

1.50

2.50

3.00

2.75

1.36

0.95

1.33

0.99

1.08

1.35

0.85

1.10

1.03

0.83

0.67

1.00

0.77

0.09

0.77

0.89

0.87

0.15

0.22

0.83

no

no

no

yes

no

no

yes

no

yes

No

yes All 3.10

yes some 3.70

yes all 3.28

no all 2.82

no some 2.66

no all 3.41

no all 4.00

no all 2.72

no some 2.85

Yes All 3.65

3.13

3.28

2.85

3.12

4.19

3.19

3.59

2.67

2.75

3.63

2.25

1.00

2.50

2.25

3.13

2.50

3.00

3.38

Caseload profile But how does all this relate to the quality of care provided by this sub-sample of CMHTs? In the intensive phase of the research, team practitioners were asked to:

Complete a caseload analysis describing the diagnosis, severity, chronicity and complexity for all clients. These factors were used to (a) describe the teams caseload profile; and to (b) select a representative sample of around 40 CMHTs clients following procedures defined by the research team (see Methods)

Approach selected clients, and if appropriate their carers, and invite them to participate in the research by completing a Service Satisfaction questionnaire Participate in interviews structured to elicit biographical, resource use, and clinical information about the teams 40 selected clients, retrospectively for the past 6 months. Data on resource usage were collected relating to number of contacts with practitioners in the team; number of inpatient days; use of day hospital, day care or drop-in; medication; outpatient appointments

Complete the HoNOS monthly for selected clients over the following 6 months or up to discharge To provide information during a second site visit relating to resource use and clinical outcomes for the teams sample of representative clients.

Clinical and cost data were collected for 372 CMHT clients for the first 6 months of the 12-month period, although an attrition rate of around 35 per cent meant that, for the second 6 months, data for only 241 of those clients were available. The high attrition rate was accounted for primarily by key workers or care co-ordinators leaving the CMHT, and other key workers or care co-ordinators being unable to supply clinical data during the second site visit.

Psychiatrists completed a caseload analysis in only two teams, and for Team J in particular, this increased the number of clients on the teams caseload. Figure 7.2 shows the client caseload team by team, in terms of severity and diagnostic group. Oneway ANOVA tests showed that some teams differed in the severity of their caseloads F(9,362) = 4.44, p < 001. Post hoc Bonferroni tests indicated that Team Bs caseload (mean 2.18) was significantly milder than Team Es (mean 2.70), p = .003; than Team Is (mean 2.64), p = .019; and Team Js (mean 2.62), p = .021; and marginally milder than Team Hs (mean 2.65), p = .054.

Figure 7.2: Client caseload profile by severity and diagnostic group6

Diagnostic category Substance misuse 3 Depression 42 Anxiety 50 Psychosis 55 PD 14 Substance misuse 3 Depression 43 Anxiety 51 Psychosis 85 PD 9 Substance misuse 3 Depression 43 Anxiety 44 Psychosis 85 PD 12 Substance misuse 8 Depression 59 Anxiety 8 Psychosis 54 PD 8 Substance misuse 6 Depression 24 Anxiety 9 Psychosis 141 PD 18 Substance misuse 3 Depression 86 Anxiety 45 Psychosis 68 PD 8 Substance misuse 7 Depression 38 Anxiety 30 Psychosis 79 PD 8 Substance misuse 8 Depression 41 Anxiety 24 Psychosis 154 PD 17 Substance misuse 0 Depression 85 Anxiety 42 Psychosis 144 PD 14 Substance misuse 29 Depression 180 Anxiety 43 Psychosis 170 PD 16 Mild 5 Moderate 87 Severe 82 TOTAL 174

Team A

Team B

33

97

78

208

Team C

20

83

92

195

Team D

28

57

70

165

Team E

56

136

198

Team F

52

90

74

216

21

91

70

182

Team G

Team H

10

88

153

251

Team I

38

84

200

322

Team J

25

174

236

435

Some disorders (e.g. adjustment to disability) presented in low numbers, therefore were not included
6

During the site visits, all CMHTs emphasised that they experienced a tension between policy requirements that their case load included enduring mentally ill people, and the referral patterns of local GPs who continued to refer all adults with mental health problems. The tension was increased by a third demand in some localities that CPNs from within the team should conduct clinics in primary care practices. Most teams had attempted to clarify with GPs the appropriate referrals, but reported little guidance from their Trust management, and little change in types of GP referrals.

Varying responses to these conflicting demands emerged from the data.

An

important finding was that the three teams (B, C and F) carrying caseloads which had clients who were relatively moderately ill, all provided clinics in primary care, and inclusion of their primary care patients accounted for their lower overall caseload severity.

Mental health status, health economic costs and clinical outcomes Mental health status, first 6-month period

In this section we present more detailed information about the sample of mental health status clients selected by teams and their clinical outcomes. The period was 12 months, although as stated there was a 35 per cent attrition rate for cases. Some of the selected clients who completed the

service satisfaction questionnaire criticised continuity of care, as a result of practitioners leaving the team and being allocated to a new key worker. Figure 7.3 gives a summary of the mental health status of a sample of clients selected in terms of severity and CPA level. These data were collected 6 months into the review period. Comparing the information in Figure 7.3 with Figure 7.2 above shows that the severity of selected clients' mental health condition was broadly representative of teams entire caseload profiles. Looking at Team J, for example, Figure 7.2 showed that 236 (54%) clients were severely ill; Figure 7.3 below shows that 28 (62%) of Team Js selected clients were severely ill.

Figure 7.3: Summary of selected clients mental health status

Mild Team A Team B Team C Team D Team E Team F Team G Team H Team I Team J 1 5 3 3 1 3 3 0 1 0

Moderate 14 23 17 12 10 14 19 9 12 17

Severe 19 12 16 26 29 19 13 17 26 28

CPA level 1 12 22 11 6 9 12 19 5 6 -

CPA level 2 13 7 21 14 21 16 13 9 23 -

CPA level 3 6 7 1 6 4 5 1 3 1 -

On section 2 2 0 3 4 3 2 4 3 2

Not on CPA 3 3 1 13 5 0 0 9 9 -

During the site visits, it emerged that teams used different criteria for determining CPA level. Some teams had produced guidelines for assigning CPA level, while others had not. Most teams used the labels level 1 as least severe, but others used level 1 as most severe. One team in the sample used level A as the most severe. These had been recoded to represent severity as lowest, level 1, highest, level 3. Thus how CMHTs used the CPA was non-standard, and indicated variation in the management of CPA.

More importantly, in terms of the role of CPA in maintaining service standards, we observed firstly that planned CPA reviews for selected clients were often overdue, and secondly, that part of the difficulty was getting access to all the practitioners who it was specified in the work plan must be present at the review. Some teams had a particular difficulty with accessing psychiatry.

Across the teams, the percentage of CMHT clients who had a formal carer, in the sense that their key workers judged that clients needed carer support to live in the community, ranged from 22 to 55 per cent across the teams. We found that many clients did not have a care plan with formal targets; however, this varied across teams and was related to severity of caseload. Figure 7.4 gives a team breakdown of selected clients diagnosis and severity.

Figure 7.4: Selected clients diagnosis and severity

Diagnostic group Team A Substance misuse Depression Anxiety Psychosis/PD Substance misuse Depression Anxiety Psychosis/PD Substance misuse Depression Anxiety Psychosis/PD Substance misuse Depression Anxiety Psychosis/PD Substance misuse Depression Anxiety Psychosis/PD Substance misuse Depression Anxiety Psychosis/PD Substance misuse Depression Anxiety Psychosis/PD Substance misuse Depression Anxiety Psychosis/PD Substance misuse Depression Anxiety Psychosis/PD Substance misuse Depression Anxiety Psychosis/PD

Mild

Moderate

Severe 1 1 5 9/2

TOTAL

1/0 2 3

1 6 6/1 6 8 8/1 7 5 3/1 3 7 2/0 1 8/1

33

Team B

3 8/1 2 1 12/1 2 7 2 12/2 2 1 25/1 1 5 2 10/1 2 3 7/1 3 2 10/0 1 5 6 10/4 13 1 13/1

40

Team C

35

Team D

1 2

40

Team E

40

Team F

2 0/1

Team G

5 3 5/0 2 3 6 8/0

35

35

Team H

8/0 1 7 4/0 9 4 4/0

23

Team I

38

Team J

45

Figure 7.4 illustrates that, although these 10 teams were similar in terms of diagnostic categories for which care was provided, the numbers seen within the serious and enduring categories varied. This was most apparent in team E, in which

83 per cent of selected clients were diagnosed with psychosis, in contrast to team G, in which a diagnosis of anxiety was almost as prevalent as psychosis.

Health economic costs

Figure 7.5 gives a summary of costs 6 months into the period. Costs are derived as mean cost per client, and rounded to the nearest pound. Economic costs for inpatient days, outpatient appointments, day care sessions, and contact with practitioners were calculated using Unit Costs of Health & Social Care compiled by Ann Netten and Jane Dennett at the PSSRU University of Kent at Canterbury. We commissioned advice on the analysis and interpretation of these data from the Centre for Health Economics, University of York.

The following assumptions were made when calculating costs:

In-patient days were calculated at 136 per day

Outpatient appointments were calculated at 97 per appointment

Day care costs were calculated at 32 per session. One session equates to half-a-day. All contacts reported were assumed to be one session. Day care included day hospitals, day centres, drop-ins and workshops

Calculations for contacts with practitioners were based on a generic cost for all members of the Community Mental Health Team. The unit cost used was face to face contact calculated at an hourly rate. All contacts were assumed to last one hour. Both a minimum cost of 26 per hour and a maximum of 50 per hour were calculated

Medication costs were calculated using the British National Formulary. Costs of generics were used in all calculations except where these were not available

Contact costs have been computed using the minimum generic costing (26 per contact), so this is an underestimation of the cost of contacts, although consistent across all teams

Figure 7.5: Summary of health economic costs: first 6-month period

Medication Team A Team B Team C Team D Team E Team F Team G Team H Team I Team J 247 199 210 100 150 169 193 225 180 325

Contacts 2655 950 939 2140 5001 1138 2530 2727 3129 999

Day care 1156 842 1043 1304 1294 1003 1197 1816 1092 440

OP appointments 120 99 119 213 172 136 77 131 246 229

Inpatient days 1572 126 162 1012 3515 382 1480 1496 1806 48

TOTAL 5750 2216 2473 4769 10132 2828 5477 6395 6453 2041

One way ANOVA was used to determine significant differences between teams. Significant differences were apparent in terms of outpatient (F = 3.94, p < .001), inpatient (F = 3.41, p < .001) and practitioner contact costs (F = 3.01, p < .01). caseload Using univariate analysis of covariance, controlling for differences between teams remained statistically

severity,

significant, as shown in figures 7.6 to 7.8.

Figure 7.6: Inpatient costs, first 6 months Source Corrected Model Intercept SEV TEAM Error df 10 1 1 9 359 F 3.785 1.014 6.688 2.854 Sig. .000 .315 .010 .003

Figure 7.7: Outpatient costs, first 6 months

Source Corrected Model Intercept SEV TEAM Error

df 10 1 1 9 231

F 4.192 .570 7.039 2.972

Sig. .000 .451 .009 .002

Figure 7.8: Contact costs, first 6 months Source Corrected Model Intercept SEV TEAM Error df 10 1 1 9 353 F 6.178 3.166 32.308 2.648 Sig. .000 .076 .000 .006

Post hoc Bonferroni tests indicated that Team I had significantly higher outpatient costs than Teams B, C and G. Post hoc tests also indicated that Team E had significantly higher inpatient costs than teams B, C, F and J, and that Team E had significantly higher practitioner costs than teams A, C and J.

Pearsons R correlations were carried out on caseload severity with all health economic costs for the first 6-month period. Medication costs were not associated with other costs or caseload severity. Practitioner contact costs were associated with outpatient (Pearsons R .14, p < .01) and inpatient costs (Pearsons R .25, p < .01). Practitioner contact costs were also associated with caseload severity, Pearsons R .30, p < .01. Outpatient costs were associated with inpatient costs, Pearsons R .17, p < .01. Caseload severity was also associated with outpatient costs, Pearsons R .12, p < .05, and inpatient costs, Pearsons R .18, p < .01. Mental health status, second 6-month period

Whereas in some CMHTs we were able to collect cost and clinical data for most of the sample of representative clients for the second 6-month period, in others there was high attrition. Figure 7.9 shows the attrition rate within each team.
Figure 7.9: Summary of selected clients lost at second 6-month period
First 6 months Team A Team B Team C Team D Team E Team F Team G Team H 34 40 36 41 40 36 35 26 Second 6 months 21 27 17 12 32 29 31 7 % lost 38 33 53 71 20 19 11 73

Team I Team J TOTAL

39 45 372

26 39 241

33 13 35

Clients were divided into two independent groups, the first made up of clients for whom data were collected only in the first 6 months, the second consisted of clients for whom we collected data at both time points. Using independent sample t-tests, we found that over the first 6 months clients retained in the sample were marginally less severe (mean = 0.48) than those lost to the sample over the second 6 months (mean = 6.80; t = 1.70, p = .09).

For those clients who were on CPA at the first time point, and remained in the sample, 116 remained at the same level, 25 moved to a lower level, and 63 moved to a higher level. Health economic costs, second 6-month period

In Figure 7.10 below, costs are shown for the second 6-month period, based only on the clients still in the sample, and calculated as described above.
Figure 7.10: Summary of health economic costs: second 6-month period
Medication Contacts Day care OP appointments 185 86 102 162 173 171 122 83 224 259 Inpatient days 648 584 0 533 1466 1913 715 2273 675 883 TOTAL

Team A Team B Team C Team D Team E Team F Team G Team H Team I Team J

136 179 188 136 258 269 440 436 310 258

433 216 177 295 597 267 265 505 307 223

4302 551 452 891 747 164 364 1563 411 340

5704 1616 919 2017 3241 2784 1906 4860 1927 1963

One way ANOVA was used to determine differences between the teams in terms of costs. Significant differences were found in practitioner contacts and day care costs, but not for other costs. Post hoc Bonferroni tests indicated that practitioner contact costs were accounted for by Team E being significantly higher than all others except A, D, and H (F = 4.50, p < .001).

Day care costs were significantly higher for Team A than all other teams, except teams C and H (F = 2.75, p < .01). Pearsons R correlations were carried out on client caseload severity with all health economic costs for the second 6-month period. There were clear associations between severity level and all costs except medication: with practitioner costs, R = .30, p < .001; with outpatient costs, R = .23, p < .001; with inpatient costs, R = .20, p < .01; and with day care costs, R = .13, p < .05.
Clinical outcomes, first and second periods combined It was not until the end of the 12-month period that practitioners were asked to judge whether clinical targets for each selected client had not been met at all, had been partly met, had been fully met, or had been exceeded. Often, given the diagnosis and chronicity of many of these clients, the targets were simply that they should be stable or maintained in the community. Sometimes, though, there was a clear

objective to discharge the client, and tests were carried out to see how many of the proposed discharges subsequently took place.

At the team level, in order to determine team differences, client severity level and whether predicted targets were met were analysed using Oneway ANOVA. The analysis indicated that Team E clients continued to be

significantly more severely ill (mean 7.19) than those of Teams B (mean 5.63) and G (mean 5.77), F = 2.97, p < .01. A marginally significant difference was found in the level of predicted targets met (F = 1.66, p = .10) and post hoc Bonferroni tests indicated that Team Js perceived performance in meeting predicted targets was marginally better than Team Cs.
Team-level tests were carried out to determine whether the costs incurred by different teams were related to the diagnostic profile of their clients. Univariate

analyses of covariance were carried out, controlling for the effects of diagnosis and severity, as shown in Figure 7.11.

Figure 7.11: Costs and clinical targets met controlling for diagnosis and severity

Dependent variable

Effects of diagnosis

Effects of severity

p Inpatient costs F <1 ns F 1.70

Team effects (controlling for diagnosis, severity) F (11,351) p

.19

3.06

< .01

Outpatient costs Df 11,227 Medication costs Df 11,266 Overall costs Df 11,352 Clinical targets met

3.00

.08

6.24

< .05

3.39

< .01

<1

ns

1.56

ns

1.30

ns

<1

ns

6.65

.01

3.68

< .001

3.94

< .05

<1

ns

1.81

.07

Taking into account variation across teams in diagnosis and severity of clients, teams were significantly different in terms of some health economic costs incurred, although not for medication costs.
Inpatient costs (computed at 136 per day)

In relation to inpatient costs, Team J (mean cost per client 48) made little use of inpatient facilities; teams B and C (means 125 and 167 respectively) also incurred low inpatient costs. Inpatient costs for Team E, some of whose clients were severely psychotic and at high risk, were higher than all others (mean 3515).
Outpatient costs (computed at 97 per contact) Outpatient costs were considered only for those clients who attended appointments. As expected, both diagnosis and severity had an effect on outpatient costs, with more serious diagnoses and higher levels of severity positively associated with higher costs. At the team level, H and B (mean cost per client 97 and 161

respectively) incurred lower costs than did teams I (mean 589) and J (mean 499). Medication costs

Medication costs were not influenced by diagnosis or severity, and teams did not differ significantly in terms of their outlay on medication.

Overall costs

Overall cost per client was strongly influenced by severity, though not by diagnosis. Teams differed significantly in relation to overall costs, with Team E (mean 5001) higher than all other teams, a dimension partly accounted for by the high use of support workers visiting clients at home, sometimes daily, and high inpatient costs. Teams B, C and J (means 950, 964 and 999 respectively) incurred lower overall costs than the other teams in the sample. We saw above that Teams B and C had relatively more moderate caseloads than other teams, but this did not apply to Team J.
Clinical targets met

Teams differed marginally in terms of practitioners judgements of whether clinical targets were met, although this was also influenced by client diagnosis. Teams B and J (mean per client 2.67 and 2.74 respectively)

scored relatively high on this dimension, while Teams A, C and H scored low (means 2.25, 2.07 and 2.25 respectively). However, a score of over 2.00 indicated that for the average client, clinical targets had been at least partly met.
Psychiatry input A serious problem for this sample of CMHTs was their lack of effective input from psychiatrists. This was more of a problem for teams with a higher number of

severely ill clients, for example, Teams E and I. How this problem manifested varied across the teams. One team reported that they had direct access to a psychiatrist for over 2 years. Other teams had negative views about access to medical input. Their clients had to visit outpatient departments for CPA reviews, rather than this being conducted at the team base. Some teams reported that although a psychiatrist was based within the team, and clients attended clinics at team premises for some CPA reviews, these reviews were not conducted as frequently as specified in the clients care plan, because no medical input was available. It was difficult to find locum psychiatrists so long term absence or secondment also resulted in a lack of psychiatric cover for the team.

Team processes, team performance, team stress, clinical outcomes, user satisfaction and cost effectiveness

At the team level, in order to determine how the different dimensions of effectiveness were related, correlation analyses were performed on relevant compositional (age, tenure, severity of caseload, number of linked GPs, MINI), process (team processes, clarity of leadership, ability to communicate across disciplines, relationships with GPs, Trust and Social Services, pooling of referrals, how quickly emergencies are seen), performance (overall self report and external evaluation, external evaluation of innovations) and outcome (targets met, overall costs, and user satisfaction variables). associations were apparent, as shown in Figure 7.12. Some

These data must be

viewed as exploratory and interpreted very cautiously, in view of the probability of both Type I and Type II errors: calculating such a large number of correlations invites Type I errors, whilst the small sample of teams incurs substantial risk of Type II errors.
Figure 7.12: Summary of associations across composition, process and outcome
Association Caseload severity/annual costs Caseload severity/self report effectiveness Teams feel ineffective when they have severe caseloads Clinical targets met/team stress level Teams whose members feel stressed also report disappointing clinical outcomes External evaluation of innovation/clarity of team leadership Good team leadership recognised by innovation raters Team processes/self report effectiveness User satisfaction/how quickly emergency referrals are seen MINI/team uses integrated case notes Integrated case notes tend to be used in more deprived areas Pearsons R .763 -.655 probability .010 .040

-.628

.052

.560

.091

.890

.001

.789 .766

.011 .016

Referrals pooled in the team/number of linked GPs A central referral system in response to larger N of GPs CMHT relationship with Trust/CMHT relationship with Social Services Tenure/external rating of effectiveness (n = 8)

.708

.049

.642 .699

.045 .054

The survey results suggested that good processes and outcomes are associated with: small team size few part-time workers Health Service only commissioning single, clear line of leadership or co-ordination rapid response to emergency referrals effective communication processes external judgement about its effectiveness if the team itself rates its functioning highly

When we looked in more depth at exemplar CMHTs, we found a wide variety in practice. During the researchers' visits to the 10 CMHTs in four NHS regions, the team numbers spent an average of around 20 hours with the research team, which yielded rich anecdotal evidence in support of the hard findings at both survey and intensive stages of the study.

Size We found that in one very large team CHMT where three smaller teams had been created, and separate meetings were held for nursing staff and social workers, co-ordination and communication were problematic. The wider team met only once every two months to debate and decide team policy and practice.

Single, clear leader or co-ordinator Seven of the sub-sample teams had a clear leader or co-ordinator, but in three teams the lack of clarity about leadership was problematic. One team had been without a leader for over two years, which was felt to be an indicator of under-resourcing and lack of support from the local commissioners. Practitioners agreed that this situation also made team meetings difficult, not only in terms of process, but in terms of the struggle to implement and communicate decisions.

Rapid response to emergency referrals

One of the primary agenda items at

CMHT meetings was the implementation of duty systems to cover emergency and urgent referrals. These clients often had to wait longer than practitioners felt was ideal. Such new systems were also described by many team members across the sample as one of the major innovations their teams had implemented in the previous

12 months. Of course, some of these teams had only been brought together within the previous 12 months.

Effective communication processes Observation of team meetings indicated wide variation in quality of team communication. In general, meetings had an agenda, either formally written and circulated before the meeting or informally presented at the start of the meeting. Most teams kept to the agenda and covered all business. However, there were wide differences in process. In the most effective teams,

interaction was quick, responsive and supportive, and participation was equal; in some teams though most people attending did not take part and merely listened in. The issue then arises that those people who do not participate in discussion or decisions do not feel they own decisions and are slow to implement them.

Inclusion of social workers At the statutory level, social workers must be involved in the care of the CMHT client group. Our survey findings indicated that social workers tended to rate their teams effectiveness lower than other disciplines. What we discovered in carrying out the study was that the research process itself was not so thoroughly underwritten by Social Services employees as by Health personnel. At some site visits, we also observed the negative attitudes of health personnel towards social workers, and vice versa. In open-ended statements in the survey, this crossdisciplinary hostility was evidenced in many teams, partly because nurses were expected to take on the duties previously seen as only related to social work, for example, giving housing or benefits advice. However, in the most effective teams in the sub-sample, social workers were well integrated to provide the delivery of care for this client group.

Discussion

This intensive analysis of a sub sample CMHTs draws attention to some key issues in delivering mental health care, as well as providing a demonstration methodology for looking in detail at the effectiveness of services delivered in terms of participant evaluations in relation to health care costs. However, in view of the small number of teams we were able to study at this level of detail, our substantive findings cannot be interpreted as more than tentative.

The intensively-analysed teams were selected on the basis of team member ratings of team effectiveness, team processes, and personal stress. The 3 teams whose

members perceived them as effective were seen as more innovative, but not as more effective, than the remaining teams in the sub sample.

We found that teams varied in the perceived severity of their caseloads, in the proportion of their caseloads who were suffering from severe and enduring mental health problems, and in the health care utilisation costs incurred in the treatment of patients on their caseloads. Two of the 3 teams with relatively low overall costs per case had relatively high proportions of only moderately severe cases. Health care utilisation costs differed across teams even when controlling statistically for the variation in caseload severity across the 10 teams. The different categories of cost incurred (CMHT practitioner contacts, outpatient costs and inpatient costs) were positively inter correlated across teams. This suggests that teams differ in terms of their use of more or fewer services of all kinds, rather than differing in the priority or availability of the 3 types of service in the care packages delivered by each team. Only medication costs were unrelated to the other categories of cost.

Further analyses controlled statistically for both diagnosis and caseload severity and showed that inpatient, outpatient, and overall costs all differed across the 10 teams. In terms of meeting the clinical targets set by keyworkers themselves, however, teams differed only marginally ( p = .07) when diagnosis and severity were controlled. The associations between severity and costs were well-illustrated by the team with the highest costs being the one that judged its patients to be most severely ill.

In the course of collecting these data we made important supplementary observations concerning the targeting, integrity, and likely effectiveness of CMHT care. The most significant problem was the frequency of overdue Care Programme Approach reviews, suggesting that CPA is often not implemented effectively. This was attributed to the unavailability of psychiatric input. This echoes the finding from the larger survey sample that only 12% of CMHTs included all core disciplines. However, effective CPA requires availability of all relevant staff for review meetings, a requirement that goes beyond the mere inclusion of relevant disciplines in the membership of the team. A further problem was the threat to continuity of care

arising from the staff turnover that accounted for most of the 35% attrition when we returned to collect data for the second 6-month period.

We were also forcibly reminded of the tension experienced by CMHTs between policy requirements to focus on severe and enduring mental health problems and the demands of GPs continuing to refer many adults with mental health problems. This was reflected in the finding that the 3 teams with relatively moderately ill caseloads all provided clinics in primary care, with inclusion of their primary care patients accounting for their lower overall caseload severity. Fulfilment of CMHTs remit under the National Service Framework will require primary care to shoulder its full burden in relation to the less severe but considerably more prevalent disorders with which it is tasked by the framework.

Finally, this intensive analysis of clinical data lent some further validation to the measures used in the full sample of CMHTs: teams with relatively severe caseloads considered themselves less effective; teams whose members felt under stress also reported disappointing clinical outcomes; users of services provided by teams reporting rapid response to emergency referrals were more satisfied with their teams service.

Although not definitive, on account of the small sample, this intensive analysis highlights key issues in mental health service delivery.

Chapter 8
Secondary Health Care Team Research methods and Sample Details
Introduction Three studies were conducted with secondary health care teams. The first examined the relationship between team processes and team member stress. The second explored whether team membership per se, was associated with stress, and if so, what factors in the team environment accounted for this association. The third was a longitudinal study of the relationship between team functioning and subsequent member turnover from the team. The Sample in Study 1 in Secondary Care The Association between Team Membership and Sress. The purpose of this study was to determine whether team membership conferred upon NHS employees better mental health than did membership of loose working groups or a work situation which did not involve working in a team or loose staff grouping. Four Trusts were selected from nineteen Trusts included in a larger study of stress in the National Health Service (Borrill et. al., 1998) and 4,500 names were selected from the Trusts staff lists. The next step involved selecting individuals from these hospitals for possible involvement in the study. For small occupational groups, where the number of possible respondents was a hundred or fewer, all those on the staff lists supplied by the hospitals were included. For larger occupational groups, individuals were randomly selected from staff lists, with a minimum proportion of 20%. Hence, greater proportions were sampled from smaller occupational groups and from smaller hospitals. A total of 4,500 names was thus selected. Individually addressed questionnaires were distributed either by hand to the persons area of work or through the internal post. 2,263 people returned completed questionnaires, representing a response rate of 50%. Questionnaires were sent to all those selected for the sample The questionnaire was in four sections:

Section 1: Biographical information Respondents were asked to give information about age, gender, marital status, number of children, number of other dependants, job title, length of service, hours of work.

Section 2: Work Characteristics Scales were used to measure perceived job, work and hospital characteristics (e.g., work demands, autonomy, role conflict, influence in decision-making). Full details of these measures can be found in Haynes et al., (in press) and are available from the first author of this report. Measures of organisational climate were also included, which examined 12 dimensions of climate.

The climate measure we employed is based on the Competing Values Model of organisational effectiveness (Quinn & Rohrbaugh, 1981; Hill, 1998). This model

posits two fundamental organisational dimensions: internal versus external orientation and emphasis on control versus an emphasis on flexibility. These two orthogonal dimensions create four domains of organisational emphasis:

Rational Goal Approach - external focus with tight internal control Open Systems Model - external focus and flexible relationships with the environment Internal Processes - internal focus with an emphasis on tight internal control Human Relations - emphasis on well-being, growth and commitment of employees.

These approaches reflect the rich mix of competing views and perspectives within an organisation and Quinn (1988) argues that a balance of these competing organisational values is required for organisational effectiveness.

The organisational climate questionnaire (Hill, 1998) was developed by selecting scales from an existing team processes measure considered appropriate to service organisations. Following extensive pilot work in four NHS Trusts, this Organisational Climate Questionnaire (OCQ) was used to survey 5,275 health service employees from 27 Trusts (Hill et. al., 1997). The results of this survey suggested seven core

dimensions: innovation, performance monitoring, autonomy, co-worker co-operation, training, communication and resources (Hill, 1998). Innovation This dimension measures the extent to which the Trust is seen to be responsive to change. More specifically the scale explored the extent to which senior staff were interested in suggestions and the development of new ideas. This was a six-item scale with five intervals and three stems ranging from strongly agree to strongly disagree. An example item from this scale is New ideas are readily accepted in the Trust. The coefficient alpha for the current sample was 0.91. Performance Monitoring This dimension addresses the perception of how adequately job performance is monitored within the Trust as a whole, and how well staff are informed about their work performance. The scale consisted of five items with five intervals and three stems ranging from strongly agree to strongly disagree. An example item from this scale is Staff performance is measured on a regular basis. The coefficient alpha for the current sample was 0.86. Autonomy This dimension measures the extent to which employees feel that they have the freedom to work in their own way and are given adequate scope and responsibility to work without constant upward consultation. The scale has six items with five intervals and three stems ranging from strongly agree to strongly disagree. An example item from this scale is Management tightly control the work of those below them. The coefficient alpha for the current sample was 0.87.

Co-worker co-operation This factor measures the extent to which there is co-operation and conflict amongst staff in the Trust. The scale comprised of six items with five intervals and three stems ranging from strongly agree to strongly disagree. An example item from this scale is People can rely on one another in the Trust. The coefficient alpha for the current sample was 0.87. Training This dimension measures the employees perceptions of degree of emphasis within the Trust on skill development and the availability of training resources. The scale

comprised of six items with five intervals and three stems ranging from strongly agree to strongly disagree. An example item from this scale is Staff are strongly encouraged to develop their skills in the Trust. The coefficient alpha for the current sample was 0.86. Communication This dimension measures the employees perceptions of information sharing throughout the Trust, particularly top-down/vertical communication between

management and workers. The scale comprised of five items with five intervals and three stems ranging from strongly agree to strongly disagree. An example item from this scale is Communication between management and staff is excellent in the Trust. The coefficient alpha for the current sample was 0.85. Resources This dimension measures employees' perceptions of resource allocation and usage within the Trust. The scale consisted of seven items with five intervals and three stems ranging from strongly agree to strongly disagree. An example item from this scale is There is very little waste of financial resources in the Trust. The coefficient alpha for the current sample was 0.77. Scale Structure and Reliability Factor analyses and other multivariate techniques demonstrated the empirical distinctiveness of the scales from each other (Hill, 1998). The scale reliabilities reported for this study compared well with the original work, which quotes a range of reliability coefficients from 0.69 to 0.89.

The notion of teamness was operationalised by using the definitions of teams employed in the literature (e.g. Alderfer, 1977; Hackman, 1987; Guzzo & Shea, 1992; Guzzo, 1996, p. 8; West, 1996b). The following characteristics are commonly used to define a team:

a) The group is perceived as a social entity by others and has an organisational identity within a defined function.

b) This is a real group with a task to perform in an organisation from which shared objectives are developed for the team. c) There is a degree of interdependence between members of the group and members interact together to achieve group objectives. d) There is a degree of differentiation of roles and duties in the group. e) There is collective responsibility for measurable outputs. f) Groups are not so large that they constitute an organisation, which has vertical and horizontal relationships and sub-groups. In practice this is usually a group of less than 20 members (although there may be some exceptions to this number). Distilling these characteristics suggests five components of teamness: Distinct roles for members of the team, Task interdependence - team members rely on each other to perform the task, Outcome interdependence - team members' achievement of team goals is dependent on other members' knowledge, skill and task performance, Team identity - team members and other organisational members regard the group as a team with a clear team level task to perform Clear team objectives - there are clear team level objectives.

Section 3: Stress The main measure of stress was the 12-item version of the General Health Questionnaire (GHQ-12; Goldberg, 1972; Goldberg & Williams, 1991). The GHQ-12 was designed as a self-administered screening test for detecting minor psychiatric disorder in the general population. It covers feelings of strain, depression, inability to cope, anxiety based on insomnia, lack of confidence and other psychological problems.

Section 4: Team working Respondents were asked to indicate, by ticking a yes or a no response option, whether they worked in a team. To differentiate between those who did and did not work in a clearly defined team according to our criteria of teamness, but who indicated in answer to the categorical question that they did work in a team, we summed responses to 4 questions:

Does your team have relatively clear objectives? Do you frequently work with other team members in order to achieve these team objectives? Are there different roles for team members within this team? Is your team recognised by others in the hospital as a clearly defined work team to perform a specific function?

Those who did not answer, yes to all four questions were categorised as being in a quasi team. Out of the total sample, 283 responded clearly that they did not work in a team. Of the 1,980 who answered yes to the question Do you work as part of a clearly defined team? 692 answered no to one or more of these questions and were therefore categorised as members of quasi teams. Thus 283 (12.5%) did not work in a team, 1,288 (56.9%) worked in a team, and 692 (30.6%) worked in a quasi team. The Sample in Study 2 The Relationship between Team Processes and Team Member Stress Using data from official records and the expertise of members of the National Health Service Executive, ten Trusts were selected for inclusion in this part of the study

Identifying a sample of teams in each Trust was a lengthy process. Discussions were held with senior managers, who identified teams in their organisations and suggested contact persons from each team. The researchers then telephoned the contact person, negotiated their collaboration in the study, and once agreement was reached, secured the names and location of team members. Contact persons were asked to distribute questionnaires to their team members.

There was considerable variation in the types of team in Trusts, and it was not possible to identify a sufficiently large sample of a single type of team that was common across all Trusts. Six team types predominated: nursing care management medical multidisciplinary support quality improvement teams

Members of 225 teams were invited to take part in the study. Members of 14 teams declined this invitation. Over a period of 16 months, 193 teams in 10 NHS Trusts continued to collaborate in the research. Questionnaire responses were received from 1,237 team members. The numbers in each profession/occupation were: 752 nurses, 114 doctors, 98 administrative staff, 78 managers, 125 professions allied to medicine (PAMs), 26 professional and technical staff, 26 ancillary staff and 18 of unknown occupational group. Team sizes ranged from 2 to 44 (mean 11.4, SD = 6.93). Figure 8.1: Characteristics of NHS Trusts Type Budget* (in million) Teaching Teaching Teaching Teaching District District District Community Community Community 100 125 120 90 78 56 38 40 57 45 Number of Staff* 3,000 5,000 5,000 5,500 3.250 2,500 1,200 2,000 1,200 2,500 Year of Trust Status 1991 1990 1994 1992 1994 1994 1991 1992 1992 1993 City City City City City Rural Rural City Rural Rural Location

* Data available from 1996

The ten NHS Trusts included four teaching hospitals, three community Trusts and three full District Trusts. Numbers of staff ranged from 1,200 to 5,500 (as shown in Figure 8.1).

Women formed 86 % of the sample. Mean age of team members was 39.58 years (SD = 10.52, range 17 to 64 years). Mean team tenure was 4.3 years (SD = 4.65, range one month to 38 years). 5% of the sample had worked in their team for less than one year, 17% between 1 to 5 years, 25% between 6 and 11 years, 18% between 12 and 16 years, and 36% had over 16 years service.

The mean caseness of teams was 23.3% (SD = 0.25), with a mean GHQ Likert score of 0.95 (SD = 0.24). This level is comparable to a group of 71 primary health care teams (caseness = 21.8%, Borrill & West, 1998).

At individual level this can be contrasted to 26.7% for a larger group of British health care employees (n = 22,298, SD = 3.09, Mullarkey et al., 1999) and 18.4% for the general working population (BHPS, Taylor, Brice, Buck, et al., 1995).

NHS employees often belonged to three or more teams (48%), with only 14% belonging to one team. Teams either met infrequently (30% had not met in the last month, and 39% ha met once), or frequently (21% of team had met four or more time in the last month). Most people worked in permanent teams (90% of members). Questionnaire completed by individual team members Section 1: Biographical Information Participants were asked for their job title, age, gender, and duration of tenure in the NHS. Section 2: Team Composition Team members were asked to indicate the size of their teams (number of members), how frequently they interacted together, whether they were members simultaneously of other teams, team tenure, and the nature of the teams task. Teams were classified as coming from Teaching, District General Hospitals or Community Trusts.

Section 3: Team Processes This contained eight measures of team working. Four of these were drawn from the Team Climate Inventory (Anderson & West, 1994,1998) that is based on a welldeveloped theoretical model of team functioning (West, 1990). The four measures assess levels of: team participation clarity of and commitment to team objectives emphasis on quality support for innovation.

Four other measures were included:

Reflexivity the extent to which team members reflect upon their team objectives, strategies and processes and make changes accordingly (West, 1996; West, 2000)

Teamness The extent to which the team functions as a team versus a loose
grouping

Roles - Team members' understanding of the distinctiveness of their own role and
the degree of differentiation of roles within the team. Team members are asked to consider their understanding of their job in the team and the appropriate use of skills and knowledge needed to carry out the work. Then they are asked to consider these themes in relation to other members roles.

Interdependence - Task interdependence is when group members interact and depend on one another in order to accomplish work.

Section 4: Outcomes This section of the questionnaire elicited members perceptions of team performance and shared understanding of team goals.

Section 5: Objectives A single item invited team members to describe their team objectives. Study 3: Do team inputs and processes predict team member retention? Research Design This was a longitudinal research design with data collected six months after participating teams had completed Study 2. 76 teams were selected from four NHS Trusts (two community, one teaching and one District General Hospital).

Following Study 2 teams were sent feedback reports. This created opportunity for dialogue with the teams. A single sheet of questions and an introductory letter was sent to each team contact. Materials were clearly marked with the team name as some contacts were members of more than one team. A stamped, addressed envelope was included to return responses.

Response Fifty-seven teams participated in the survey (31 nursing care, 13 management, 5 multi-disciplinary, 5 Quality Improvement, 2 medical and 1 administrative support team). Community Trusts gave the most enthusiastic response (17 of 19 teams, 89%) and teaching Trusts the least (15 of 26 teams, 58%) (overall = 57%). Data from the Quality Improvement teams were removed from analyses as these teams are not permanent; and several teams had ceased to exist having completed their tasks. The final sample for analysis comprised 52 teams (mean size 11.8, SD 6.03, range 2 to 25).

Nineteen teams did not respond to this survey (mean size 9.0, SD 4.9, range 2 to 19). No information is available as to the turnover in these teams. A comparison of means was undertaken across study variables obtained at Time 1 to identify any differences between responders and non-responders. emerged. No significant differences

Team size and type of Trust are associated with turnover. People are more likely to leave a team if they work in a teaching Trust (mean rank = 36.0 Kruskal-Wallis one way ANOVA; Chi-square 8.43, df 2, p < .05) and are members of a larger team (Pearson correlation: r = .23 between log percentage leaver and team size p < .05, one-tailed test).

Regression analyses showed team size and Trust type account for between 15% to 18% of the variance in turnover.

There is no evidence to suggest that stress at Time 1 is associated with team viability at Time 2. Perceptions of clear team objectives and high levels of participation are significantly associated with low levels of turnover

Regression analysis (controlling for team size and Trust type) reveals that TCI variables treated as a block (participation, support for innovation, team objectives and task orientation) explain 10% of the variance in team retention.

Team Tasks Secondary health care teams diagnose illnesses, plan and administer treatment for various conditions, conduct health screening, and provide maternity care. These are complex tasks that require co-ordination and management, both, in a professional sense to ensure the best outcomes for patient care, and, in an organisational sense to ensure that the work conforms to organisational objectives, budgets, and internal and external standards. There were seven main categories of team included in the sample and these are shown in Figure 8.2 below:

Figure 8.2: Types of Secondary Health Care Team

Classification Medical Teams Nursing Care Teams

Description These are teams of doctors. This is a broader notion of team covering all nursing care to patients / clients. These teams included groups of staff in addition to nurses such as health care assistants, auxiliaries, ancillary staff, clerical staff, and professional staff. These were teams which undertook the task of managing a department, group of wards, or specialty/directorate. Therefore, a senior nursing team that manages a number of wards would be described as a management team. These were teams, often of professional staff, which had the task of delivering care or a service to patients/clients often in a boundary spanning role across departments, wards and specialities. For example, an Endoscopy team would contain medical, nursing, and professional staff who may work in a medical, surgical, or investigative context. These were teams, often of administrative and clerical staff, which provided support to the four team types above. This support may be secretarial, administrative or record keeping in nature. For example, a medical records team would be responsible for the storage, retrieval, and distribution of patients hospital records. These were teams of ancillary staff such as porters, domestics, and catering staff who provided support services for both patients and staff. For example, a portering team would provide support to the whole Trust to transfer patients and goods between locations, provide access to restricted areas and other duties such as security and staff protection. These were temporary teams assigned discrete tasks in order to improve quality of services provided in various

Management Teams

Multi-Disciplinary Team

Support Teams: Administrative

Support Teams: Ancillary

Quality Improvement Team

health care areas. For example, a bed hire team would be monitoring in-patient admission activities and ensuring appropriate mechanisms were created to facilitate the provision of hospital beds in the appropriate locations depending upon demand.

The frequency of each type of team in the sample is shown in figure 8.3.

Figure 8.3: Frequency of Team Types in Study 2

Key to Task Type
100

NUR = Nursing care 45%

86

MAN = Management teams 15%

80

MED = Medical teams 5%

f r e q u e n c y

MD = Multidisciplinary teams 25%

60 51

SUPP = Support teams 6% QI = Quality improvement teams 3%

40 29

20 10 11 6 0 NUR MAN MED MD SUPP QI

Type of team

In order to portray the variety of work that takes place within secondary health care we offer a few examples from each task category and describe the team composition, accountability, and work undertaken.

Nursing Care Teams The Paediatric Nursing Team This team is part of an urban community Trust. Team members provide for the nursing care needs of acutely or chronically sick children at home, or, in other community settings. There are 12 members of this team: a nurse manager, 2 paediatric community nurse sisters, 4 paediatric community nurses, 2 staff nurses, 2 district nurses, and a paediatric diabetes nurse specialist. The nurse manager is the team leader. Organisationally this team is part of the Child Health Care Group and is accountable to the Child Health Management Team. This team can be described as a complex decision making team which performs multiple tasks using both basic and specialised equipment. This team interfaces with many other specialised teams in hospital and community organisations.

The Coronary Care Unit This team is part of a busy urban teaching Trust. This team provides care for patients with heart disease and problems associated with acute and chronic capacity. There are 23 team members: 2 nursing sisters, 15 staff nurses, 2 enrolled nurses, one domestic, and 3 doctors (a consultant, a senior house officer, and a house officer). This is a complex decision making team that provides specialised care delivered using highly specialised and technical equipment. Team members are involved with patients and their families and take an essential role in rehabilitation. In addition, this unit undertakes training and education of staff. The team leader is the medical consultant who is responsible, at Trust Board level, to the Clinical Director of Medicine. Ward One Ward One7 is a busy surgical ward that is part of an urban teaching hospital Trust. The ward practices team nursing and divides patient care between four teams. Team members provide care for patients in conjunction with other professional staff such as occupational therapists and physiotherapists during patients post-operative rehabilitation. Ward One has four members: two health care assistants, one primary nurse (nursing sister), and one associate nurse (staff nurse). The primary nurse is the team leader and will co-ordinate with the other three team leaders on the ward. This team is responsible to the ward manager who is part of the Medical Directorate Management Team. Management Teams The Child Health Management Team This team is part of an urban community Trust. The team co-ordinates the Childrens Service in the community. There are 10 members of this team: a general manager, an assistant general manager, a primary care manager, a district dental officer, manager of speech and language therapy, finance manager, care group planner, consultant paediatrician, personnel manager, and a nurse manager. The general manager is the team leader and is accountable for this team at Trust Board level. This is a complex decision making team which can be considered the top management team for the Child Health Care Group.

Team names are fictitious in order to maintain confidentiality.

The Clinical Management Team This team is part of a busy district general hospital. The function of the team is to manage the General Medicine Directorate and they are responsible to the Trust Board. There are seven team members: one clinical head of service and a deputy clinical head (both doctors), a speciality manager, two nurse managers, a bed manager, and a secretary. The clinical head of service is the team leader. This team can be described as a top management team.

Multi-disciplinary Teams The Surgical Oncology Team This team provides treatment for patients with breast cancer in a busy urban teaching hospital Trust. There are 14 team members: two consultant surgeons, two ward sisters, two medical secretaries, two senior house officers, four surgical house officers, and two breast care nurses. This is a complex decision making team which provides diagnostic services, treatment and follow-up care for breast cancer patients using in-patient and out-patient resources. The team leader is the senior consultant and he is responsible, at Trust Board level, to the clinical director of surgery. The Medical Practice Team The team is part of an urban community Trust providing a comprehensive health care service to a practice population that involves working across the boundaries of local and community care. There are nine members of this team: two district nursing sisters, a community staff nurse, a district staff nurse, a community nursing auxiliary, and four health visitors. A district nursing sister is the team leader responsible to the Locality Management Team. This is a complex decision making team that works across the boundaries of primary and secondary health care.

Medical Teams The Transplant Team This team is part of a busy teaching hospital Trust. The team provides bone marrow transplants for adults and children and carers for patients before, during and after their transplant. This specialist team is involved in the development of new

techniques, training, and research at an international level within the medical community. There are five members of this team who are all doctors: one acts as programme co-ordinator, and two other consultants take responsibility for adult and paediatric patient care. The team leader is the specialty director. This is a complex decision making team which operates across the boundaries of several medical

specialities. The clinicians are involved in highly complex work that requires the use and understanding of complex medical and surgical techniques, drug therapies, and highly technical equipment. The General Medical Team This team has a broad remit providing treatment for any medical or social problem and is part of a large district Trust. This team has four members: a consultant physician, a registrar, a senior house officer, and a pre-registration house officer. The consultant physician is the team leader and there is a strict medical hierarchy within this team. This is a complex decision making team which provides care for patients within the hospital and the community for a wide variety of acute and chronic conditions. Administrative Support Teams The Clinic Notes Team This small team provides medical records for patients with outpatient clinic appointments. This team is part of a busy rural whole district Trust. The team is composed of three administrative staff. This team will undertake tasks that require co-ordination across many departments within the Trust and will undertake many problem-solving tasks. Information technology skills are well developed within this team. This is a non-hierarchical team, which is responsible to the medical records manager. Ancillary Support Teams St. Jane's Domestics St. Janes is a small community hospital in a busy urban Trust. The hospital is due to close in the next two years and services will be moved elsewhere. St. Janes

domestics are a team of 14 ancillary staff who provides domestic services throughout the hospital over a twenty-four hour period. The team is lead by a supervisor who is responsible to the domestic services manager. Each domestic assistant will have a designated geographic area of work but will be required to work in other areas as the need arises. The team performs domestic duties and assists ward staff in providing food and beverages, keeping the ward clean, feeding patients, and helping visitors. It is likely that the team members will feel more part of the ward team than the domestic team. Although this type of work would appear to be of low complexity, domestics need to be able to carry out their work in harmony with the health care environment of the ward. This requires understanding of the health care process in

order to communicate appropriately with patients and visitors and to adequately clean highly complex equipment. These tasks are of medium complexity. Quality Improvement Teams The Outpatients Quality Improvement Team This is a temporary team, which is part of a busy whole district Trust. The aim of this team is to improve the quality of patient care within the Outpatients Department, which covers a broad range of care across medical and surgical specialities. There are six team members: two outpatients' service managers, a quality assurance coordinator, a senior midwife, an outpatient senior sister, and a medical records manager. This is a complex decision making team which is involved in generating many problem solving strategies in order to achieve standards set by the Trust and the external Patients Charter standards. This is not the principal team for most of its members and has many similarities to a quality circle.

In summary, a majority of these teams are complex decision making groups, which undertake multiple health care tasks. With the exception of the quality improvement teams, all are permanent teams with an on-going work remit. The composition of a majority of these teams is made up of a variety of different occupational groups. In addition, there are differences of status, pay, conditions of service and hours of work across these teams. Conclusion The diversity of team types, tasks, composition and organisational contexts in secondary care argues against the use of research designs we employed in primary care and community mental health. There are no unitary measures of effectiveness common across these diverse types of teams. Moreover, the nature of their tasks varies across organisational settings as well as across team types. Consequently, we focused on three questions: Does membership of teams buffer NHS employees in secondary care from the negative effects of stress at work and, if so, why? To what extent and in what ways are team inputs and processes related to team member mental health in secondary care? Do team processes predict team member retention in secondary care? We provide the answers to these questions in Chapter 9.

Chapter 9
Secondary Care Teams Results from Surveys
Summary of Findings

Those working in clearly defined teams in secondary care have lower levels of stress than those not working in teams or working in loose groupings (quasi teams).

Differences between team membership types in stress could be accounted for by the higher levels of social support and role clarity experienced by those working in clearly defined teams.

Those working in teams also perceive greater co-operation amongst all staff and clearer feedback from the organisation on staff performance than those not working in clearly defined teams.

This finding suggests that team membership somehow buffers individuals from the vagaries of organisational climate. Poor training; resistance to innovation, low levels of resources, co-operation, feedback on performance, autonomy, communication and training, appear to affect stress levels deleteriously much less among those working in clearly defined teams, than among those not working in teams or working in looser groupings.

It implies that teams can somehow compensate for the limitations and frustrations of organisational factors in the work experience of their members, and that this can significantly influence the level of stress experienced by organisational members.

Team processes are significantly associated with stress better team functioning is associated with lower team member stress.

The more frequently team members interact and meet, the better does the team function.

The longer team members work together, the clearer their understanding about each others roles.

The more teams people were members of, the less clear they were about the teams objectives. However, they reported higher levels of emphasis on quality of care and understanding of others roles.

Those working in larger teams reported lower levels of participation in team decision making and less clear understanding of team objectives. Around 10% of teams in the sample ceased to exist in the six-month study period and there is an average 6% turnover of team members in the remaining teams.

Clear team objectives and high levels of team participation positively predict member retention.

Data from our recently completed study at the Aston Centre for Health Service Organisation Research8 show that the percentage of people working in teams in acute trusts is associated with lower levels of patient mortality. The more people who work in teams in Trusts, the lower the number of patient deaths measured by the Sunday Times (Dr Foster) Mortality Index, deaths within 30 days of emergency surgery and deaths after admission for hip fracture. Is team membership associated with lower stress?

The data from this study revealed that 283 (12.5%) respondents did not work in a team, 692 (30.6%) were members of quasi teams and 1,288 (56.9%) worked in a clearly defined team (i.e. they conformed to the criteria of teams specified in the research design see chapter 8).

Analysis of variance within and between the three groups (team, non-team, quasiteam) in relation to stress scores on the GHQ-12 revealed significant differences between those who worked in teams (mean = .95), those who did not work in teams (mean = 1.09) and those who worked in quasi teams (mean = 1.03) (f = 15.68; df = 2,2250; p = > 0.001). Using the GHQ caseness method of scoring, 98 of those who definitely did not work in a team were categorised as cases (equivalent to 34.9%). 275 of those who worked in a team were categorised as cases (21.8%) and 203 of those who worked in a quasi team were categorised as cases (29.7%). Caseness implies the individual is suffering from a sufficiently high level of stress that they require and would benefit from some professional help.

The next step in the analysis addressed the question of what could explain these differences in GHQ scores between the teamworking types. We examined, in turn, demographic, work role and organisational climate factors. Demographic factors To determine the extent to which demographic factors accounted for GHQ differences between those not working in teams, those working in teams and those working in quasi teams, we conducted Chi-Squared tests of these groups by demographic factors.

These included occupational group, gender, whether they had children, marital status (single/married/living with a partner/separated/widowed/divorced), the

organisation, (i.e., membership of which of the four Trusts that participated in the study), time employed in the National Health Service, time employed in current post, and age. There was a significant Chi-Square value only in the case of occupational group (Chi-Square of 47.73; df =12; p = <0.001). We therefore conducted an

analysis of variance by team type (non-team, team, quasi team) and occupational group by GHQ scores. There were significant main effects for occupational groups

(f = 5.572; df = 6, 2232; p = <0.001) but this did not remove the effect of team type on GHQ score (f = 18,645; df = 2, 2232; p = <0.001). There were no significant

interactions between occupational group and team type group on the GHQ. Work role factors

Details are available from West or Borrill (co-authors of this report)

Next we conducted analyses of variance to determine whether work role factors varied between team membership types (those not working in teams, those working in teams and those working in quasi teams). There were significant differences

between these types in role clarity, supervisory leadership, social support, feedback, autonomy and control, and influence over decision-making, with those working in teams reporting higher levels that those in quasi teams, who in turn report higher levels than those not working in teams. A directly opposite pattern was found in relation to role conflict and role ambiguity. There were no differences between the groups in perceived work demands and hours worked.

Which of those clear differences in work role factors between the groups might therefore account for the variation between team membership types in GHQ scores? To answer this question we conducted separate analyses of covariance to examine the variation between team groups in GHQ scores controlling for each of the work role factors in turn. Figure 9.1 shows the results, which reveal that none of these

work role factors alone accounts for the difference between the team membership types in GHQ scores. However, the effect is most reduced by using role clarity and social support as covariates. Indeed, when these two variables are entered as

covariates together, the difference between team membership types in GHQ scores is no longer significant (F= 0.955; df = 2,221; P = 0.385). Thus it appears to be the differences in social support and role clarity between those who work in teams (high social support and role clarity) and those who do not work in teams or work only in quasi teams, which account for variations in stress levels between these team membership types. Figure 9.1: Analysis of variance of GHQ scores by team membership type (team, quasi team, non-team) controlling for work role factors Covariate F DF Significance of F <.0001 <0.001 <0.001 <0.001 Main Effect (F) 4.51 9.12 4.99 7.15 DF Significance of F 0.011 <0.001 .007 .001

Role Clarity Role Conflict Feedback Influence over Decisions Supervisory Leadership

213.03 284.36 273.24 60.90

1,2231 1,2233 1,2217 1,2232

2,2231 2,2233 2,2217 2.2232

124.55

1,2179

<0.001

5.37

2,2239

.03

Role Conflict Autonomy and Control

148.16 5.30

1,2147 1,2216

<0.001 .02

12.38 14.67

2,2147 2,2216

<0.001 <0.001

We then examined the data to determine if there were any significant interactions between work role factors, which were significantly associated with GHQ scores, and team membership types. There was a significant interaction only between team type and role conflict level in predicting GHQ scores (R2 change = .0025; F = 3.13; p = .044). This showed that at low levels of role conflict there were no differences

between the team membership types in GHQ scores, but at high levels of role conflict there were very large differences, principally between those who were members of teams and the other two team membership types) those not working in teams or those working in quasi teams). Perhaps health care teams enable their members to manage role conflict through an overt process of shared role negotiation, whereas those who do not work in teams do not have easy access to such a process. Organisational climate We also conducted analysis of variance to determine whether organisational climate factors varied between team membership types. There were highly significant

differences between the team types in perceived autonomy, training provision, level of organisational resources, organisational attitude to change, organisational feedback on performance, communication and staff co-operation, with more positive perceptions among those working in teams and the most negative perceptions amongst those not working in teams. To determine which of these organisational climate factors might account for the variation between team membership types in GHQ scores, we conducted separate analyses of variance entering each climate variable as a covariate. The results revealed that none of these variables could entirely account for the difference between team membership types in GHQ scores. However, when we entered co-operation and feedback on performance together, the difference in GHQ scores between team membership types was removed (F = 2.01; df = 2, 2208; p = 0.134). Thus, the difference in perceptions of co-operation between staff working in the organisation, and the perceptions of the quality of the organisation's performance feedback to staff, between those working in teams, those not working in teams, and those only working in quasi teams, appears to account for the differences between these team membership types in stress levels.

Finally, we checked the data to determine whether there were any significant interactions between perceptions of organisational climate factors and team membership types that predicted GHQ scores. There were significant interactions between team membership type, innovation, feedback on performance, co-operation between staff, communication, autonomy, and training and organisational resources when predicting GHQ levels. They show that those working in clearly defined teams seem less strained than those not working in teams, or working in quasi teams, by perceptions of low levels of these organisational climate factors. It is as though, by working in a team, team members achieve a shared level of self-sufficiency that buffers team members from the inadequacies of their organisations. Those who are not members of clearly defined teams seem more affected both positively and negatively respectively, by the relative presence or absence of those organisational factors. Discussion The results suggest that being part of a team in the high-strain setting of the NHS is associated with lower levels of stress than if one is not a member of a team or belongs to only a loosely defined and weakly interdependent team (what we have called a quasi team). The results could not be accounted for by demographic

factors, or by individual work role and organisational climate factors. However, the results clearly suggested that differences between team membership types in stress could be accounted for by the higher levels of social support and role clarity experienced by those working in clearly defined teams. This finding is consistent with theoretical explanations of some of the beneficial effects of teamworking that propose that teams contribute to a greater and shared sense of role clarity and social support (Cohen & Bailey, 1997; West, Borrill & Unsworth, 1998; Mohrman, Cohen & Mohrman, 1995). Roles are socially negotiated sets of mutual expectations and, by working closely with those in ones role set, role clarity results. Moreover, since teamworking, by definition, involves interdependent working with close social contact and communication, it is likely that team members will experience more support from colleagues than those whose working relationships are less tightly linked.

Similarly, in relation to organisational climate perceptions, those working in teams derive a sense of greater co-operation amongst all staff and clearer feedback from the organisation on staff performance, as a consequence of their team membership than those not working in clearly defined teams, and this accounts for the differences

between team membership types in stress levels. It is easy to appreciate how membership of a team (whose members co-operate to achieve shared goals) might lead to the illusory sense of high levels of co-operation among staff more generally in the organisation. Another possibility is that staff do co-operate more with those who are members of teams, perhaps because of their clearer roles and goals, or because of the greater power conferred by their membership of a group, or as a result of their more clearly defined social and functional identity - "This person is a member of the Accident and Emergency Resuscitation Team and I know about their functional significance and understand what information or resources they require"

Similar explanations can be offered for the effect of the relatively high level of organisational feedback perceived by team members which accounts (in concert with perceptions of staff co-operation) for the differences between team membership types in stress levels These may be illusory perceptions with team members mistakenly assuming that the higher level of feedback on their performance that they experience (as a consequence of their team members' feedback to them), can be attributed also to organisational feedback to staff on performance. It could also be that as a result of the clear functional identity of the team in the organisation, the team does get clearer feedback on performance.

Particularly intriguing is our finding of significant interactions between team membership types and organisational climate perceptions as predictors of stress. This finding suggests that team membership somehow buffers individuals from the vagaries of organisational climate. Poor training, resistance to innovation, low levels of resources, co-operation, feedback on performance, autonomy, communication and training, appear to affect stress levels deleteriously much less among those working in clearly defined teams, than among those not working in teams or working in poorly defined teams. The consistency of these interactions and the fact that they are largely absent when we examine interactions between work role factors and team membership types, suggest their robustness. It implies that teams can somehow compensate for the limitations and frustrations of organisational factors in the work experience of their members, and that this can significantly influence the level of stress experienced by organisational members

We can speculate about a number of possible explanations for our findings First, it may be that those who have lower levels of stress, self select into teams People who are relatively well-adjusted socially and have lower levels of anxiety may be attracted

to work closely with others since the challenges of teamworking are more manageable for them than they are for those who experience high levels of anxiety and uncertainty. Alternatively, it may be that those who are more relaxed and cope with stress better are selected into teams by existing team members as they offer less of a threat to the effective social functioning of the team. Finally, of course, it may be that those who have relatively low levels of stress are less likely to leave teams, than those with high levels of stress.

Thus those who experience high levels of stress may find teamworking too demanding and challenging and their stress may also create social dysfunction that leads to their attrition from the team. These attraction-selection-attrition explanations for our findings are credible alternatives to the suggestion that it is the effect of working in teams upon stress that we have discovered in this first study. Study 2 Do the composition of and the way secondary health care team members work together affect member stress?

Team processes are significantly associated with stress better team functioning is associated with lower team member stress.

The more frequently team members interact and meet the better does the team function.

The longer team members work together the clearer their understanding about each others roles.

The more teams people belonged to, the less clear they were about the teams objectives. However, they reported higher levels of emphasis on quality of care and understanding of others roles.

Those working in larger teams reported lower levels of participation in team decision making and less clear understanding of team objectives.

Questionnaires were distributed to team members as described in the previous chapter. 193 teams from 10 NHS Trusts responded. The responses consisted of 1,

237 team members (752 nurses, 114 doctors, 98 administrative staff, 78 managers, 125 professions allied to medicine (PAMs), 26 professional and technical staff, 26 ancillary staff and 18 of unknown occupational group). Team sizes ranged from 2 to 44 (mean 11.4, SD = 6.93).

The overall response rate to this survey was 54%. There were variations in response across organisational type (from teaching 42% to community 72%) and across team task type (Support team 30% to Management team 65%). Women formed 86 % of the sample. Mean age of team members was 39.58 years (SD = 10.52, range 17 to 64 years). Mean team tenure was 4.3 years (SD = 4.65, range one month to 38 years). 5% of the sample had worked in their team for less than one year, 17% between 1 to 5 years, 25% between 6 and 11 years, 18% between 12 and 16 years, and 36% had over 16 years service in their teams.

Nearly a quarter of those working in teams scored above the cut-off point on the GHQ, indicating a high levels of stress. The mean caseness of teams was 23.3% (SD = 0.25), with a mean GHQ Likert score of 0.95 (SD = 0.24). This level is comparable to the primary health care and community mental health teams (caseness = 21.8%) but somewhat lower than the 26.8% recorded amongst a larger group of NHS employees (n = 22,298, SD = 3.09, Borrill et al., 1998) though higher than the figure of 18.4% for the general working population (BHPS, Taylor, et al., 1995).

Team Interaction Teams either met infrequently (30% had not met in the previous month, and 39% had met once), or frequently (21% of teams had met four or more times in the previous month). Team interaction frequency was significantly related to all team process variables (with eight of the ten possible relationships being significant). All relationships were in a positive direction suggesting that higher frequency of team interaction facilitates team processes.

Team Task A simple classification of six team task types was used (see previous chapter). Quality improvement teams rated support for innovation higher than did medical and ancillary support teams. Managerial teams described their team objectives in greater detail than support and multidisciplinary teams. Administrative support teams report

greater understanding of each others roles compared to teams undertaking medical and managerial tasks.

Multiple Team Membership Most people worked in permanent teams (90% of members). Nearly half of this sample belonged to three or more teams (48%). Multiple team membership was negatively associated with clarity of and commitment to team objectives and positively associated with emphasis on quality of care, interdependence, and role understanding. This suggests a lack of clarity about team objectives may be counterbalanced by a more vigorous emphasis on quality of care: working in other teams may enable members to be more vigilant about and aware of quality issues.

Team Size Team size is positively related to interdependence and negatively related to participation and clarity of and commitment to team objectives. Many of the larger teams are ward or department teams, which administer care or undertake investigations over a 24-hour period. Dependency on other team members is critical for the safe and efficient delivery of these services. However, membership of larger teams is also associated with low levels of information sharing and influence over decision making, and less clear understanding of the teamsobjectives. These findings are consistent with previous studies (Blau, 1970; Shaw, 1981; Stahelski & Tsukuda, 1990; Sundstrom et. al., 1990) that demonstrate that as group size increases there are increased difficulties in communication, co-ordination and interpersonal relationships. Team Tenure Team tenure was not related to stress but was related to mutual role understanding. Relationships between Team Processes and Stress Partial Pearson product moment correlations were undertaken with pair-wise deletion, controlling for team size. Eight of the ten-team processes variables examined were significantly and negatively associated with stress. This suggests that good team processes are associated with lower stress. Five process variables predicted stress: participation, support for innovation, emphasis on quality, team objectives, and role understanding.

Team Processes and Stress Regression analysis at team level revealed team processes accounted for 22.8% of between team variance in stress. These results are summarised below. After controlling for team size, three process variables negatively predicted stress (task reflexivity, p < .01; emphasis on quality; p < .05; interdependence, p < .01). Figure 9.2: Hierarchical Regression Analysis for Team Processes Predicting Stress (n=193) Variable Step 1 Team Size B SE B .002 .5 df 1 .069 .060 .045 .051 .05 .046 .043 .02 .059 .054 df 10

.0043 R2 .016

.13 p <.08 .20 -.014 .14 -.31* -.32** -.039 .24** .030 -.19 .0068 p <.001

Step 2 Participation Support for Innovation Team Objectives Emphasis on quality Reviewing Processes Social Relations Interdependence Team Objectives Role Outcome

.098 -.0074 .065 -.12 -.14 -.016 .11 .008 -.11 .0033 R2 .228

Note * p < . 05, ** p < . 01, *** p < . 001. Relationships between Stress and Team Processes at the Individual Level of Analysis

Process variables were entered in blocks to avoid suppresser effects and finally entered together. Several process variables significantly predict stress at the individual level: participation, emphasis on quality (p < .01), social reflexivity (p < .001), role understanding, and knowledge of team outcomes (p < .05). Stress and Team Processes Across Two Levels of Analysis

The analysis of this study is complex as a cross-level model (Rousseau, 1985) is being explored: individual level of stress and team level processes. These levels are nested (hierarchically ordered systems) and exist within the context of a

higher level unit (organisation, the NHS Trust). Strategies that examine relationships at one level of analysis ignore the influence events at another level may have on the data. In this case events may occur in the team that might influence individual stress as well as individual events influencing team morale (such as team members leaving).

Hierarchical Linear Modelling (HLM; Bryk & Raudenbush, 1992) is a strategy that investigates data at more than one level of analysis. This analysis takes into account both individual and team level variance. To explore the relationships in this study the following research questions were examined. How much does stress vary across teams? Do team process variables influence stress?

Data from 136 teams (n = 1,121) were analysed using HLM (teams with 4 or less members were removed from analysis to improve reliability). Variables were chosen that significantly correlated with stress: participation, support for innovation, team objectives, emphasis on quality, reflexivity, social relations, role understanding, knowledge of team outcomes and teamness (a composite of the four criteria variables to identify team membership). Team tenure, age and gender were used as control variables.

HLM demonstrates that 3.4% of the total variance in stress is explained by team factors (this is significant, given that the GHQ measures general life stress rather than work-related stress in particular). More than75% of this variance can be explained by team processes (as measured by the TCI scales clarity of team objectives, emphasis on quality, participation and support for innovation).

Study 3: Do Team Inputs and Processes Predict Team Member Retention?

Around 10% of teams in the sample ceased to exist in the six-month study period and there is an average of 16% turnover of team members in the remaining teams.

Larger teams have lower levels of retention. Clear team objectives and high levels of team participation predict member retention.

Results 57 teams from 4 NHS trusts (75% of those originally participating) participated in the survey (31 nursing care, 13 management, 5 multi-disciplinary, 5 quality improvement teams, 2 medical teams and 1 administrative support team). Data from the quality improvement teams were removed from the data set as these teams were not permanent and several had ceased to exist having completed their tasks. The final sample for analysis comprised 52 teams (mean size 11.8, SD 6.03, range 2 to 25).

Nineteen teams did not respond to the survey (mean size 9.0, SD 4.9, range 2 to 19. No information is available as to the turnover in these teams. A comparison of means was undertaken across study variables obtained at Time 1 to identify any differences between responders and non-responders. No significant differences emerged.

A high rate of turnover was reported amongst the teams surveyed. 10% (5) of the teams ceased to exist and only 12% (6) of teams reported no change in membership (either joiners or leavers) over the six months period of the study. On average, there were two leavers per team (mean = 1.90; SD = 1.94; range 0 to 8), and an average 16% turnover during the study period (mean = 16.1; SD = 15.8; range 0 to 71.4%). 26.3% of teams reported no leavers.

Team size and type of Trust are associated with turnover. People are more likely to leave a team if they work in a teaching Trust (Chi-square 8.43, df 2, p < .05) and are members of a larger team (Pearson correlation: r = .23 between log percentage leaver and team size p < .05, one-tailed test). Regression analyses revealed that team size and Trust type accounted for 15% and 18% of the variance in turnover respectively. There was no evidence to suggest that stress at Time 1 was

associated with team turnover at Time 2. Perceptions of clear team objectives and high levels of participation were significantly associated with low levels of turnover, such that in teams whose members were clear about and committed to the team objectives, and who reported high levels of team participation at Time 1, there were lower levels of turnover between the two measurement points. (See Figure 9.3).

Regression analysis (controlling for team size and Trust type) revealed that team process variables treated as a block (participation, support for innovation, team objectives and emphasis on quality) explained 10% of the variance in team turnover or team member retention. Conclusions Overall these three studies suggest the value to NHS employees in secondary care of working in teams, and particularly in teams that are characterised by clear objectives, high levels of participation, emphasis on quality and support for innovation. Taken together with the findings reported in earlier chapters, it

suggests teamworking is a means for promoting effectiveness in the NHS and the well-being of employees. Perhaps most striking is the finding from a recently completed study by the Aston research team (West, Borrill and colleagues) revealing that the percentage of Trust staff working in teams in acute trusts is associated with lower levels of patient mortality. The more people who work in teams in Trusts, the lower the number of patient deaths measured by the Sunday

Times (Dr Foster) Mortality Index, deaths within 30 days of emergency surgery and deaths after admission for hip fracture.

We now turn to examine the results of the intensive analysis of team meetings and communication amongst primary health care and community mental health teams, carried out by the Human Communications Research Centre at the Universities of Glasgow and Edinburgh in collaboration with the Leeds and Aston research groups.

Chapter 10
Meetings and Communication Research Methods
Introduction

Meetings are important for effective teamworking, providing teams with their main opportunity for communicating. They are particularly important in health care teams where the team members often work in different locations and therefore have little opportunity for communication. For this part of the work, we have compiled two different sources of information about team meetings. The first source details the range of meetings held within the team and who is invited or expected to attend them. This information was derived from the practice manager interviews in PHCTs and the CMHT. This allows us to assess how much communication took place in a team, especially across different disciplines, where links are usually the poorest. However, even within teams which hold many meetings, with good links across disciplines, communication can be good or poor, depending on how those meetings are conducted. Therefore our second source of information is recordings of team meetings, from which we draw both quantitative measures and observations about meeting practice. Communication and decision making in teams

Communication in meetings is important in teams for two different reasons. First, effective teamworking requires everyone to be both well-informed and to be invested in the team's overall goals and plans. Although there are other methods for keeping team members informed, such as newsletters, bulletin boards, and informal conversations, meetings are a common way of doing it. Meetings are the most effective method of involving a group of people in activities which require discussion. Team members who have been involved in the discussions of the team's goals or plans, or in the decision-making process itself, are more likely to feel that they "own" those goals and plans and to work actively to bring them about (Weldon & Weingart, 1993). Second, the plans which a team develops are likely to be better, the wider the pool of views they take into consideration. Each individual in a team will have their own unique perspective on how health care can best be provided in that team's circumstances; in particular, team members from different disciplines encounter very

different situations in their daily working lives and therefore will have different ideas about what should be done. Providing the best care means synthesising this

diversity of views into a co-ordinated plan which is understood and accepted by everyone on the team. Therefore two important properties of team communication are that everyone participates and especially that every discipline is involved in discussion. Barriers to effective communication

One of the classic problems for all teams, and not just ones from the health care sector, is that when there are status differences between team members, higher status members are more likely to attend the most important meetings. Even within a single meeting, higher status members are likely to make the contributions which drive the meeting, such as giving information, asking questions, and making suggestions; lower status members are usually restricted to relatively short, responsive contributions such as answering questions or expressing agreement with something that has been said (Berger, Rosenholtz, & Zelditch Jr., 1980; Berger, Fisek, Norman, & Zelditch Jr, 1977). This tends to make lower status members less invested in the team's plans and to limit the range of ideas about potential changes to increase effectiveness that the team actively discusses. In manufacturing industry, for instance, it has been argued that strictly hierarchical management structures are insufficient to deal with the rate of change in the modern business world because they promote the flow of information downward but not upward, making it difficult to adapt using information gained "on the ground." (Burns & Stalker, 1966)

Another classic problem, again universal, is that the larger a group discussion, the more one person will come to dominate that discussion and the more people will sit silently rather than contributing actively (Bales, Strodtbeck, Mills, & Roseborough, 1951). In fact, the optimal group size for free discussion is five people, and in any group larger than around eight, no more than eight people say virtually anything which is said. In status-differentiated groups, it tends to be the high status individuals who speak and the low status ones who remain quiet (Berger et. al., 1980); where one person has authority for decision-making, they tend to control the interaction (Carletta, Garrod, & Fraser-Krauss, 1998). This can create difficulties for larger

teams unless they can find a way to discuss freely issues in small, cross-disciplinary groups and then pass ideas forward from them. Finally, there are differences in the properties of very small groups which make them more suitable for free discussion.

Even in relatively small groups, active discussion in a meeting tends to involve just a few people at a time, but there is evidence that people who participate actively in a discussion understand and react to it differently from those who simply overhear it (Schober & Clark, 1989). In non-status-differentiated groups of five people,

discussions are highly interactive, with people's opinions influenced by whoever they interacted with the most. Even in groups of just ten people, speakers make longer utterances as if lecturing to the whole group, and whoever speaks the most has the most influence (Fay, Garrod, & Carletta, 2000). This means that for good discussion across disciplines, it is important to keep the groups small and make sure that status is not an issue as far as that is possible, so that everyone has at least some chance to interact with everyone else. Because people are more likely to have informal conversations outside meetings with people they encounter and see as similar to themselves i.e., staff from the same discipline this makes cross-disciplinary discussion in meetings all the more important. Methods Our general interview methods have already been detailed in chapter 3; interviews were held with the practice manager for primary health care teams and the team leader in the community mental health care teams answered a questionnaire. Information was collected about the set of meetings held within the team, who was invited or expected to attend them, how long the meetings were, the purpose of the meetings, and how frequently they were held. The sample sizes for primary health care teams and community mental health teams are 67 and 92, respectively. The rest of this section describes the methods used in order to obtain a corpus of recorded meetings. Team Selection Within the primary health care sector, all teams undertaking the questionnaire and interview section of the Health Care Team Effectiveness study were invited to participate in recording of meetings for further study; selection was first-come firstserved, with no additional selection criteria. In particular, we did not select teams based on meeting size or on the results of the effectiveness questionnaires. Twelve teams volunteered for this part of the study. Meeting size ranged from three people to twenty-five. The teams recorded were reasonably representative of the larger sample of primary health care teams. Meeting Selection Recordings were made of multi-disciplinary decision-making meetings as it was expected that these meetings would best reflect effective teamworking. Multi-

disciplinary meetings were defined as those attended by a range of disciplines.

Decision-making meetings were defined as those where in addition to exchanging information decisions were made during them, again with active participation from the different disciplines. Primary health care teams provided access to the teams business meetings, in which the day-to-day running of the practice was discussed. In one team, all decisions affecting the practice were made solely by the partners, all of whom were doctors, and therefore there were no multi-disciplinary decision-making meetings to record. In this case the partners meeting was recorded. For community mental health teams, the meetings again were multi-disciplinary business meetings in which decisions were made about the running of the team. Wherever possible, two meetings of the same type were recorded for each team. Dates for meeting

recording were at the discretion of the practice; therefore the two meetings recorded were not always in sequence.

Before each meeting was recorded, the primary contact for the team, usually the meeting chairman, was briefed that the meeting was to be kept as naturalistic as possible. It was requested that recorded meetings should be held in their usual locations, with their usual meeting protocols (agendas, minutes, chairing procedures, etc.), and that attendance should be the same as if the meeting were not being recorded. The researcher who tended the recording equipment made herself as unobtrusive as possible. Equipment used Audio recording was conducted with two omni-directional PZM tabletop microphones linked to different channels of a high quality audiotape recorder; the microphones were set up so as to maximise channel differentiation but to be unobtrusive enough that participants would not move them. A single static video camera on a tripod was trained to record the gross movements of as many of the participants as possible; this record was used only to aid speaker identification during transcription. Transcription Before each meeting was opened all participants introduced themselves and their occupation and on the basis of this was allocated a speaker number. Therefore the first person to introduce him/herself became speaker 1, the second speaker 2 and so on. Each participant was referred to by the same speaker number for the transcripts of both meetings regardless of when they spoke during the second meeting.

Meetings were transcribed from the audiotapes by an audio typist who had not attended the meeting. Audio typists transcribed complete contributions in order according to when they began, labelling each contribution by speaker number, but

did not code finer timing information. Speaker identification was facilitated both by the video recording and by a seating plan drawn up during the meeting by the person recording the meeting. In previous work using these methods on four to twelve person meetings, transcribers were able to agree very reliably who made any one contribution; using the kappa statistic, K=.93, k = 2, N = 230, with an average of 2% and a maximum of 6% non-backchannel contributions left as unidentified (Carletta et al., 1998). A contribution was defined as a period of speech from one individual in which the only major pauses coincided with silence from the other speakers, so that the pause was likely to be caused by the speaker thinking and not by the speaker listening to someone else's contribution. Under this definition, speakers cannot follow themselves in the speaking order. Overlapped speech was transcribed, with the extent of the overlap roughly marked. Infrequently, parts of the meetings were

omitted because they were so badly overlapped that we could not track individual contributions. After transcription, the transcripts were completely anonymised taking out all staff, patient, place names, place and local authority names or possible team or person identifiers.

An example transcription excerpt is given in Figure 10.1. Transcription proceeds one contribution per row. Column one contains the speaker number. Column two

contains the words said, combined with some coding information, and column three contains any notes which the transcriber wished to make (for instance, about people entering or leaving the room). Column two codes include /num for the approximate location of the start of another contribution during overlap, italics for anonymised text, @ for words from the tape which could not be heard clearly, and some common descriptions of non-linguistic behaviour such as general laughter within the group. To make the example clear, everything but transcribed speech is indicated in red.

Figure 10.1: An example of the format used for meeting transcription

1 4

Shall I open /4 the meeting Yep, lets get on with it. phone rings

My apologies I am going to have to leave before the end. I have an appointment in Place 1. Are you skiving off? Group laughter

Analytical techniques The primary interest is in how well teams communicate not just overall but also across disciplines. The analysis therefore relies on a classification of meeting

participants by occupation. For ease of reference, categories are identified by colour as well as number. For primary health care teams, we have used the following categories. 1 2 3 4 5 6 GPs practice managers practice nursing staff, including nurse practitioners attached staff (mostly health visitors and district nurses) administrative staff (mostly secretaries and receptionists) Miscellaneous

For community mental health teams, the categories are instead: 1 2 3 4 5 6 7 Psychiatrists nursing staff occupational therapists psychologists, psychotherapists, and other therapists Managers staff from social services Miscellaneous

Both kinds of teams rarely had miscellaneous staff or miscellaneous meeting attenders who were not included in the analysis. In most cases, these were visitors, students, or staff associated with the physical location such as caretakers and security staff; for community health care teams, there were also sometimes representatives of user, carer, or voluntary groups and liaisons to local GPs. For both the entire range of meetings described in interview and for the recorded meetings, descriptions were produced of which staff categories interact with each other. These descriptions are best explained in terms of diagrams. For instance, consider the following diagrammatic representation of one of the recorded primary health care team meetings:

Figure 10.2: Communication in a primary health care team meeting

6 7 8 9 10 11 12 13 14

4 3 2 1 heavy: > 2 1 medi um: > 1 4 19 18 17 l i ght : > 7

70 minutes 68% of team present

15

16

In the diagram, each person is represented by a coloured circle, where the colour represents their occupational category. How often each person spoke immediately before or after each other person is represented by the line between their two circles. Heavy lines mean the people took adjacent turns relatively often; light or no lines means that they took adjacent turns relatively rarely. When people take adjacent turns in meetings of this type, they are usually (but not always) communicating directly with each other and addressing the same topic. The actual line darknesses are determined by the maximum number of times anyone followed anyone in the meeting and using that to construct quartiles; no line is shown when the number of adjacent turns the two people took is less than a fourth of this maximum, a light line when it is less than half, and so on. For instance, in the diagram shown, nineteen people attended the meeting, of which seven were doctors (indicated in red), and the heaviest interaction was between participants 3 and 6, 3 and 8, and 8 and 1. Despite the fact that no pairs were able to interact very many times (as indicated by the numbers in the legend) this was a quite long meeting. Although the numbers in the legend are affected by meeting length, smaller meetings are more likely to have high numbers because there are fewer possible pairs to interact, and therefore the potential for any given pair to interact is greater. High numbers in the legend of a large meeting, unless it is unusually long, indicate that most of the possible pairs of

people do not interact with each other at all, and usually means that many of the people attending the meeting say little or nothing.

In the diagram just observed, quite a few pairs of people interact, showing that, at least among those who participate actively, the interaction is quite free. However, although there are many people present at the meeting and they represent all of the occupational categories, the interaction is almost exclusively among the GPs and practice manager. It is also possible for interaction to occur primarily between one person and others within the meeting (see Figure 10.3).

Figure 10.3: Communication in a primary health care team meeting

6 7 8 9 10 11 12 13 14

4 3 2 1 heavy: > 7 medi um: > 4 19 18 17 l i ght : > 2

30 minutes 40% of team present

15

16

This usually indicates strong chairing of the meeting, although the chair may only be choosing who will speak next rather than actively controlling the topic of the discussion or making contributions to the topic him or herself. Alternatively, it may indicate that the purpose of the meeting was for the dominant person to give a report, with others asking clarification questions as needed. Whether the diagrams indicate interaction among many pairs or interaction through one central person, the interaction which exists can be more or less cross-disciplinary, depending on the mix of people who actively contribute to the meeting.

The same sort of diagram serves for the interview data. Consider the following depiction of the set of meetings in a community mental health team:

Figure 10.4: Communication in a community mental health team

v er y l i gh t : < 2 1 3 6 l i gh t : < 4 2 7 m edi u m : < 6 4 1 4 5 h eav y : < 8 5 5

Here, instead of representing individuals, the coloured circles represent staff categories. Instead of representing the number of times people took adjacent turns, the lines show the number of minutes per month which representatives of the two categories spend in the same meetings. Lines around a circle show all meetings which someone from that staff discipline attended. For instance, this diagram shows a team in which psychiatrists never attended meetings, not even ones only with other psychiatrists. As in the diagrams for the recorded meetings, the darker the line, the more communication occurred. As well as providing a descriptive account of meeting practice in health care teams, we also derive quantitative measures of communication from what we observed. Recall that good communication among the pairs of individuals in a team and among the pairs of disciplines is theoretically important for effective teamworking. To

measure this, we have devised a score for freedom of interaction (Carletta et. al., 1998). For a recorded meeting, the freer the interaction, the more pairs of

participants take adjacent turns. This is reflected in the diagrams by how "starry" they appear. Similarly, starriness in the diagram for a set of meetings reflects how free the interaction is in general among the different staff disciplines. Freedom of interaction is scored based on either the meeting transcripts or the interview data. The scores vary between 0 and 1, with high scores reflecting high freedom of interaction. Similarly, for the individual meetings, since it is important to know

whether high status individuals are over-represented in the communication, we score

equality of participation from 0 to 1. Meetings with equal participation have the same darkness of lines coming from each of the participants if they are added together, but do not necessarily link all the pairs and therefore do not necessarily have very free interaction. Other measures for both individual meetings and the set of meetings for a team as a whole consider the amount of communication which occurs, sometimes divided by individual or discipline, and who attends meetings.

Chapter 11
Analysis of Communication in PHCT teams

Summary of Findings

Meetings in primary health care are often badly managed and dysfunctional Attached staff (i.e. health visitors and district nurses) often miss team meetings, yet support for innovation is higher in teams where they attend. Better meeting attendance in PCHTs is associated with care that is more patient centred. In primary health care, team meetings are often re-arranged, cancelled or start late.

When teams do meet many PHCT team members remain silent throughout the meetings. In over half the meetings we recorded no group decisions are taken.

Types of meetings

"A meeting" is defined as a set of people, who meet usually at some regular interval, for a particular purpose. Adding together the time devoted to the different meetings reported - all the meetings which involved any part of the team - shows that primary health care team members spend relatively little of their time in meetings. On

average, there was a meeting involving some part of the team for 325 minutes per month (range 22 - 1190, S.D. 240); that is team members spent about 3% of their time in meetings. The primary health care teams in the sample had between 1 and 6 meetings. The frequency of these meetings ranged from weekly to yearly. Primary health care meetings tended to fall into the following categories, divided by who attended them:

Figure 11.1: Primary health care meetings

1 2

Whole team meetings, usually held monthly and attended by either the whole team or at least by representatives of each of the disciplines (22% of sample).
5

1 2

5 3

Single discipline meetings for doctors or for administrative staff, sometimes with practice management also attending (28% and 3% of sample, respectively). Doctors meetings were typically weekly or monthly; administrative meetings were usually monthly or every two months.

1 2

5 3

Practice clinical meetings attended only by those clinicians based in the practice, and sometimes by the practice management (43% of sample). These meetings were usually monthly but some practices held them weekly.

1 2

Full practice meetings, usually held monthly and attended by all staff based at the practice: management, doctors, practice nurses, and administrative staff (7% of sample).
5

1 2

Clinical staff meetings, usually held monthly and attended by all clinicians included attached ones, and sometimes by the practice management (16% of sample).
5

1 2

Nursing meetings attended by the practice nurses and attached staff such as health visitors and district nurses (6% of sample).
5

Diagrams shown are representative of the types, but not all meetings in the category conform completely to the diagram. Nursing meetings were the only ones which were never attended by the practice management. For each meeting type, there was no relationship between whether or not a team held a meeting of that type and the

team's size (unrelated t-tests, allowing for unequal sample sizes).

The set of

meetings which a team held divided teams into the following categories, with the following typical diagrams. For each category, we give the mean, minimum, and maximum size of team with that meeting practice.

Unitary: Teams with whole team meetings and nothing else (mean team size 14, min. 8, max. 26).

Multiplex: Teams with both a clinical staff meeting which included attached staff and either a full practice, practice clinical, or single discipline doctors meeting. One-quarter of these teams also held a whole team meeting (mean team size 23, min. 10, max. 51).

Unitary-plus: Teams which hold whole team meetings plus either a separate doctors meeting or a separate practice clinical meeting (mean team size 23, min. 10, max. 45).

Practice-based: Teams whose most inclusive meetings were full practice clinical ones. These teams sometimes had additional single discipline meetings. In this category, attached staff such as health visitors never attended any meetings and administrative staff never met with anyone outside of their single discipline meeting (mean team size 21, min. 8, max. 64).

Isolated: Teams which had nothing which could be categorised as a team meeting. In these cases, the only cross-disciplinary meetings might mix practice nurses and health visitors. These teams tended to report some single discipline meetings (mean team size 24, min. 17, max. 37).

Of these types, multiplex, unitary-plus, and practice-based were the most common, with relatively few teams having just a whole team meeting (the unitary category) or no true cross-disciplinary meetings (isolated). (See Figure 11.2).

Figure 11.2: Proportions of multiplex, unitary-plus, unitary, practice-based, and isolated Teams

9% 29% Multiplex 27% Unitary-plus Unitary Practice-based Isolated

7%

28%

In practice-based and isolated teams, communications with attached staff could be insufficient, causing inefficiencies and lack of direction. Unitary teams might require a great deal of informal communication to supplement meetings; this is more likely to be a successful strategy for fairly small teams, which is in fact where the meeting practice tends to occur.

One might expect teams which have more inclusive meetings - those with whole team, full practice, or clinical staff meetings - to have higher self-reported team-working effectiveness scores, because these teams tend to have more chances for cross-disciplinary communication. Although the occurrence of a whole team

meeting is unrelated to self-reported teamworking effectiveness (unrelated t-test, t = -1.32, NS), the occurrence of full practice and clinical staff meetings is (for full practice meetings, t = -2.44, df = 6S, p<.O2 with a mean teamworking score of 4.66 for teams without a meeting and 5.30 for teams with one; for clinical staff meetings, Levene's F = 4.64, p <.04, with teams without a meeting having the more varied scores; t with unequal variances = -2.11, df = 48.78, p<.05; mean without meeting 4.63, mean with meeting 5.01). Since the meetings we recorded are whole team

meetings, we will return to why this might be the case after we have considered what form they take. Processes within meetings Meeting practice in PHCTs varies considerably. In this section we describe the meetings recorded in terms of who attended, how long they were, meeting practice, and interactional characteristics.

Interpreting date relating to the relationships between communication in the recorded sample and other variables such as effectiveness and team processes, is challenging. Primary health care teams are highly variable, not just in terms of input factors, involving team context. For instance, one of the recorded teams, identified as Team A, was made up half of new staff who had been brought in with the explicit aim of changing the ethos of the practice. We observed that the atmosphere in this practice was quite tense, and, they were one of only three teams in the wider sample not to hold a Christmas party. This team had very low self-report effectiveness

scores compared to the rest of the recorded sub sample for all but the sub scale reflecting professional delivery of care. They also declined to have a second meeting recorded. For this reason, the team was omitted from analyses involving In addition,

effectiveness and team processes variables, but shown in graphs.

differences in meeting practice made it necessary to omit further teams for certain parts of the analysis.

Because of the small sample size, it is not always possible to test whether assumptions of normality and linearity are warranted for our statistical analysis. In

order to make the analysis more robust, we categorise teams into two sets, high and low, for each of the properties of communication which we investigate, and employ tests. Where these show a difference, we then go on to show the relationship graphically and to characterise it using correlations.

In most cases, the teams held one meeting which they considered to be for the whole team. These meetings were open, with all team members expected to attend. Although the meeting remit was not always clear, the teams used these meetings as their opportunity to discuss matters affecting the practice. There were three obvious exceptions. One of the teams, identified as Team B, was strictly controlled by the partners and never held multi-disciplinary meetings. Their team meetings were

attended by the partners and practice manager only, and even took place away from the practice, in one of the partners' kitchens. This team had very low effectiveness and team processes scores compared to the other teams in the sub sample. Like Team A, they also declined to have a second meeting recorded. Another team, identified as Team C, did not hold one whole team meeting, but had two highly multidisciplinary sub-teams with specific remits which were meant to improve the working of the practice. One sub-team discussed how to make the best use of the nursing staff within the practice, while the other discussed initiatives to improve preventative care (for instance, an anti-smoking campaign). A third team reported that they held team meetings and gave us permission to record them. However when we

attempted to arrange to record meetings, the team claimed that they were not holding any meetings which would be appropriate. This team had average self-reported effectiveness scores.

Figure 11.3: Length of recorded PHCT meetings

meeting length (minutes) The recorded sample is larger than the set of analysed meetings due to recording difficulties.

PHCTs had regular time set aside for weekly or monthly team meeting. However they were quite often rearranged or cancelled completely. Meetings often started late, with people coming in late and leaving early in order to complete their other duties. Many of those attending were silent throughout and appeared bored: many commented informally to the researcher that the issues discussed in meetings were irrelevant to them. Meeting agendas were quite vague. Early business in the

meetings, tabled on the agenda, tended to consist of items which the practice manager felt it was important to discuss. However, most of the meeting time was

taken up by "any other business" raised by other people present. In many cases, items which the practice manager were put off in order to accommodate unscheduled discussion. We observed that for many teams, items raised at the beginning of the first meeting we recorded had not been dealt with by the end of the second meeting. Only one team kept rigidly to the agenda, with no unscheduled discussion. Meeting chairs were usually practice managers, GPs, or practice nurses. (See Figure 11.4). However, the degree of control which chairs exerted over the meetings varied considerably, with some chairs, particularly those who were nursery and administrative staff, merely announced the next step on the agenda as prior discussion came to a close. Figure 11.4: Who chaired the recorded PHCT meetings, by occupational class

Who cha ir ed t he r eco r d ed PHCT m eet ing s, b y o ccup a t io na l cla ss.

ADMIN

MISC GP

ATTACHED

PN

PM

In the meetings we recorded the decisions made which affected the entire team. Often the issues discussed were logistical; financial or business issues were discussed in different meetings. The official agenda in individual meetings tended to focus on one or two large issues, such as auditing team performance or clinic management. However, the majority of meeting time was spent discussing less weighty, more social issues such as what to do on practice nights out, whether to have a fish tank in the waiting room, and where to go for a Christmas party. These discussions rarely remained focused therefore they tended to take up more time than the critical issues.

Although the meetings we recorded had been identified by teams as decisionmaking, we found that decisions were not made in the meetings. Where decisions were made, they were often about how to proceed with the issues discussed; for instance, in these meetings, the team might decide to call another, often smaller, meeting for more discussion. Major decisions affecting the team members, such as changing a clinic date or recruiting more staff, were taken in a different forum and reported back to the 'decision making' meetings. Thus these meetings were largely for exchanging information and a forum so that there would be a place where team members could express their opinions. Some of the individual participants complained informally to the researcher carrying out the recordings that the meetings were boring and that the issues which they addressed were completely irrelevant to them. Figure 11.5: Primary health care meetings vary considerably in size

Number of People attending PHCT meetings

.
M e e ti n g s i z e

. .
20 75 100 125 200 225 300

Number of People

To score general attendance, one can use the average proportion of team members who attended recorded meetings. All of the recorded teams considered their meetings to be open to all team members except for Team B, which openly restricted

attendance to GPs and the practice manager, and Team C, which was organised into sub-teams. Omitting these two teams, there is a relationship between team size and the proportion of members attending recorded meetings (see Figure 11.60 (dividing teams into two sets, small and large, (t = -3.64, df = 7, p = .01 two-tailed; small teams have the higher proportions). Figure 11.6: The relationship between team size and meeting attendance for whole team meetings

.9 .8 .7 .6 .5 .4 .3 .2 .1 10

The relationship between team size and meeting attendance for whole team meetings.

Team C Other teams 20 30 40 50

team size
Team B is omitted from the graph because team size is unavailable. Fit shown without Team C.

One likely reason for this is the workload of health care team members. The bigger the team, the harder it is to schedule meetings at times that are suitable for everyone, and the more likely is that members will be unable to attend. This needs to be taken into account when interpreting results based on this general attendance score. Team A was omitted from the analysis because it had disproportionately low effectiveness and Teams B and C because they did not hold whole team meetings. When the average attendance was divided into two sets, low and high, a relationship was found with one of the self- reported effectiveness sub scales, patientcentredness of care (see Figure 11.7) (t = 2.42, df=6, p= 0.5 two tailed). Among the teams that have a higher proportion of team members attending meetings, the effectiveness score is higher. If a linear correlation between general attendance and

this effectiveness sub scale is assumed, the same result emerged (r = .73, df = 8, p = .04). This is the case despite the fact that there is no relationship between team size and patient-centredness of care, either in the sample of teams we recorded or in the wider sample. Figure 11.7: The relationship between general attendance and patient-centredness of care
The relationship between general attendance and patient-centredness of care.
6.5

6.0

5.5

5.0

4.5 Team A 4.0 .1 .2 .3 .4 .5 .6 .7 .8 .9 Other Teams

average proportion of team members attending meetings

Teams B and C have been omitted because their meetings are not open to the whole team. Fit shown without Team A.

These results show that higher levels of attendance at meetings is associated with effectiveness with respect to quality of patient care. The explanation for this, even though there are no relationships with the other effective variable, can be found if we consider how the team use meetings. Even during the meetings, agenda items were often delayed while team members passed on information about individual patients to other members who were also involved in their care. Team members also used the time just before and after meetings to have such discussions, although there the opportunities were less certain because people often came late or left early. Although these discussions were only useful to a few of the people present and therefore might be seen as wasting team time, this was the only opportunity many had to exchange information.

Team members from the differing occupational groups were not equally likely to go to the meetings recorded; whereas GPs and practice managers nearly always attended.

(See Figure 11.8). The graph includes team members whether they work full or parttime with the team. Although some categories are more likely to be part-time than others, and therefore have difficulty attending meetings, all team members still need some opportunity to communicate with each other. Figure 11.8: Who attended at least one meeting (in solid), by occupational class, versus total team membership (complete bars)

Who attended at least one meeting (in solid), by occupational class, versus total team membership (complete bars).
90 80 70 nu m be r in sa m pl e 60 50 40 30 20 10 0 GP PM PN ATTACHED ADMIN MISC

occupational class

In addition, not all meetings recorded had at least one representative from each of the constituent disciplines; whereas all meetings had at least one GP and one practice manager present, nearly half the recorded meetings did not involve any attached staff.

Figure 11.9: For each occupational class, percentage of meetings recorded which had someone from that class attending

F o r e a c h o c c u p a t io n a l c la s s , p e r c e n t a g e o f m e e t in g s r e c o r d e d w h i c h h a d s o m e o n e f r o m t h a t c la s s a t t e n d in g .
120%

100%

80%

60%

40%

20%

0% GP PM PN A TT A D M IN M IS C

oc c upat i onal c l as s

Team B deliberately excluded all members except GPs and the practice manager. Two teams where we were able to record two meetings did not have any attached staff present at either one. Some practice managers remarked informally that for some individuals, failure to attend was quite regular and tended to cause resentment among the other team members.

The meetings which we recorded were in all cases the largest and most multi-disciplinary meetings which the teams held. As a result, we can use who

attended the recorded meetings as a measure of integration in between the different disciplines. Although team meetings were only one of many ways in which a team communicated internally, it is likely that team members and disciplines that had poor attendance at meetings would be less well-informed and less involved in decisionmaking. This was most likely to apply to attached staff such as health visitors and district nurses. Not only were they least likely to attend team meetings, as the

interview data shows, in most team these were the only meetings they were expected to attend. Much of the work of attached staff was carried out away from the practice premises, therefore they were least able to communicate with other team

members in other ways. From the content of the meetings it was clear that because attached staff do most of the home visits for the practice more than other team members, they were more in touch with both the circumstances of individual patients and patient needs in general. The recordings of team meetings also suggested that detached staff were the most aware of inefficiencies within the practice arising from GP's performing tasks that other team members could do (for instance/GPs making home visits on the same day as an attached team member). Thus for this analysis, we use two measures to reflect the discipline mix of meeting attendance. The first, a score for multi-disciplinarity of meetings, is the average number of occupational categories present at a team's recorded meetings, out of our list of five. The second is simply whether or not at least one attached staff member was present at one of the meetings we recorded. These scores are highly related, since no teams ever had an attached staff member present unless all of the other disciplines were represented as well; that is, for teams that ever had an attached staff member present, the multidisciplinarity rating was over 4.

There is no relationship between self-reported effectiveness and either of these measures. However, the data suggest a relationship between self-reported support for innovation and multi-disciplinarity of team meetings, if we omit Team A on the grounds of its disproportionately poor team processes. If Team A is omitted, support for innovation is higher when attached staff are present for at least one of the team's recorded meetings (t = -3.76, df = 8, p = .006 two-tailed).

Dividing the multi-disciplinarity score into two sets, low and high, more multidisciplinary teams have higher support for innovation (t = -2.8, df = 8, p = .02 twotailed; r = .8347, df = 10, p = .003 two-tailed).

Figure 11.10: Fit shown with (solid) and without (dotted) Team A

A.
4.0 3.8 3.6 3.4 3.2 3.0 2.8 * 1 .5

it sh o w n w ith (so lid ) awithout (dotted) Team nd A

Team A O t h er t e am s

2 .0

2.5

3.0

3.5

4.0

4 .5

5.0

5 .5

number of occupational categories present

A t least o ne attached mem ber w as p resen t fo r at least o n e m eetin g if nu mb er o f catego ries presen t ex ceed s 4.

This result suggests that team members viewed these meetings as their chance to raise new ideas within the team. For this sample, as for the wider study, only around a quarter of the team members providing self-reports for team processes were attached staff (in this sample, mean 22.56%, min 8.33%~ max 38.46%). Therefore it is unlikely that the differences were a result of attached staff themselves reporting that they feel the team supports innovation; a more likely explanation is that their presence affects the entire team.

We used the 'freedom of interaction' and 'equality of participation' scores previously to study interactions during meetings in small to medium-sized manufacturing firms. In that study we confirmed that the scores differentiate groups which operate as teams, with equal responsibility among the members, from groups in which one person has overall authority. Equality of participation and freedom of interaction were higher for the teams, showing that they engage in freer discussion. Primary health care teams behave like equal responsibility teams and not like the managed groups. Restricting consideration to teams in the same size range (fewer than

thirteen members), PHCT scores are higher and less varied than the industry scores as a complete set (for equality of participation, F = 6.725, p = .014 two-tailed; t = 3.76, df = 3l.53, p = .001 two-tailed; for freedom of interaction, F = 5.028, p = .032 two-tailed; t = 2.50, df = 33.05, p = .017 two-tailed), but indistinguishable from the

equal responsibility subset (for equality of participation, F = .72l, NS; t = l.57, df = 22, NS; for freedom of interaction, F = 3.582, NS; t = .60, df = 22, NS). Although the teams have nominal chairs, for the most part the meetings are not strictly led. This is surprising because it is generally difficult to have free discussions in such large groups. In addition, status differences tend to make interaction less free, and GPs are both traditionally high status and the employers of many of the team members. Under the circumstances, if free discussion is what is required, these teams are doing better at allowing them to occur than one would ordinarily expect.

Conclusion In primary health care teams, good general attendance at team meetings was linked to self-reported patient-centredness of care. An explanation for this seems to be that meetings gave individuals the opportunity to have conversations and exchange information about patients. However, whole team meetings do not appear to make the team believe they are more effective in other ways. In particular, teams that have whole team meetings do not believe they are any better at teamworking than teams that do not, even though full practice meetings and clinical staff meetings do improve a team's impression of their teamworking skills. These differences may arise from a sense of the purpose of a meeting. Team members may have felt that it was

important to meet, but have been unsure about who should go to the meeting and what should be discussed. Being aware that a meeting is necessary requires that team members know that people need to communicate, but knowing how to communicate requires more preparation. Where there is uncertainty about the

purpose of the meeting, the practice was to suggest that everyone (or at least representatives from each staff group) attended just in case something important was discussed. It was only possible to know which staff could be excluded when the remit of the meeting was clear. Although whole team meetings could be useful if they had a clear purpose that included everyone who attended, the meetings that we observed often did not have this character. Instead, because of the lack of direction, many staff members saw the meetings as irrelevant and a waste of their time. Attending meetings that are seen as irrelevant may have a demoralising effect on staff with further ramifications for the team's work. Therefore it is important for teams to consider their meeting practice and to make sure it is designed to best fit their circumstances.

Chapter 12
Analysis of Communication in CMHT's
_____________________________________________________________________

Summary of Findings

CMHT members spent three times more time in meetings than PHCT members

The more cross-disciplinary meetings held in a CMHT the lower the stress levels in the team

CMHT meetings were generally well organised and multi-disciplinary

In 90% of CMHT meetings effective group decisions were taken

The prototypical CMHT meeting combined operational and clinical decision making and contained about 9-13 members

CPNs and Social workers were the best represented in the meetings but there was also regular attendance of occupational therapists, psychologists and psychiatrists

Types of Meetings The number of different meetings was larger and more varied in community mental health teams than in primary health care teams (range 1 to 11, mean 4.36, SD 1.77). (See Figure 12.1). Figure 12.1: Number of meetings

30

20

10

0 1 2 3 4 5 6 7 8 10 11

Number of meetings in total

Teams identified over 20 types of meeting, which we considered to fall within four categories: Clinical, including audit/quality; day care meetings; ward rounds; representation at PHCT meetings; CPA reviews; allocation; referrals

Operational, including MDT meetings; business; locality, sector or patch meetings; team leader meetings; management; communication

Strategy, including planning meetings, away days, and team building Professional development, including education or training meetings;

professional group meetings; supervision; support Some of the meetings described, for example, PHCT, ward rounds, locality, sector, patch, or team leader meetings, were not strictly team meetings, although the CMHT was represented and received clinical, operational or strategic input from such

meetings. There is some confusion between clinical and operational meetings in the data set because teams tended to mix these two functions, for instance, by calling one meeting which first performed case allocation and then handled business issues. 81% of teams reported having clinical meetings and 97% reported having operational meetings; these two types of meetings are probably ubiquitous, with the teams reporting no meetings of a type performing that function as part of another meeting. Strategy and professional development meetings were reported by 16% and 48% of the teams, respectively. Probably because they held so many different meetings, counting up the number of minutes per month that at least part of the team is in a meeting gives an average of around 1000 minutes per month (range 140-2940, S.D. 608). This means that on average, there was a meeting happening a tenth of the time involving at least part of the team. Because of the variety of meeting types, it is less useful to characterise team practice as a whole in terms of the set of meetings which a team holds than to consider which disciplines engage in meetings with each other. In these teams, communication was usually very strong across the constituent disciplines. Teams generally fell into one of four categories of practice. (See Figure 12.2).

Figure 12.2: Four Categories of Practice

4 5

Teams exhibiting complete connectivity might have some direct links missing --- for instance, in the example shown, managers never met directly with occupational therapists --- but all disciplines were involved in some cross-disciplinary meetings. In just under half of these teams, all disciplines encountered all others in meetings. Team with one isolate exhibited complete connectivity for five of the six disciplines, but one discipline was never involved in cross-disciplinary meetings. In two-thirds of these cases, the isolated discipline was management; the remaining cases were distributed evenly among psychiatry, occupational therapy, and psychology. Teams with a psychiatry + nursing + social services axis showed good connectivity for these three disciplines. Just over half of these cases only ever had these three disciplines communicate together in meetings. In the remaining cases, these three disciplines were included, but so was one other, with all the others equally likely to be the additional inclusion. Teams with a nursing + social services axis showed good connectivity between these two disciplines, and also usually included cross-disciplinary meetings with one or two other disciplines, but never with psychiatry. In over half of these cases, occupational therapists were involved in meetings, but there were also examples with management and psychology involvement.

4 5

4 5

4 5

All of these categories were reasonably common in the sample, but the categories showing better overall connections were more prevalent:

Figure 12.3: Cross-disciplinary communication in CMHT meetings

Cross-disciplinary communication in CMHT Meetings

12%

33%

Complete connectivity One isolate

23% Psychiatry+nursing+social services axis Nursing+social services axis

32%

Processes in Meetings

What we can see from this analysis is that in terms of communication in meetings, nursing and social services staff tend to form the core of the team, with psychiatry in close contact and management most likely to be isolated. Although what happens in meetings is not necessarily indicative of communications in the team as a whole, meetings provide opportunities to discuss work and develop good relationships not just during the meetings themselves, but also beforehand and afterwards. Therefore we would expect this pattern to hold for the teams overall, even outside their meetings.

Which meeting communication pattern a team has is not completely arbitrary. Omitting three teams with more than 35 members and six teams which were strangely constituted (usually nursing-only teams rather than cross-disciplinary teams), freedom of interaction is related to commissioning (one-way ANOVA F (2,79) = 3.41, p<.05). In our descriptive analysis, the categories divided according to

whether psychiatry was connected to the core team and whether management was ever involved in cross-disciplinary meetings. Whether or not psychiatrists ever met with staff from social services is related to how the team was commissioned (X2 = 6.78, df = 2, p<.05).

Teams which are commissioned by health service and social services separately are more likely to have psychiatrists meet with social services than teams which are commissioned jointly or by the health service only. Whether managers were involved in meetings at all was also related to commissioning (X2 = 6.45, df = 2, p<.05 two-tailed); again, they were more likely to be involved in teams commissioned separately. The general pattern is that teams which are commissioned separately have stronger cross-disciplinary links than the other types.

The complete details of how our quantitative measures of communication in meetings relate to other variables within our theoretical model of teamworking are given in Chapters 1, 2 and 5. Summarising from this analysis, the number of minutes per

month a team's meetings takes is related to the length of time the team has been set up, the percentage of staff full-time in the one team, and the age of the team members (younger team members spend more time in meetings). Although we can not be certain why these relationships occur, they do have reasonable explanations. Part-time staff are simply harder to schedule into meetings than full-time ones. Teams which have younger and presumably less-experienced members may provide a somewhat less complicated service for their local clientele or may be less statusdifferentiated than other teams, and therefore meet together rather than dividing by function into smaller meetings. Mature teams have had more opportunity to structure themselves to fit their circumstances; they presumably divided their meetings by function to involve smaller sets of people, or, as sometimes happens, they may accumulate new meetings for new functions without remembering to end ones which are no longer useful.

Social workers are one of the key disciplines involved in meetings. If there are social workers in the CMHT itself rather than accessible from outside the team, then the overall meeting time for the team was lower. This may be because under these circumstances more of the team's communication can occur informally. The higher the percentage of men on the team, the more overall meeting time the team had; percentage of men is almost certainly standing in for some hidden variable, but the real cause is not clear. The more meeting time a team had, the more quickly the team deals with emergency referrals and the more quickly emergencies are seen; this is probably a result of being able to communicate information related to the emergency itself. Finally, the freer the interaction among disciplines exhibited by the set of meetings the team holds, the lower the average level of stress for team members and the lower a team's self-reported innovativeness. This relationship is

probably complex, since communication with colleagues should, by and large, reduce work stress and allow team members to express ideas which eventually turn into team innovations, but innovative teams undergo more changes than non-innovative teams, and change increases stress. Results derived from CMHT recorded meetings Meeting practice in CMHTs followed a more consistent pattern than with the PHCTs. Overall we found the communication in these meetings to be extremely effective. The meetings were used to make important group decisions and the content of the meetings was appropriate to their stated purpose. In this section we describe the recorded meetings in terms of their purpose, meeting practice, multidisciplinary representation, and the general interactional characteristics. We then use the analysis to identify good practice in such meetings and highlight what we believe to be important contributory factors toward good practice.

As in the case of the recorded PHCT meetings we did not think it appropriate to try and draw strong conclusions about the relationships between communication in the recorded sample and other variables such as effectiveness or team processes. The sample is not sufficiently large or diverse to do this. Instead we describe the results in more qualitative terms and use them to identify prototypical meeting practices in a CMHT. On the basis of the purpose, content and general interactional characteristics of the meetings we then define a good practice prototype for CMHT meetings of the kind recorded. The purpose of the recorded CMHT meetings

Teams were asked to select for recording routine meetings with strong multidisciplinary membership. Generally, they chose meetings that fell into the operational category described earlier. So the meetings were typically weekly team business meetings, but frequently they also had a clinical component. We recorded 18 meetings from 9 teams and where possible ensured that they were two consecutive meetings of the same type from each team in the sample. Unfortunately, for one of the meetings there was a technical problem in recording so it had to be dropped from the sample. Hence, the analysis was based on a sample of 17 meetings in 9 CMHTs.

Although the meetings fell into the operational category, as business meetings, they did vary in terms of purpose and this affected their style. In one case the team selected for recording two special meetings designed to respond to and influence health department policy on CMHTs. Membership was much larger than for other meetings in the sample and was quite different. For example, it included health care managers and a ward manager. The content and communicative style of these meetings was also quite different from others in the sample. We refer to these as High Level Policy (HLP) meetings to differentiate them from the others.

A second kind of meeting associated with two of the teams involved dissemination of policies imposed from above. The purpose of these meetings was to ensure a detailed understanding of Trust or DoH policy initiatives. In many respects

communication in these particular meetings was less effective than in others. For one of the teams much of the meeting time was spent reading documents that would have been better dealt with outside the meetings and decisions about how to deal with the policy initiatives were regularly put off to subsequent meetings. However, this team had adopted a rotating chair practice for their meetings and this limited the ability of the team leader to control the teams decision making. It could well have been this factor which led to the apparent ineffectiveness of the decision making. We take up this issue in the section on meeting practice. We refer to this kind of meeting as a Policy Dissemination (PD) meeting.

The most frequent kind of meeting in the sample was the weekly team meeting in which both clinical issues, such as deferrals, and team policy were formulated. Typically, teams split the meeting into a section on clinical reporting and case allocation and a subsequent business section. In most cases the business section of the meeting dealt with team operational issues rather than high-level policy issues. We shall refer to these as Mixed Purpose (MP) meetings.

Running meetings in CMHTs

Members of CMHTs spend much more time at meetings than do members of PHCTs. This greater experience is reflected in a generally high standard of meeting management. In the majority of recorded cases, team meetings were held regularly, they were well organised and chaired, and had clear agendas. However, there was some variation in the organisation and style of chairing that did affect the quality of the decision making at the meetings.

Three important factors in the running of meetings are their regularity, their size and their duration. In all cases except the rather anomalous HLP type, the team meetings took place on a weekly basis. The HLP meetings only occurred every two months. Twenty-one people attended the HLP meetings, but this was also not representative of the sample. For the meetings as a whole the average size was 11 and it varied between 5 and 21. (See Figure 12.4). The average duration of the meetings was 71 minutes and it ranged from 38 to 140 minutes. (See Figure 12.5). In most cases the meetings were scheduled for no more than an hour, but there were a few occasions when this was extended to two hours for both clinical and business meetings.

Figure 12.4: Size of meetings in terms of number of people present

People present 21
Number of people

18 15 12 9 6 3 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Meeting

Figure 12.5: Duration of meetings in minutes

140 120

Length of Meeting

100 80 60 40 20 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

Meeting

For multidisciplinary decision making teams choosing the right size of meetings is always going to be difficult. On the one hand, groups of more than about 8 or 9 tend to fall into a non-interactive mode of communication in which each speaker broadcasts information to the rest of the group. This hinders the process of finding a true consensus in the group. On the other hand, it is important to have a sufficient number of people present to ensure effective representation of the different disciplinary interests. In this situation meetings should be restricted to no more than about 12 or 13 members and there needs to be an experienced chair. The chair can then ensure that all relevant voices are heard and promote interactive discussion when it is needed. This was the practice in a large proportion of the MP type

meetings that we recorded.

For example, in one of the MP meetings the team was confronted with a crucial decision about how to respond as a team to a change in the sectors they were to cover. This change, which had been introduced without consultation, meant that there would be a reduced overlap between the areas covered by team members. The question was whether they should respond by splitting into two separate groups for their meetings or remain as a single group. The meeting contained 13 members and up until this point was characterised by a non-interactive broadcast style of discussion. So the team leader and chair of the meeting first promoted an extensive discussion of this issue in which he did not intervene. Then, when everyone had had

their say, he went round the whole group and allowed each member to indicate their feelings on the issue before confirming the decision. So although the group contained 13 members it was possible to establish a clear consensual decision about this important matter. Interestingly, this meeting also managed to get through as much business as most of the others. Yet, this was all achieved in 40minutes; it was one of the shortest meetings in our sample.

By contrast, one of the much larger HLP meetings (21 members) faced a situation in which a group decision should have been made, but it was left up in the air. A young CNP raised a problem of risk management and safety in the light of a recent traumatic experience. She had been on a routine visit and on arrival confronted a suspected suicide. Because she had no portable phone and was in a poorly provisioned area of the City, she had to return to base before being able to call for support. The question confronting the meeting was whether to push for provision of portable phones to all staff engaged in domiciliary visits. This item received considerable discussion, but because the group was so large the discussion amounted to a series of long broadcasts where different members expressed their opinions on everything to do with risk management. After nearly 10 minutes discussion the problem had not been resolved and was not deferred for subsequent decision. This does not reflect on the will of the people at the meeting or the ability of the chair, but rather on the extreme difficulty of making effective group decisions in meetings of 21 members. In a smaller MP meeting, when confronted with a less dramatic example of the same problem, the team managed to come up with a coordinated policy, including the provision of portable phones, for just such cases as this one.

The point illustrated here is that meeting size is crucial to effective group decision making and that even with relatively small groups of around 12 or 13 members the process relies on skilful chairing. In relation to this point, two of the teams in our sample adopted a practice of rotating chairs for meetings. Both had relatively small meetings (between 6 and 9 members at each). However, it was apparent that the quality of the group decision making was affected by the practice. For instance, in one case there was real confusion about who was to monitor and control the decision; whether it should be the team manager or the chair. In another case at least 5 minutes was wasted establishing who was to chair the meeting and who to take the minutes. Although it may seem helpful to give members experience of

chairing meetings, such cases illustrate that ineffective chairing will certainly reduce the effectiveness of the teams decision making processes.

For about half of the teams in this sample meeting size was restricted to less than 9 members, which is quite an appropriate size for effective interactive discussion and group decision.

The content of CMHT meetings and decision making

The content of the recorded meetings fit in with the goals of the meetings. There were three main areas of discussion: policy, team operations and clinical allocations. The proportion of meetings in the sample that covered each of these topics is shown in the Figure 12.6.

The three broad categories of meetings HLP, PD and MPD discussed the three kinds of issue according to their goals. The HLP meeting predominantly discussed a green paper on mental health care provision in their city. This was quite appropriate to the meeting because the team had been chosen to elicit feedback from relevant community groups on the content of the paper. The PD type meetings also discussed policy, but more in the context of detailed policy documents that had been sent to their team leader. In MP type meetings there was often also reference to policy, but only in so far as it was pertinent to particular issues arising from either clinical cases or team operation.

The second main topic of discussion was what we have called team operation: by this we mean practices or policies to be adopted by the team that affect the way the team works. Not surprisingly this topic arose in most of the meetings we recorded. It was also the source of most team decision making that occurred in the meetings. Finally, the MP teams also discussed clinical matters. In the meetings we recorded most clinical (i.e., patient oriented) discussion concerned allocation of cases. However, there was also discussion of particular problems associated with difficult cases. For example, in one such case a sectioned patient had been on leave in her hometown in India. She had written to indicate that she was being held by the family against her will and was requesting repatriation as a British subject. The team had to work out an appropriate response to this situation.

Figure 12.6: Content and decision making in CMHT team meetings in terms of % meetings discussing these topics and making group decisions

90% 80% 70%

% of meetings

60% 50% 40% 30% 20% 10% 0% Policy Team operation Clinical Group Decision

Type of decision

The figure also shows the proportion of meetings which resulted in 1 or more group decisions. As can be seen such decisions were made in 88% of the recorded meetings. These decisions varied from straightforward matters of how to co-ordinate reports between the nurses, social workers and occupational therapists to more complicated matters such as formulating an effective security policy for staff on domiciliary visits. Interestingly these two issues arose in a number of the meetings and the teams tended to come up with slightly different solutions. The diversity of decision in operational matters reflected the different circumstances of the teams and seemed perfectly appropriate.

Attendance at meetings

The CMHT meetings were attended by a broad range of different categories of staff representing different professional groups. The figure shows percentages of meetings attended by each of the major staff categories (see Figure 12.7). Figure 12.7: Percentage of meetings with a representative from each staff category

% of meetings

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%
ps ych iat ris ts nu oc rsi cu ng ps pa ych sta tio na ff olo l th gis era ts, pis ps ts ych oth era p.. . sta ff f ma rom na ge so rs cia l se mi rvi sce ce llan s eo us sta ff

All the meetings had representation of both the nursing staff and the social services. In fact, community psychiatric nurses and social workers were nearly always in the majority at the meetings. Across all meetings 37% of attendees were CPNs and 24% were social workers. In more than half of the meetings there were also occupational therapists, but they only represented 9% of the membership across all meetings. The two other staff categories of psychiatrists and psychologists were less well represented. Psychiatrists represented 6% of the attendees and psychologists only 4%.

In a situation where different disciplines are associated with marked differences in status (e.g., between consultant psychiatrists and CPNs) multidisciplinary

representation at meetings can become a barrier to group decision making. On the one hand, it is important to have representation from as many groups as possible; but on the other hand, it is also important to minimise any major disparities in the

status of the group members. This is especially true for larger groups in which high status dominant speakers exert a disproportionate influence on the discussion. This means that notional group decisions tend to become individual decisions that do not reflect the broader interests of the group.

So it is interesting that there were psychiatrists at only 1 in 10 of the meetings. Furthermore, we had the strong impression that the meetings in which they were present were not so effective at making group decisions. They tended to be longer (average 91 minutes compared to 72 minutes for the others) and generally the communication was less interactive.

One strategy adopted by CMHTs was to have nursing staff from the hospitals represent the consultants cases at the meeting. This seemed to be a very effective way of ensuring broader representation without introducing marked disparities of status within the meeting itself.

Communication dynamics in CMHT

Unfortunately, we encountered some problems with the quality of the video recording for CMHT meetings. This made it difficult to establish quantitative measures of the degree of interaction of the kind made with the PHCT meetings. Also the mixed nature of many meetings with a clinical reporting section followed by a business section would have made it difficult to interpret overall measures of interactivity and participation.

From other sources of information, such as the relative length of contributions it is possible to draw general conclusions about the interactivity of the discussion in the meetings in the sample. The major contributory factor to interactivity was simply the size of the meeting. In general, the larger the discussion group the lower the interactivity in the discussion and this was reflected in the CMHT meetings.

The prototypical CMHT meeting: recommendations on good practice

On the basis of this sample, which represents around 20 hours of discussion in CMHTs, it is possible to define a prototypical meeting. This can be used both to

describe the overall nature of CMHT team meetings as they occur in our sample and as a framework for giving recommendations about good practice.

Within our sample of meetings a clear pattern of practice emerged that corresponded to the results from the interview data. The most common type of meeting was an operational meeting, which included both a clinical component and a separate business component. This was the case for more than half of the recorded meetings. In terms of size this prototypical type of meeting ranged from 5 to 13 members. In terms of multi-disciplinary representation it fit into the one isolate style of meeting. In other words, the meeting always contained representatives from nursing and social work but had either psychiatrists or occupation therapists or psychologists not represented. We feel that this is probably quite satisfactory in terms of group decision making because it allows for a sufficiently broad representation without compromising the size of the group or producing disparities in status.

In relation to this prototype, meetings with around 9-13 members were generally the most satisfactory for both routine reporting and case management as well as more general group decision making. However, for the larger meetings to be effective there had to be skilled chairing of the meetings. We would strongly recommend that meetings be chaired by team leaders where possible and that they be given some training in effective meeting management.

In contrast, the three meetings with membership of 15 or more were noticeably less effective in terms of decision making. It seemed that the very small team meetings with 5 or 6 members were also less effective. Despite often being very interactive in terms of the discussion they seemed to have more trouble making group decisions and on one occasion ended in conflict between two of the groups being represented. This was the only occasion in all the recorded meetings where such a conflict arose.

Conclusions The prototypical CMHT meeting contains between 9 and 13 members, it combines clinical discussion with team business and is used to make team decisions. Communication in community mental health care teams was much more effective than in the primary health care teams. On average team members spent three times as much time at meetings as members of PCHTs. Also, there was strong crossdisciplinary interaction over the range of meetings that they held. Interestingly, the

freedom of cross-disciplinary interaction was associated with reductions in team stress levels (as derived from the interview data). Furthermore, the meetings we recorded showed evidence of high quality chairing and were generally well organised. Unlike many of the PHCT meetings group decisions were made in 90% of the meetings and many teams came up with effective ways of ensuring multidisciplinary representation without introducing too much disparity in the status of the members.

General Conclusions In Chapters 11 and 12 we have looked at communication in both primary and community mental health care teams. The analysis was based both on the interview data from a large sample of teams in the two health sectors and the detailed analysis of recorded team meetings from a sub-sample of teams. The findings highlight major differences in the quality of communication and group decision making between PHCTs and CMHTs. In PHCTs there are few occasions where the whole team gets together in a recognized forum to discuss their activities as a team. When they do so in team meetings the communication and decision making is not particularly effective. In less than half the meetings we recorded not group decisions were taken. Team members who are attached to a practice regular do not attend the team meetings. Yet, there was evidence that teams were they do attend have stronger support for innovation. To the extent that the team meetings are effective, they enable pairs of members to sort out bilateral problems, such as coordination of patient visits. In contrast, meetings play an important role in the day-to-day operation of CMHTs. On the basis of the interview data it seems that members spend about 3 times as much of their time at meetings as do members of PHCTs. Most of the meetings that they attend are multidisciplinary and there is generally good connectivity across the different disciplines in teams at these meetings. Overall, the recorded meetings were well managed and spent most of their time discussing topics on their agendas. In about 90% of meetings appropriate group decisions were made and there was evidence in some meetings of skilled management of this process by the chair. In relation to the primary health care teams we would recommend that some attention is given to training in the management of multidisciplinary meetings. We feel that PHCTs should recognize the importance of group decision-making processes in supporting an effective team. For CMHTs the communication and meeting practices are in general more professionally organized and on occasion match the highest standards observed in industrial team meetings in which the members have had training in communication and decision making. However, we would also like to underline how important it is for these meetings to be chaired effectively. This was one of the most striking observations from the CMHT meetings.

Chapter 13
Conclusions and Recommendations
We began this research programme by asking whether team working in healthcare was associated with better quality patient care. We also asked what factors were associated with effective teamwork in healthcare organisations. Using a model that related team inputs and team processes to healthcare outcomes we were able to conduct research which has directly addressed these questions. The results paint a clear picture and provide firm answers to the questions we addressed.

Inputs such as team composition and organisational factors have a strong influence upon innovation and effectiveness in healthcare teams.

Team processes are directly related to the innovation and effectiveness of health care teams across sectors.

The quality of teamwork is directly and positively related to quality of patient care and innovation in healthcare.

There is a significant and negative relationship between the percentage of staff working in teams in acute hospitals and the mortality rate in those hospitals, taking account of local health care needs and hospital size1. Where more

employees work in teams the death rate among patients is significantly lower (calculated on the basis of the Sunday Times Mortality Index Dr. Foster; deaths within 30 days of emergency

surgery and deaths after admission for hip fracture)9.

Effective teamwork in primary healthcare teams is associated with lower stress among team members.

In those teams characterised by clear leadership, high levels of integration, good communication and effective team processes, team members have good mental health and low stress levels. In secondary health care settings the retention rates of staff are higher in those teams characterised by good team processes.

In this setting also, the research demonstrated that team membership itself confers role clarity and social support on team members, helping them to achieve better mental health or lower stress than their counterparts not working in clearly defined teams.

Good team processes means clear, shared objectives amongst team members; high levels of participation including frequency of interaction, quality of information sharing and shared influence over decision making; emphasis on high quality patient care within teams and a preparedness to encourage constructive controversy but to discourage interpersonal conflict; and practical support for ideas for new and improved ways for providing healthcare.

Effective and innovative teams are characterised by a pattern of reflexivity. Team members collectively and individually take time out to review the objectives, strategies and processes of the team; they prepare plans for making changes accordingly; and they implement those plans in action.

Such reflexivity leads to both better quality healthcare and higher levels of innovation.

This finding is based on separately funded research recently completed by the research team at the Aston Centre for Health Services Organisation Research (further details available from West or Borrill).

Leadership also emerges as critical in healthcare teams but is often absent. A single clear leader in highly complex large teams is associated with higher levels of effectiveness and innovations

As teams develop and become more sophisticated in their ability to work effectively, distributed leadership or shared leadership amongst different functions is associated with higher levels of effectiveness, innovation and better quality teamwork

Clear shared leadership is associated also with better team processes - clear objectives, participation, emphasis on quality, support for innovation and reflexivity.

But conflict over leadership is disastrous for teams.

Where conflict over

leadership exists, teams are ineffective, not innovative and team processes tend to be very poor.

The implications of our research are clear. First it is important that teamwork is encouraged in healthcare organisations and second that leaders are trained and encouraged to work effectively in teams. Leadership should be group-centred rather than traditional.

The traditional approach to leadership of healthcare teams is that the leader is responsible for the group and has control over final decisions. He or she guards their position power and perceives the group as individuals to be managed by the leader. The leader shapes the task for the team and He or she

ignores the socio-emotional processes within the teams.

discourages expression of needs or feelings during team meetings.

Our research suggests that healthcare teams needs group centred leaders who see responsibility as shared by both the leader and the team; where control over final decisions is vested in the team; where leader position power is de-emphasised; where the leader perceives the team as a collective entity and shares responsibility for shaping the tasks of the team. Moreover, the team leader should emphasise and share with the group the responsibility for maintaining the group as a social entity. He or she should

closely observe socio-emotional processes in the group and encourages discussion in meetings of team members' needs and feelings.

The research reveals that multi-professional working is associated with high levels of innovation in healthcare. Where a variety of professional groups are involved in healthcare teams it is much more likely that innovation will be a consequence. This is particularly so when the team processes are healthy.

Larger teams are seen as more innovative and effective, partly because they have the resources, organisational structures and processes in place which enable radical changes in the delivery of healthcare to be accomplished.

Policy makers should be cautious about how they respond to these findings. Teams cease to exist above around twelve to 14 members. They become small organisations. Primary health care "teams" consisting of around 40 members are in effect medium sized enterprises. As these teams transform at around 15 to 20 members into organisations it is likely that they will divide into sub-teams. This is a normal structural process in organisations.

Leaders and managers then have to make decisions about the appropriate sub-teams to be formed. These should be formed taking account of the team's context and tasks. The purpose of these sub-teams should be to manage uncertainties in relation to particular patient groups, particular stakeholder groups and particular illness categories.

A wise differentiation within healthcare teams is necessary as they develop into organisations. This differentiation or specialisation of activities is

necessary for the team to cope effectively with its environment. But at the same time it creates new pressures upon the organisation. The sub-groups must learn to integrate effectively with one another to communicate, liase and co-operate. The danger is that sub-teams, particularly if they are

composed of like professional groups, end up competing rather than collaborating and co-operating for the greater good of patients. Therefore, managers and leaders must ensure that these groups and sub-teams liase and collaborate together. It is harder to work as a single team as the

organisation grows in size and the required differentiation and integration processes necessary require sophisticated leadership. Leaders of health

care teams must therefore be trained in appropriate leadership knowledge, skills and attitudes.

These observations are reinforced by our finding that higher levels of integration in healthcare teams are associated with innovation and effectiveness. Communication, quality of meetings and information sharing activities to increase integration are all associated with innovation and effectiveness in healthcare teams. However, our research also reveals that quality of communication and meetings varies considerably. In general meetings are very poorly managed in primary health care and professionals in this domain have much to learn from their colleagues in community mental health teams. Indeed, another important conclusion which can be drawn from the research is the importance of training for those working in primary health care to better plan and manage their meetings.

Recommendations There are a number of key elements to effective teamwork (Guzzo and Shea, 1992). 1. Conditions for effective teamwork

First, Individuals should feel that they are important to the success of the team. When individuals feel that their work is not essential in a team, they are less likely to work effectively with others or to make strong efforts towards achieving team effectiveness. Roles should be developed in ways which make them indispensable and essential.

Individuals' roles in the team should be meaningful and intrinsically rewarding. Individuals tend to be more committed and creative if the tasks they are performing are engaging and challenging.

Teams should also have intrinsically interesting tasks to perform. Just as people work hard if the tasks they are asked to perform are intrinsically engaging and challenging, when teams have important and interesting tasks to perform, they are committed, motivated and co-operative (Hackman, 1990).

Individual contributions should be identifiable and subject to evaluation. People have to feel not only that their work is indispensable, but also that their performance is visible to other team members.

Above all there should be clear, shared team goals with built-in performance feedback. Research evidence shows that where teams are set clear targets at which to aim and they receive feedback on their performance, their performance is generally improved.

2. Selecting team members Regardless of their task specialism, there are certain attributes that all team members need to demonstrate if the team is to achieve its goal. Selection should focus not just on professional skills but also on knowledge skills and attitudes (KSAs) for teamworking. (See Figure 13).

Figure 13: Knowledge, Skills and Abilities for Teamworking

Conflict resolution KSAs

Collaborative problem solving KSAs

Communication KSAs

Fostering useful conflict, while eliminating dysfunctional conflict. Using integrative (winwin) strategies rather than distributive (winlose) strategies. Having the right level of participation e.g. for any given problem. Avoiding obstacles to team problem solving (e.g. domination by some team members). e.g. Employing communication patterns that maximise an open flow. Using an open and supportive style of communication. Using active listening techniques. Paying attention to non-verbal messages. e.g. . e.g. Setting specific, challenging and attainable team goals. Monitoring, evaluating and providing feedback on performance. Co-ordinating and synchronising tasks, activities and information. Establishing fair and balanced roles and workloads among team members.

Goal-setting and performance Management KSAs

Planning and task co-ordination KSAs e.g.

Source: Stevens & Campion, 1999 (Reprinted with permission).

3. Team composition Todays health care teams are being formed to respond to increased complexity and demands in the environment; and they bring together people from diverse professional backgrounds. Such diverse teams therefore embody different attitudes and working practices as a result of differences in age, gender, educational background, nationality, organisational culture, etc.

There is growing evidence, supported by the results from our research, that teams that are diverse in terms of skill and educational specialisation produce high levels of innovation and clear strategic thinking.

To date however it seems that other forms of diversity produce less positive effects on team performance. Teams whose members have diverse cultural backgrounds initially tend to perform more poorly than culturally homogenous teams, although this effect diminishes over time. Turnover rates are higher in teams which are diverse in terms of demographic features such as age, educational level, status and nonindustry work experience.

Diverse teams are not only advantageous if team performance is to be maximised: in current health care settings, it is simply not practical to select teams that are highly homogeneous. The challenge is to achieve the positive effects of diversity whilst building stable teams that will grow and develop together. It is essential therefore to provide induction and training for individual team members which will minimise the impact of differences that can be disruptive.

Susan Jackson (1996) has drawn a distinction between team-member differences that are task related (for example, educational level, work specialisation, organisational function) and those that are relations-oriented (e.g. gender, age, nationality, political views). It would seem that difficulties in teams are more often, in the short term at least, related to relations oriented-differences. In these areas, individuals will tend to make shallow or stereotyped decisions about others. Both awareness training and opportunities for social or informal contact between team members therefore play an important part in breaking down stereotypical reactions and developing more appropriate judgements.

4. Team Leadership

Leadership is creating alignment around shared objectives and strategies to attain them.

Leadership is increasing enthusiasm and excitement about the work and maintaining a sense of optimism and confidence.

Leadership is helping people appreciate each other and helping them to learn how to confront and resolve differences constructively.

Leadership is helping people to co-ordinate activities, continuously improve, develop their capabilities, encourage flexibility, encourage an objective analysis of processes, and foster collective learning about better ways to work together.

Leadership is representing the interests of the group or organisation, protecting its reputation, helping to establish trust with external stakeholders and helping to resolve conflicts between internal and external partners.

Leadership is creating a unique group or organisational identity.

Organisations that introduce team based working stress the importance of selecting the right people to lead teams at the very beginning. Leaders who find it difficult to move from a directive/controlling supervisory role to one of participative leadership can cause lasting problems. Those organisations which have successfully overcome this difficulty actively encourage all members of staff to apply for team leader positions. This begins the process of breaking down stereotypical thinking about who can lead teams, the criteria for application, etc. In addition, these organisations provided considerable training and support for new team leaders in the initial months of their appointment.

In the early stages of a teams development, training should be made available for team leaders to enable them to design and implement appropriate team processes and to develop the skills needed for effective team leading.

Team leaders need to be skilled in responding appropriately to meet the needs of their teams, i.e. to be more or less directive in supporting a team. Their aim should

always be to move as quickly as possible away from being directive and towards allowing the team to be autonomous. When supported by a team leader who provides an autonomous environment, a team can achieve more highly by becoming self-directing in its development and its work.

5. Organisational Support for Teams

Hackman and his colleagues at Harvard University have concluded that there are six principal areas within which teams need organisational support: targets, resources, information, education, feedback and technical/ process assistance in functioning. Examining the extent to which organisations provide team support in these areas can help in discovering the underlying causes of team difficulties.

Targets
Teams need support from an organisation in determining targets or objectives. Surprisingly few health care teams are given clear targets by their organisations often because organisational targets and aims have not been clarified sufficiently. It is striking, when team members are asked to outline their objectives and team targets, how few have clear notions of what is required of them. There is an implication that teams should derive their targets and objectives by scrutinising the organisational objectives or mission statements. However, these are often such vague good intentions or positive but abstract sentiments that it is almost impossible for a team to derive clear targets and objectives. Where, through a process of negotiation, teams are able to determine their targets in consultation and collaboration with those hierarchically above them, there is usually a better level of performance.

Resources
The organisation is required to provide adequate resources to enable the team to achieve its targets or objectives.

Resources include: having the right number and skill mix of people; adequate financial resources to enable effective functioning; secretarial or administrative support; adequate accommodation; adequate technical assistance and support (such as computers, blood pressure testing equipment, or appropriate equipment for testing infants' hearing, etc).

Information
Teams need information from the organisation which will enable them to achieve their targets and objectives. Changes in strategy or policy which are not communicated to teams can hamper their effective functioning. Ensuring that relevant information reaches a team to enable it to perform effectively is an essential component of an organisation's management. For example, GPs need to provide health visitors with ready access to age/ sex registers, medical records and other information about the practice population, in order for the health visitors to function effectively within the teams.

Education
Part of an organisation's responsibilities for effective team functioning is to provide the appropriate levels and content of education for staff within teams. The purpose of such training and education is to enable team members to contribute most effectively to team functioning and to develop as individuals. This includes on-the- job training, coaching via supervisor, training courses, residential training courses or distance learning courses. There should be adequate access to training which is relevant to the team's work and of a sufficient quality and quantity to enable them to perform to maximum effectiveness. And, as indicated above, team members should be trained in the knowledge, skills and abilities, for team working.

Feedback
Teams require timely and appropriate organisational feedback on their performance if they are to function effectively. Timely feedback means that it occurs as soon as possible after the team has performed its task, or occurs sufficiently regularly to enable the team to correct inappropriate practices or procedures. Appropriate feedback means that it is accurate and gives a clear picture of team performance. For some teams it is difficult to gain accurate feedback. For example, primary health care teams have almost no feedback at all. For a team responsible for providing

training in one division of, say, a major oil company, organisational feedback might take the form of senior managers' satisfaction with improved performance. This could include measuring the results from technical training courses in customer service in retail outlets (i.e. filling stations). Such information could come from surveys of

customer satisfaction with retail operators' services.

Clearly there are large

differences in the extent to which organisations can and do provide feedback to teams, but the aim should be for the organisation to improve continuously in the extent to which it provides useful, accurate and timely feedback to teams.

Technical and process assistance

Organisations have to provide the specialised knowledge and support which will enable teams to perform their work effectively. A primary health care team engaged in developing its practice objectives, by identifying the health needs of the practice population, might need the health authority to deploy a community medical officer to advise the team on patterns in local health and ill-health. For a training team in an oil company, technical assistance might take the form of specialist computing experts and marketing strategists, advising the company on how to communicate most effectively to managers throughout Europe, in order to market their training courses to managers in different functions.

Process assistance refers to the organisational help available when team process problems are encountered. Are consultants and facilitators available to help the team identify, diagnose and overcome problems of team functioning from time to time? But the implication of this work is that NHS organisations should not simply create teams. They should recreate themselves as team-based. this important issue. 6. Developing team-based organisations Teams working within team-based organisations have more discretion and scope than those working within traditionally managed organisations. In practice, teambased organisations reflect a management philosophy that incorporates certain fundamental principles. We now turn to address

In team-based organisations, most employees are clear about and committed to the objectives of the organisation as a whole. Senior management take time to communicate information to all employees about organisational objectives and also encourage team members and teams to influence the development of organisational objectives.

In team-based organisations employees are more fully involved. They are encouraged to contribute ideas, opinions and information to decision-making processes, and their teams have influence over decisions that are made. The organisation as a whole promotes acceptance of and commitment to processes of debate about how to perform work most effectively.

Managers

within

team-based

organisations

are

committed

to

encouraging

constructive debate within the organisation. They listen carefully to the views of team members and take time to explore diverse views and differences of opinion. They also encourage the expression of minority points of view and value opportunities for careful discussion about the best ways of delivering products and services.

In team-based organisations there must necessarily be a climate supportive of creativity and innovation. Teams are hothouses for creative ideas, and the organisation must encourage the expression and implementation of ideas for new and improved health care processes and ways of working. If it fails to do this, both the impetus for and the value of team-based working are lost.

To ensure the achievement of these aims, team-based organisations must reflect the belief that organisational goals will largely be achieved not by individuals working separately but by groups of people who share responsibility for outcomes and who work in efficient and effective teams.

In traditional organisations, there tend to be individual command structures with various status levels representing particular points in the hierarchy. There are supervisors, managers, senior managers, assistant chief executives and so on. In team-based organisations, the structures are collective. Teams orbit around the top management team or other senior teams, both influencing and being influenced rather than being directed or directive. The gravitational force of different teams affects the performance of the teams around them. This is a flexible, fluid structure in contrast to the mechanical, hierarchical structure of traditional organisations.

In traditional organisations, the manager monitors the performance of employees. In team-based organisations, the team monitors the performance of members within the team and the team as a whole is appraised by those it provides services and products for. Thus the Human Resource Management team may be appraised by all of the teams within the organisation for which it provides services.

In traditional organisations, power is invested in the hierarchy. The further up the hierarchy you go, the more power you find located there. In team-based organisations, the emphasis is on integration between teams and on reducing the number of levels in the organisation so that there is less vertical difference between different teams and groups. Whereas in traditional organisations the emphasis is on maintaining power and control through the use of a clear hierarchy of command (which may be important for example in an organisation dealing with crises), in teambased organisations the emphasis is on achieving shared purpose across teams and achieving shared understanding and integration across teams.

In traditional organisations the emphasis is on stability and keeping things the same. Rules and regulations, formalisation and bureaucracy encourage uniformity and control. In team-based organisations the emphasis is on encouraging innovation, change and flexibility in order that the organisation can adapt appropriately to its changing environment and be innovative health care services.

Traditional organisations tend to adopt one best way and to seek for universal models of effective organisational functioning. The team-based organisation emphasises its uniqueness, adopts ways of working that are appropriate to the organisation in its current circumstances, environment and economic context, and adapts as the environment changes.

In traditional organisations, managers manage and control; whereas in team-based organisations, the teams are self managing and take responsibility for setting their (perhaps in consultation with senior managers) and monitoring the effectiveness of their strategies and processes. Changes in the process of achieving the team-based organisation are therefore deep, wide and pervasive.

In the face of the inevitable complexities within organisational environments, within teams themselves and between the people who constitute those teams, there are no simple prescriptions for implementing effective team based working. In order to be effective, team members must therefore learn to reflect upon, and intelligently adapt to, their constantly changing circumstances as the team develops.

There are however certain areas where problems can be predicted and where effective initial design greatly improves the chances of success. Many of the

common problems in the introduction of team-based working (TBW) result from impatience: effective TBW takes time to implement and requires multiple changes that create almost inevitable difficulties. Long-term benefits can only be achieved through persistent and consistent action in each of these three key areas: organisational context team structure team processes

Organisational context The top management teams level of commitment towards TBW is a key factor in an organisation's introduction of TBW. But the attitude towards TBW amongst employees generally is another powerful issue under this heading. Reward systems that focus on competition between individual employees for bonuses undermine the introduction of TBW. Similarly, information systems that are characterised by secrecy rather than openness will impede the implementation of TBW. The training and education priorities of the organisation must also be geared towards developing the knowledge, skills and abilities required for TBW, including leadership skills and teamworking skills. The organisation will also need access to coaching expertise to support teams both during their development and when they experience difficulties in the course of their work (such as conflicts between team members), either provided by someone within the organisation or an outside consultant. Team structure Teams structure refers to the composition of the teams, i.e. who will be the team members. This is not simply a matter of the skills required to perform the task, but also raises questions of variety in functional background and balance in demographic characteristics such as background culture, gender, age and even personality.

A key aspect of team structure is the nature of the task that the team is required to do. The goals should be clear, the task should be motivating and team members should have clear feedback on how effective their performance has been. It also refers to effective team leadership, as we have emphasised above, and the need to appoint team leaders who know how to lead teams and are not hierarchical, traditional supervisors.

Team processes When TBW is introduced, most organisations focus initially on team processes and send prospective team members on team-building workshops. Though the motive behind this is valid, i.e. to build cohesion, the first step in building effective teams is to ensure that team members: make sufficient effort (that they are motivated to perform the task).

have adequate knowledge and skill within the team both to perform the task and to work in a team.

have developed appropriate ways of performing their task, i.e. task performance strategies.

As organisations implement TBW, there are major pitfalls in each of these three key areas that must be avoided, any of which could considerably delay or impede the process. the creation of teams throughout the organisation, regardless of the need or the nature of the tasks.

setting up teams but continuing only to appraise, reward and manage individuals.

creating teams but neglecting to train people to function effectively within and across teams.

introducing TBW while leaving teams without expert assistance when problems such as major conflicts arise.

creating well functioning teams but ignoring the vital need to ensure these teams communicate with each other, integrate their work and otherwise liaise effectively.

failing to negotiate with the teams clear and challenging team-level objectives.

giving the teams challenging objectives but not the training, skills and resources to meet those objectives.

Conclusion

The following quotation illustrates just how fundamental team working is to our species and we include this to remind the reader of the importance of groups and teams to human societies throughout their development. There is little new about teamwork.

"He makes tools (and does so within more than one technical tradition), builds shelters, takes over natural refuges by exploiting fire, and sallies out of them to hunt and gather his food. He does this in groups with a

discipline that can sustain complicated operations; he therefore has some ability to exchange ideas by speech. The basic biological units of his

hunting groups probably prefigure the nuclear family of man, being founded on the institutions of the home base and a sexual differentiation of activity. There may even be some complexity of social organization in so far as firebearers and gatherers or old creatures whose memories made them the data banks of their 'societies' could be supported by the labour of others. There has to be some social organization to permit the sharing of cooperatively obtained food, too. There has to be some social organization to permit the sharing of co-operatively obtained food, too. There is nothing to be usefully added to an account such as this by pretending to say where exactly can be found a prehistoricial point or dividing line at which such things had come to be, but subsequent human history is unimaginable without them." [Extract from J.M. Roberts (1995), The History of the World, page 18].

The activity of a group of people working co-operatively to achieve shared goals via differentiation of roles and using elaborate systems of communication is basic to our species. The current enthusiasm for team working in and in health care reflects a deeper, perhaps unconscious, recognition that this way of working offers the promise of greater progress than can be achieved through individual endeavour or through mechanistic approaches to work. That is what this report has demonstrated in

relation to health care teams to and quality and innovation in patient care.

Appendix I Measures Used

Primary Health Care Team Questionnaire Section 1 Team working. This contained seven measures of team working. Five of these form the Team Climate Inventory (Anderson and West, 19xx): participation, a 15 item scale covering information sharing ( = ); innovation, an eight item scale covering support from new ideas ( = ); team objectives, covering clarity and relevance of objectives ( = ); task style, covering the monitoring and appraising of work in the team ( =). Three other measurers were included: reflexivity, covering the reviewing process in the team ( = ) and XXX was measured using XX (xx 19xx) ( = ); and team innovation measured using (West?) ( = ). Respondents were also asked to list the major changes introduced by the team in the previous 12 months. Section 2 included measures of team effectiveness adapted from Poulton and West (199x). This includes three dimensions: team working ( = ); , patient orientation ( = ); and organisational efficiency ( = ).

Section 3 included a measure of psychological stress, the GHQ-12 (Goldberg, 1991) ( = .88 ).

Section 4 included questions eliciting biographical and team information (e.g. age, gender, ethnic origin, job title, employer, team composition, team leader).

Initial construction of the effectiveness measure Effectiveness criteria were generated using an iterative process within the constituency model approach (Connally et al., 1980). After consultation with the local Health Authority and Community Mental Health Trust, the Department of Health, and local community mental health teams, representatives of the range of stakeholders in the provision of mental health care were invited to a one-day workshop. A total of 13

interest groups were approached: users, carers, advocacy agencies, mental health charities, consultant psychiatrists, community mental health nurses, occupational therapists, psychologists, social workers, managers, policy makers, researchers in mental health, and general practitioners. All groups were represented by the 50 people attending the workshop.

The aim of the workshop was to provide the basis for an agreed set of definitions of effectiveness in CMHTs. Stakeholders sharing a perspective were grouped together, so that consensus could be achieved more easily within each working group. Participants were asked to generate a set of criteria which they agreed would measure CMHT effectiveness, with the proviso that any criterion must be supported by concrete examples of how good practice could be audited. reconsidered their criteria in order to prioritise aspects of practice. Groups then

Output from the workshop was analysed and categorised by the research team. Duplication and ambiguity were removed. The 76 remaining criteria, grouped into the three broad categories of user and carer issues, team development and viability, and organisational issues, were re-circulated to all workshop participants. They were invited to comment on wording and clarity, to suggest modifications, indicate significant omissions, and approve priority ratings. Items rated as less important by a majority of respondents were removed, and any items rated down in the final consultation were respositioned. Remaining ambiguities and duplication were

removed, together with items already covered in other sections of the proposed survey questionnaire.

The reduced set of CMHT effectiveness criteria, together with supporting measures of good practice, were piloted amongst local community mental health nurses, a CMHT, psychologists and other mental health professionals. Practitioners agreed that they captured the complexity of the work and the diversity of environments in which CMHTs operate. The final set of 27 criteria was incorporated into the main survey questionnaire, along with the Team Climate Inventory (Anderson & West, 1994) and the 12-item General Health Questionnaire (Goldberg, 1970.)

Each statement was clarified by additional concrete examples of elements of practice which individuals could use to aid their rating. A 5-point Likert-type scale was used to rate how effective the team was on each criterion. Thus, to illustrate, the content of the first criterion was

Accessibility of the service to users and carers has been identified as a measure of CMHT effectiveness. [For example: identification and contactability of a key worker; clear referral procedures; time taken to respond to users and or carers; a clear point of access.]
Not at all To a great extent

Overall, to what extent does your CMHT make services accissible to users and carers?

Appendix I I Knowing the Way: Effectiveness in Primary Health Care

A description of national workshops aimed at defining effectiveness criteria for primary health care

Introduction The World Health Organisation define primary health as: ..essential care based on practical, scientifically sound and acceptable methods and tehcnology made universally accessible to individuals and families in the community through their full participation and at a cost that the community and the country can afford to maintain at every stage of their development in the spirit of self reliance and self determination. It forms an integral part of the countries health care system, of which it is a central function and main focus, and of overall social and economic development of the country. It is the first level of contact of individuals, the family and community with the national health system, bringing health care as close as possible to where people live and work. World Health Organization, 1978.

In order to enable primary health care professionals and the populations they serve to develop the health and stress of those populations, it is important that clear objectives for primary health care are established. Moreover, what constitutes effectiveness in primary health care also has to be determined by each primary health care team or organization. Indicators of effectiveness, once established, provide professionals with clear guidelines over how best to allocate resources in order to achieve effective primary health care. This consultative document represents the endeavours of 63 professionals working in the area of primary care who spent four days in workshops designed to help achieve this overall aim. 10

The vision of this exercise is to promote health care for the population nationally, by providing primary health care professionals and others concerned with the health and stress of the population with clear indicators of effectiveness. These can be used as a basis for discussion and development in primary health care teams across the country, taking into account their local circumstances, philosophies of primary care and the needs and views of their local populations.

Effectiveness in Primary Health Care In an important analysis of the evaluation of health services effectiveness, St Leger, Schneiden & Walsworth-Bell (1992) observe that Surprisingly, routine data [sources relating to primary care] are relatively sparse, especially when one considers that the majority of contacts that the general public have with the health service are with general practice. (p.41). Indeed, it is striking how little research has focused on developing theoretically grounded or practically useful indices of effectiveness of primary health care services. In a rare exception, Pearson & Spencer (1995) employed a two-stage Delphi questionnaire to determine agreed indicators of effective teamwork in primary care. Using responses from 137 people involved in primary care teamwork - primarily from FHSAs - they rated the importance of twenty indicators. Four emerged as particularly significant:

10

Within primary health care, terminology is a source of concern and conflict. For example, some prefer the term primary care. There are sensitivities around the terms general practice and primary health care being used interchangeably; and a distinction between medical and nursing care is also sensitive. Those who attend for treatment or advice are called patients by some and clients by others. We have preferred the terms primary health care and clients in this document.

Agreed aims, goals and objectives Effective communication Patients receiving the best possible care Individual roles defined and understood

What is effectiveness?
How can the effectiveness of primary health care teams therefore be judged? At the simplest level, effectiveness might be viewed as the influence the primary health care team has in improving health and ameliorating ill health within the practice population. Such a criterion of effectiveness begs a number of questions, however. Examples include:

Is the population in an area of social deprivation? Does the team have unusually good resources, in terms of number of staff or technical equipment? Will the effects be long-term?

In the literature on organizational effectiveness, a distinction is made between efficiency (doing things right) and effectiveness (doing the right things) (Sundstrom, DeMeuse & Futrell, 1990). Efficiency may be defined as the output for a given input and how a team compares with other similar teams in this regard. Effectiveness can also be seen as the teams capacity to perform, adapt, maintain itself and grow (where growth may refer to size, innovation or skill development) (Goodman, 1986).

Clearly, teams can be seen as more or less effective depending upon the criteria adopted. Consequently, the assessment of team effectiveness has come to be seen as much a political as an empirical process. Below we consider (briefly three examples of approaches to managing this problem.

The constituency approach

Recognition that effectiveness is a political concept has led to the use of the constituency approach (Connally et al, 1980; Bedeian, 1986) which seeks to incorporate all significant views in the judgement of team effectiveness. Each of the major constituents is identified (e.g. clients, carers, staff health authorities, professional organizations) and the effectiveness criteria they would use are adopted as indicators. Effectiveness is then measured using multiple indicators rather than an aggregate, since, in many cases, effectiveness in one area will necessarily imply ineffectiveness in another (consumer satisfaction may not always coincide with quality of care if consumers require prescriptions for drugs, the use of which is not in their best interest).

From this conceptual, background Poulton & West (1994) developed a set of 23 effectiveness criteria for primary health care teams using a focus group methodology involving multiple stakeholders in primary care. In a study involving more than 500 practice nurses (Poulton, 1995) the criteria were then factor analysed revealing four underlying factors: good teamwork, task excellence, organizational effectiveness and patient-centred care. More recent analyses with larger samples suggest only three: quality of health care, organizational efficiency, and teamworking. Poulton and West (1997) employed these outcome measures in a six month longitudinal study of the impact of teamworking on effectiveness. The research demonstrated that clarity of and commitment to objectives was the most important and statistically significant predictor of effectiveness in all four areas.

Multidisciplinary audit
Another important approach to measuring effectiveness is multidisciplinary audit. A major centre for research and advice is the Eli Lilley National Clinical Audit Centre (Hearnshaw, Baker & Robinson, 1994; Baker, et al 1995). In a study conducted by this Centre of three practices, those supported in the development of multidisciplinary audit showed a significant improvement in specific areas of functioning. The procedure involves teams identifying particular problems (in a diagnostic area or an area of team functioning) and a priority is identified. The team then sets standards in the specific area, observes current practice and achievements, and compares these with the standards. Discrepancies prompt changes in practice and the results are then regularly reviewed. Such an approach clearly enables teams to assess and improve performance in specific areas, although the approach is somewhat atomistic and relies on the effective identification of priorities. An excellent overview of this approach is provided by Crombie, Davies, Abraham & Florey (1993).

ProMES
In the broader organizational literatures on team effectiveness, a widely adopted approach is the Productivity Measurement and Enhancement System (ProMES) based on research by Naylor, Pritchard & Ilgen (1980). Effectiveness criteria are established in group discussions with team members and supervisors. The variables are then psychologically scaled to a common effectiveness scale. Based on group consensus about expected levels of effectiveness, which are given a zero value, maximum effectiveness levels (set at +100), and minimum levels (-100) are set. Each variable is also weighted in terms of its perceived contribution to the overall effectiveness of the team or organization. The system is then used to set objectives, develop indicators monitor and improve performance and give feedback to the team (Pritchard, 1990). This is promising for primary health care, because of the sophistication of the approach, its theoretical robustness and practical utility in complex contexts. It was therefore used as a basis for developing effectiveness measures by the Health care Team Effectiveness project.

Overall, however, it is clear that conceptual and empirical development in evaluating the effectiveness of primary health care is urgently required, if cost, clinical and community value are to be convincingly

demonstrated. In order to take this forward a series of four workshops on primary health care team effectiveness was sponsored by the Institute of Work Psychology. The methods are described below. Prior to detailing these we consider the need for a manageable set of objectives or effectiveness indicators; and the need for a theoretical framework within which to locate any set of dimension.

(i) The need for relative simplicity It is apparent from any analysis of research in the domain of effectiveness that the development of indicators and effectiveness dimensions for primary health care is a complex task (West, 1996). A major problem is that many measures of effectiveness and many indicators can be developed. Trying to use this large number of measures within an organization as an effective means of targeting resources simply becomes overwhelming for the practitioners concerned. The sheer multiplicity of potential indicators is cognitively too complex for people within the organization to cope with the task effectively. Indeed, some research suggests we can only cope with 7 plus or minus 2 categories, whereas other organizational researchers specifies to more than 8 to 12 key dimensions. Consequently, we sought a parsimonious theoretical model which would provide guidance for researchers and practitioners.

(ii) The Competing Values Model The most useful model appears to be the Competing Values Model. This model incorporates two fundamental dimensions; flexibility of the organization versus control within the organization; external orientation versus internal orientation (see figure 2).

F ig u r e A .1

C o m p e tin g V a lu e s M o d e l
F le x ib ility
H u m a n R e la tio n s M o d e l
T raining S c a n n in g o f t h e Team d e v e lo p m e n t e n v ironm ent R e s o u r c e a c q u isition & c o ordination I n n o v a tio n C o lla b o r a tio n w ith o t h e r organisations

O p e n S y s te m s M o d e l

Learning organisation Team m em ber m ental h e a lth C o m m itm e n t & satisfaction

In te rn a l O rie n ta tio n
E fficient control of resources E m phasis on perform ance C linical audit C u s t o m e r s a t i s f a c t i o n R e v ie w in g & e v a lu a t i n g effectiveness Budget m anagem ent M eeting targets

E x te rn a l O rie n ta tio n

In t e r n a l P r o c e s s M o d e l C o n trol

R a tio n a l G o a l M o d e l

Combining these dimensions identifies 4 domains of effectiveness: human relations, open systems, rational goal, internal process

Human relations model

The primary emphasis is on norms and values associated with belonging, trust, respect, skill development, growth and stress. Motivational factors are attachment, cohesiveness and organization membership. Areas of effectiveness include: development of skills team development learning organization skills team member mental health commitment and satisfaction

Open systems model

The primary emphasis is on change and innovation and appropriate scanning of the environment. Norms and values are associated with knowledge of the environment, resource acquisition, innovation and adaptation. Motivating factors are: development of services resource acquisition and co-ordination innovation collaboration with other organizations

Rational goal model

The primary emphasis in this model is on the pursuit and attainment of well-defined objectives. Norms and values are associated with good performance, goal-fulfilment and achievement. Motivators are: successful achievement in pre-determined areas. Effectiveness dimensions include: meeting production targets ensuring high quality high client or customer satisfaction

Internal process model The emphasis here is on stability, internal organization and adherence to rules and protocols, where norms and values are associated with efficiency, co-ordination and uniformity. Motivating factors are needs for order, rules, regulations and efficiency. Effectiveness dimensions include: efficient control of resources reviewing and evaluating effectiveness good budget management

The model of competing values stresses how the allocation of resources to any one area, for example the open systems model, is likely to lead to a restriction of resources in the opposite domain (internal process). Similarly a focus on external control (rational goals) may well lead to a neglect of the area of human relations (internal flexibility). Effectiveness in organizations, is likely to be maximised when emphasis is placed equally in each of the four domains of effectiveness. Using this framework as a guide we embarked on a series of 4 workshops to clarify our thinking about effectiveness in primary care.

The Workshops

Method
The methodology used to develop effectiveness measures for primary health care was developed taking account of the following: the importance of including the complexity and diversity in primary health

care and of taking into consideration the wide range of views and perspectives held by the professional groups who contribute to meeting patient needs; the need to develop a method which would most costeffectively use the time of primary health care professionals; and the importance of developing measures which were generalisable across primary health care.

The measures were therefore developed in two stages:

i) Indentifying objectives for primary health care and developing indicators for these objectives, was carried out in four workshops with domain relevant experts from primary health care, based on an approach developed by Connolly et al, 1990. Such an approach enabled the views of a range of professionals to be taken into account. Working intensively with an expert group enabled considerable knowledge and insight to be gained in a short space of time. In addition, these professionals, who had a background in primary care but were mainly working in an advisory, policy or research role, provided a broader, more generalisable perspective on effectiveness in primary health care.

ii) The measures developed were used by primary health care teams teams and feedback provided on effectiveness.

Workshops
Objectives There were three principle overall objectives for the workshops: to identify the important issues relating to developing effectiveness measures for primary health care, to develop a set of effectiveness measures acceptable to all perspectives in primary health care, for participation in the workshops to be a valuable experience for the participants.

Participants An initial stakeholder analysis identified 13 stakeholders in primary health care. Advise was sought from contacts in primary health care about key experts who could represent the views of each stakeholder group, and about whether the initial list of stakeholder was sufficiently comprehensive. The experts suggested by the contacts were sent information about the research programme, invited to attend the four workshops, and asked to suggest additional or alternative key experts who could also make a contribution. In addition, representatives from primary health care teams were invited, that is, professionals who were currently engaged in clinical practice. The majority of those contacted were keen to attend the workshops, and were able to commit themselves to attending two or three. X were able to attend all four. Each workshop was planned so that the participants covered the main stakeholder views.

Workshop process

A focus group methodology was used. The delegates were divided into three working groups. These were designed so that (a) a range of stakeholder views were represented, and (b) one or two or group members had attended most or all of the workshops and so could share with new members the learning and experience from previous workshops. Each group worked with a facilitator, trained in ProMES, and a notetaker who recorded the group discussion and the decisions made.

Workshop 1 Objective: to develop objectives for primary health care. After an initial introduction to the Health Care Team Effectiveness project and a presentation on ProMES, the delegates were presented with a set of objectives for primary health care developed by the researchers. Each group worked on (a) refining the objectives, (b) critically evaluating the objectives in relation to the criteria for objectives (see appendix xx). The outputs from each group were presented at the end of the workshop in a plenary session. After the workshop the outputs were discussed with members of four primary health care teams (who endorsed their relevance and value), combined into a single list and then circulated to delegates.

Workshops 2 and 3 Objective: to develop indicators for the objectives for primary health care. Both workshops started with a presentation on team working in primary health care and issues relating to the development of effectiveness measures. Delegates were presented with the final version of the objectives for primary health care. Each group worked on (a) developing indicators for an objective, (b) critically evaluating the indicators in relation to the criteria for indicators (see appendix xx). In Workshop 2 the working groups selected the objective to discuss. The objectives - Quality of Care and Client Satisfaction were selected. In Workshop 3 groups were assigned an objective so that each was discussed at least once. The objectives - Effective Management of Resources, Development and Satisfaction of Primary Health Care Team Members and Quality of Care, were discussed. In the third workshop the output from each working group was given to another group in the afternoon session for discussion and refinement. The output from each group were presented at a plenary session at the end of both workshops.

After Workshop 2 and 3 the outputs were amended and the circulated to delegates. In addition, after the third workshop the outputs from all three workshops was written-up in the document Knowing the Way: Effectiveness in Primary Health Care and circulated to delegates.

Workshop 4 In the fourth Workshop, those attending critically analysed the objectives and indicators developed, and considered how they could be applied in practice by PHC teams and others. A review of the data

derived from 100 teams, examining their definitions of effectiveness was also presented. The focus groups commented on the next steps in taking forward the work completed to date.

1.

Objectives were identified in the first of the four workshops and after an additional three workshops with somewhat differing attendees who worked with them, they remained unchanged. (The point is they have been tested and found acceptable by lots of different people.)

2.

In addition, they were shown to a number of primary health care teams. These teams found the objectives useful and accurate.

3.

The themes in these objectives are similar in principle to the themes of objectives that other types of professional organizations and in other settings have developed. Thus, there is some consensual validation.

Bearing in mind the problems of deluging primary health care teams/organizations with long and unmanageable lists of objectives and indicators of effectiveness, we endeavoured to produce a core list of key objectives

Below we describe each and suggest possible indicators.

Effectiveness measures are a guide not stone tablets

The key objectives identified and the indicators developed for each are a synthasis of the outputs from the four workshops. Some of the contributions made by the working groups have been modified by the researchers, and the Competing Values Model used to provide a conceptual structure. The majority of objectives and indicators described below, however, were suggested and critically reviewed by the Primary Health Care professionals who attended the workshops. The objectives and indicators are not definitive, nor are they comprehensive and applicable to every primary health care team. They are meant to be used as an aid for teams which wish to measure and enhance their effectiveness. In relation to each objective, the focus groups developed a set of indicators by which progress towards objectives could be measured. Again, these indicators are meant as examples. If they fit and make sense to individual primary health care teams, fine; but they may well not fit. Primary health care teams have different missions and the measurement must be tailored to that mission. Moreover, the focus groups produced many more possible objectives and measures of them are described below. These can be used as a resource for teams wishing to explore areas beyond those we have designated as likely to be core.

Objectives and Indicators for Primary Health

The core key objectives developed in the workshops are showh in Figure A2, mapped in to the competing value model.

F ig u r e A . 1

C o m p e tin g V a l u e s M o d e l i n P rim a r y H e a lth F le x ib ility C a re

H u m a n R e latio n s M o d e l
G o o d T e a m w o r k ing C o n tinuing professional development H ig h t e a m m e m b e r c o m m itm e n t & satisfaction

Open Systems Model

Accurate identification of h e a lth n e e d s Responsiveness to clients a n d c o m m u n ity E ffective collaboration with other organisations

Internal O rientation
E fficient use of resources H e a lth c a r e r e v i e w i n g & im p r o v i n g e f f e c t i v e n e s s C lient satisfaction

E x ternal O rientation

* Im p r o v i n g

health

*High quality of health care

Internal Process Model C o n trol

R a tio n a l G o a l M o d e l

* In d i c a t o r s m a y d e p e n d o n t h e h e a l t h c a r e p h i l o s o p h y o f t h e p r i m a r y h e a l t h
c a r e t e a m s e .g . h o l i s t i c , p r e v e n t i v e , b i o m e d i c a l

The indicators developed in the workshops for each of the objectives are listed below.

Improving health High quality of health care Improving client satisfaction Efficient use of resources Reviewing and improving health care effectiveness Good teamworking Continuing professional development High team member commitment, stress and satisfaction Accurate identification of health needs Responsiveness to clients and community Effective collaboration with other relevant organizations

Rational Goal Objectives Improving health Example Indicators Improvement in the health of the practice poulation including reductions in e.g. coronary heart disease, smoking, mental health problems. Percentage of clients improving at the expected rate after treatment. Effectiveness of preventive practice in reducing specific treatment requirements. High quality of health care Effective knowledge of and management of chronic diseases (e.g. diabetes, epilepsy, asthma) measured by conformance with evidencebased good practice. Effective health education and preventive health care programmes. The PHCT holds regular meeting to review a sample of cases. This review would include the appropriateness of who saw the client, procedures, and outcomes. Percentage of cases managed entirely appropriately, based on all staffs views. Improving client satisfaction Measures of client complaints and adequacy of procedures for complaints. Questionnaire or telephone surveys using standardised measures*. Measures of waiting times, satisfaction with consultations, appropriateness of appointments. * See Appendix III for examples

Internal Process Objectives Efficient use of resources Example Indicators Percent client contact time as a percentage of total time (there is an optimal level between extremes). Assesment of DNAs. Number and effectiveness of initiatives developed to help team members use time better. Review and evaluate budget allocation and improvements in resource utilisation. Reviewing and improving health care effectiveness Review and use of evidence-based treatment protocols (all staff). Planned clinical audit (all staff). Intra-team referral practices regularly reviewed and adapted (all staff).

Human Relations Objectives Good teamworking Example Indicators Clear, shared objectives (partly related to health needs analysis) set annually by the team. Regular (at least monthly) meetings to review team objectives, strategies, processes and procedures to coordinate sub-groups and whole team. Positive team/organizational climate assessed annually. Continuing professional development Clear and specific written annual training and development plans agreed for each staff member (percentage of staff covered; percentage of development plan items completed.) Research and development budget and plans agreed by team annually. Access for all team members to training/development resources. High team member commitment, stress and satisfaction Annual review of staff commitment, stress and satisfaction using standardized measures*. Mechanisms to deal with and review staff dissatisfaction, conflicts and complaints. Low absenteeism and staff turnover. *See Appendix III for examples

Open Systems Objectives Accurate identification of population health needs Example Indicators Collection of practice level data (demographics, disease patters, socioeconomic patterns, activity levels); and local, regional and national data. Involvement of clients, community groups and other relevant

organizations in health needs analysis. Data used to inform daily planning, and longer term strategy and direction; setting annual objectives; identifying gaps in provision and skill mix. Responsiveness to clients and community Involvement of clients and community in team/organization decisions concerning team objectives, strategies and processes. Frequency, quality and usefulness of contacts between team members and representatives of community stakeholder groups. Extent of planning within team/organization to seek feedback from clients and community stakeholders/opinion leaders/groups. Effective collaboration with other relevant organizations Appropriate admissions to hospital (and referral rate) Number of effective and appropriate contacts with agencies (e.g. palliative care, social services, education). High ratings of team/organizations on salient dimensions*. *See Appendix III for example

Application of the effectiveness measures

General Principles There are three levels of application of these effectiveness measures, ranging from simple through to comprehensive.

Simple. The simplest way of using the effectiveness measures is to use them as a basis for group discussions in the primary health care team; for members of the team to consider the areas of effectiveness described and how they can make use of the measures in facilitating of the effectiveness of the primary health care team. They may also consider what other measures they may wish to add, given their local circumstances and which of the measures are not applicable. In other words, the simple approach is to use the effectiveness measures as a basis for ongoing discussions about monitoring and improving the effectiveness of the primary health care team.

Moderate. The primary health care team can use the effectiveness dimensions and indicators to develop measures of effectiveness within their primary health care organization in relation to each effectiveness measure (as appropriate). The team can develop measures and make action plans in terms of how they can improve their performance in this area. Again they may wish to consider which of the measures are applicable in their organization and which are not, and what measures which are relevant to them are missing from document.

Comprehensive. This application involves the use of full productivity measurement and enhancement system, which has the following steps: (a) a design team, which includes representatives from all groups of staff in the primary health care team, is formed; (b) in a series of meetings the design team works with a facilitator identifying objectives and indicators for these, where necessary consulting with other members of the team; (c) the design team develops

contingencies for each indicator, that is, determines the relative contribution that improvements on a indicator will make to overall effectiveness; (d) the team uses the indicators and receives feedback on performance.

Quality of Health Care

Quality of Health Care The PHCT would have a monthly (or more frequent) staff meeting where a sample of cases was reviewed. This review would include the appropriateness of who saw the client, what procedures were done, and whether that client was handled appropriately in all aspect. The measure would be the percentage of cases which were considered as being handled appropriately. This would also be the basis for discussion of what improvements need to be made for those specific clients and for clients in general.

Some PHCTs will feel a health needs analysis is valuable but do not know how to do one or how to use it. For such a PHCT, the task of developing such an analysis could be broken down into definable steps. E.g. get information on how to do such an analysis, decide on a plan for doing the analysis for that particular PHCT, gather the information, put the information together into a form that the PHCT can use to make decisions. Each of these steps would be given a time for completion. The indicator would be the percentage of the analysis completed compared to the anticipated time for completion.

Survey on client perceptions of health improvement after treatment. For example, each client is given a questionnaire or a sample of clients are called by phone and asked about improvements. Measure is the percentage of clients improving. For the various specific targets given by agencies outside the PHCT such as immunisation rates, develop a scoring system whereby each level of meeting the objective gets a certain number of points. E.g. if the target immunisation rate was 80%, actually doing 80% would give 100 points, 60% immunised would be 20 points, 70% 80 points, 90% 130 points, etc. The number of points would be based in the importance of that target. The index would be the percentage of actual points earned compared to the maximum possible points received if all targets were met. (Note, this assumes there are lots of such targets. If this is not true, a composite measure such as this is probably not necessary.

The percentage of required reports completed on time. The number of required reports returned by agencies requesting corrections or additional information. (This would be an index of the quality of the reports.)

Client Satisfaction Establish a formal procedure where clients can make complaints including a process for following up on these complaints. Measure is the number of such complaints which were not concluded to the clients satisfaction within one week.

Effective Management of Resources Number of new initiatives developed that are designed to help team members use their time better. These initiatives should also be reviewed on a regular basis to ensure they are still effective. Percent client related time as a percentage of total time. This measure gets at how much time is devoted to clients. It does not measure how well that time is being spent. Other indicators are needed to address this issue. (RDP: Note that this indicator is one where there is probably an optimal level between the extremes. To little time with clients may suggest too much administration time. Too much time with clients may suggest too little administration time.) Percentage of staff turnover over time. High staff turnover leads to inefficient resource utilisation because it takes time to teach procedures to new staff and work is lost as a departing staff member leaves. This measure would also be an indicator for the satisfaction of team members. Percentage of appointments which are unfilled or where the client did not come.

Development and Satisfaction of Primary Health Group/Team Members Training and development. A list of training and development experiences for each person on the team would be developed each year. For example, attendance at a certain type of conference, training on a piece of office equipment, learning a new procedure, etc. This list would be the development plan for that person for that year. There would be two measures for training and development. The first would be the percentage of team members who had the written plan. The second measure would be the percentage of the development plan items actually completed. Who are reviewed, given feedback, and have a formal, jointly developed action plan for making improvements. Satisfaction. Measure overall satisfaction on a monthly or bi-weekly basis with a very brief questionnaire that would take no more than 2 minutes to compete. Measure would be the percentage of staff indicating Satisfied or Very Satisfied with their jobs. Staff turnover is also a satisfaction measure. Note this measure under Effective Management of Resources.

Figure A.2 List of participants NAME John Horder Debbie Mellor JOB TITLE President CAIPE Section Head of Workforce Non-Medical Planning Thelma Sackman Kate Andrews Rosemary Field Nicki Meade Nursing Officer Clinical Research Fellow To Be Advised Research Associate NHS Executive Dept. General Practice To Be Advised The National Primary Care Research & Development NHS Executive PLACE OF EMPLOYMENT

Centre Steven Campbell Research Associate The National Primary Care Research & Development Centre Brenda Leese * Research Fellow The National Primary Care Research & Development Centre Bonnie Sibbald Research Associate The National Primary Care Research & Development Centre Ann Richards Research Fellow Psychological Therapies Research Centre Malcolm McCoubrie Senior Lecturer in Community Based Medical Education Standards - Medical Director Sheelagh Richards * Jane Cannon* Sue Jenkins-Clarke Peter Bundred Occupational Therapy Officer Practice Nurse Research Fellow Senior Lecturer in Primary Care Judy Mead * Physiotherapist Chartered Society of Physiotherapists Richard Brown * Alan Chapman To Be Advised Management Education & Development Manager Primary Care Lance Gardner Terry Brugha * Professional Officer Senior Lecturer & Honnary Consultant Psychiatrist Rosamund Bryar Professor of Community Healthcare Nursing Practice Stephen Rogers Senior Lecturer in Primary Care Joan Lole Director of Nursing & Primary Care Paul Thomas Senior Lecturer Dept. Of General Practice Ruth Hudson Education Officer Imperial College School of Medicine @ St.Marys Community Practitioners & Health Visitors Association Mancunian Community Health University College London University of Hull The Queens Nursing Institute University of Leicester To Be Advised East Norfolk Health Authority London Larwood Surgery University of York University of Liverpool Wandsworth Community Health

Jacky Hayden Christiana Johnson

Dean of Postgraduate Medicine Health Promotion Officer

University of Manchester Princess Royal Community Health Centre

Beverley Haynes

Senior Health Promotion Specialist

Princess Royal Community Health Centre

Peggy Newton Jeanette Naish

Lecturer in Psychology Senior Lecturer in Primary Care

Dept. Of General Practice Dept. General Practice & Primary Care

Stuart Mee * Kay Robinson Brian McAvoy Susan Lonsdale Sandra Dodgson Frances Fogg Wendy Whyte

Practice Manager Primary Healthcare Facilitator Professor of Primary Health Care Senior Principal Research Officer Senior Development Manager Primary Healthcare Facilitator Regional Community Nursing Team Leader

The Crookes Practice South Humber Health Authority Dept. Of Primary Care Dept. Of Health N H S Development Unit North Notts Health Authority British Forces Overseas

Mike Sharpe

Regional General Manager of Medical Services

British Forces Overseas

Ron Pollock

Assistant Director Support & Development/Finance

Wakefield Health Authority

Mike Vaughan Sasha Wishard Marion Duffy Chris Simmonds Jane Solomon Catherine Booth Ann Netton Gwen Wilson * * * *

Total Purchasing Project Manager Research Facilitator Education Facilitator Practice Manager Locality Management General Practitioner Assistant Director of PSSRU Development Manager Community Nursing

Wakefield Health Authority Tayside Centre for General Practice Tayside Centre for General Practice Medical Centre Doncaster Nottingham Health Authority G.P. Unit University of Kent Sheffield Community Health

Appendix III Effectiveness Measures Developed for Primary Health Care Teams

Core Objectives for Primary Health Care teams Promote, maintain and improve health Enable personal and community responsibility for individual health Efficient use of resources Continuous personal and professional development High team member commitment, stress and satisfaction Responsiveness to clients and community Collaboration and partnership with other relevant organisations Provide high quality health care Accurate identification of individual and population health care needs Review and improve the effectiveness of health care provision Manage illness, injury and disease taking account of agreed standards and evidence based practice Enable patients/clients to make informed decisions about their own health. Proactively encourage positive health behaviour Implementation of health education and preventative care programmes Human resources skills, knowledge, expertise, time Physical resources budgets, equipment, premises Individual annual training plans which take account of the plans of the PHCT Equal access to training/development resources Team working Mechanisms for reviewing and acting upon staff dissatisfactions, conflicts and complaints Gather information and feedback from clients/community stakeholders/opinion leaders

Objective: Promote, maintain and improve health Techniques for reviewing whether services meet client needs A. B. What are the main aims of this service? What does the team (in collaboration or in addition to other agencies) currently do to meet a particular health/health promotion need? List all the provisions currently available in the team (and from other

agencies, if relevant). C. How do you know whether these provisions meet these health/health promotion needs? List evidence that can be used to determine this. D. Which aspects of this evidence suggest that you are meeting this health/health promotion need? Which aspects of this evidence suggest that you are not meeting this health/health promotion need? What provision would the team ideally like to have in place to meet this health/health promotion need?

E.

F.

Next steps: Use the evidence discussed in C, D and E to develop measures to enable the team to evaluate more systematically whether they are meeting clients needs.

Objective: Promote, maintain and improve health Measure 1 - Review of quality in case management

Measure =Percentage of cases judged to be managed appropriately on the most relevant quality dimensions. Steps to clarifying the measure: Determine the types of cases to be reviewed (specific condition, e.g. diabetes/asthma, or a specific age group or type of patient population). Decide which of the quality dimensions are most relevant to the cases being reviewed. Decide what is an acceptable quality level on each dimension. decide what is an acceptable % of cases to be judged as having been managed appropriately.

Using the measure: Rate each of the selected cases on the quality dimensions and give a total score. Note the dimensions where quality is above and below the acceptable level. Calculate % of cases which fall above and below the acceptable level of cases being managed appropriately. The review will result in two types of information dimension of quality for individual cases which fall below the acceptable standard. % of cases overall which are managed appropriately.

N.B. For this measure need to develop an instrument for rating cases on each dimension which suggests evidence that can be used to make judgements, emphasise the importance of standardising ratings across cases and gives guidance in how to complete the instrument.

Dimensions of Quality

Effectiveness:

Is the treatment given the best available in a technical sense, according to those best equipped to judge? What is their evidence? What is the overall

result of

the treatment?

Acceptability:

How humanely and considerately is this treatment/service delivered? What does the patient think of it? What

would/does feel if it were like? Are privacy

an observant third party think of it (How would I my nearest and dearest?) What is the setting and confidentiality safeguarded?

Efficiency:

Is the output maximised for a given input or (conversely) is the input minimised for a given level of output? How does

the unit the same

cost compare with the unit cost elsewhere for treatment/service?

Access:

Can people get this treatment/service when they need it? Are there any identifiable barriers to service - for example,

distance, or

inability to pay, waiting lists, and waiting times straightforward breakdowns in supply?

Equity:

Is this patient or group of patients being fairly treated relative to others? Are there any identifiable failings in equity - for example, are some people being dealt with less

favourably or others?

less appropriately in their own eyes than

Relevance: could the

Is the overall pattern and balance of services the best that be achieved, taking account of the needs and wants of population as a whole?

Dimensions of Quality Effectiveness To what extent...... To a very little extent Is the treatment/service being given technically the best possible? Does the treatment/service being given conform to agreed protocols/standards? Is the current outcome from the treatment/service as would have been expected, given the patients condition at the start? Acceptability To what extent...... To a very little extent Is the patients privacy safeguarded? Is the patients confidentiality safeguarded? Is the patient treated with consideration and respect? Efficiency To what extent...... To a very little extent Are the inputs to the treatment/service (e.g. staff time, medication) minimised for a given level of output? Is the unit cost the same as for this treatment/service delivered elsewhere? To some extent To a very great extent To some extent To a very great extent To some extent To a very great extent

Access To what extent...... To a very little extent Can patients access the treatment/service when they need it? Do any of the following pose a barrier to accessing the treatment/service?: To some extent To a very great extent

Location Distance Time of availability Inability to pay

Waiting lists Waiting times Lack of appointment times Lack of resources to supply treatment/service Equity To what extent...... To a very little extent Is this group of patients being fairly treated relative to others? Are the resources available for this treatment/service comparable to those available for others? To some extent To a very great extent

Relevance To what extent...... To a very little extent Are the resources used for this treatment/service appropriate in the context of the needs and wants of the practice population as a whole? To some extent To a very great extent

Objective: Promote, maintain and improve health. Measure 2 - Young Peoples health - Sexual Health

Measure = Percentage unwanted teenage pregnancies in a 6 month period Percentage of teenagers prescribed the morning after pill in a 6 month period Percentage of teenagers requesting pregnancy tests in a 6 month period Steps to clarifying the measure: Over a 3 month period monitor the number of unwanted teenage pregnancies and terminations, number of morning after pills prescribed, number of teenagers requesting pregnancy tests. This will establish a base line. Compare the numbers (or % of total number of teenager girls on the practice list) with the teenage pregnancies, use of morning after pill, teenagers requesting pregnancy tests in other PHCTs, and/or with regional figures. This enables the team to assess the extent to which they are meeting the sexual health needs of young people.

Decide what is an acceptable level of unwanted teenage pregnancies, morning after pill, requests for pregnancy tests.

Using the measure: Over a 6 month period log each: unwanted teenage pregnancy; request for morning after pill; and request for a pregnancy test. Note whether it is a small number of teenage girls who make the requests, or spread across a wide range of girls. Note whether there are any patterns (i.e. times of the week/month). After 6 months (or sooner if there are sufficient incidents of pregnancies/request for morning after pill/requests for pregnancy tests to form a judgement), collate the information collected.

Next steps: Compare the % for teenage pregnancies, requests for morning after pills and requests for pregnancy tests with (a) what were considered to be acceptable levels, and (b) with figures for other practices. On the basis of this determine whether the current provision to meet the sexual health needs of young women is (a) being met (how do the figures for the PHCT compare with the acceptable level? Are they better, worse, the same?), and (b) how the extent to which these needs are being met compares with other PHCTs/regional averages.

If the measures suggest that the sexual health needs of teenagers are not being met, introduce interventions to address this. The information about whether it is the same small number of young women requesting morning after pills/pregnancy tests will help to determine the type of interventions required.

Once interventions have been introduced, re-use the measure to determine whether the provision of services has improved.

Interventions: What type of follow-up is there when a teenager requests the morning after pill/a pregnancy test? Gather more information about why young people take risks. Implications for HIV/AIDS.

Objective: Promote, maintain and improve health. Measure 3 - Young Peoples Health - Alcohol and Drug Misuse

Measure =

Number of teenagers attending A & E after drug overdose in a 3 month period. Number of teenagers attending A & E after excessive alcohol in a 3 month period.

consumption

Steps to clarifying the measure: Over a month monitor the number of A & E slips which record that a teenager has attended A & E for drug or alcohol abuse. Compare these numbers with national/regional figures, and the number of other PHCTs. This will help establish the extent the team is meeting these health promotion needs of young people compared to other PHCTs. Decide what is an acceptable number of A & E attendances for drug and alcohol abuse among teenagers.

Using the measure: Over a 3 month period log each A & E attendance by a teenager for (a) drug abuse, (b) alcohol abuse. Note whether it is a small number of teenagers who attend A & E for drug and alcohol abuse, or if it is spread across a wide range of teenagers. Note whether there are any patterns (times of the week/month). After 3 months collate the information requested. Determine (a) number of incidents of drug and alcohol abuse at A & E, (b) number of teenagers who attend A & E once, number who attend regularly.

Next steps: Compare the number of A & E attendances for drug and alcohol abuse with (a) what the team judged to be an acceptable number and (b) with figures from other practices and regional/national figures. On the basis of this determine whether the current health promotion to raise awareness about the changes of drug and alcohol abuse are (a) being met (how do the recorded numbers compare with the agreed acceptable level? Are they better, worse, the same?) and (b) how the extent to which the health promotion need is being met compares with other PHCTs/regional figures. If the measures suggest that health promotion is not being effective, introduce interventions to assess this. The information about whether is it the same or different young people who misuse alcohol and drugs to determine the type of information required.

Measure 4 - Patient access to consulations with a GP Measure = The number of days that patients wait to see a GP of their choice

The PHC design team believed that an important part of providing quality care was to ensure continuity of patient care. The aim was to ensure that patients had access to the GP of their choice (the GP who ha most often provided health care in the past) by reducing the length of time they had to wait to see this GP. Steps to clarifying the measure
Monitor the length of time patients have to wait to see each GP over a 1 month period. If there are variations in the waiting time from week to week note the factors which might account for this (eg GP absences, increase in patient demand, services provided by other team members). Decide what is the target waiting time for each GP. Set this target taking account of the factors which increase and reduce the waiting time. This target might be the number of days a patient has to wait to see a GP of their choice, or it might be more realistic to set a target which specifies the maximum and minimum range, to allow for fluctuation which are outside the teams control. Using the measure Over a 3 month period monitor the length of time patients have to wait to see each GP in the team. If there are variations across weeks, months, or between GPs, note the factors which might account for these. Next steps Depending on the target set, calculate the average length of time that patients have to wait to see a GP or calculate the maximum and minimum lengths of time they have to wait. Compare this with the target set. If the target has been met, use the information gathered which explained fluctuations in the length of time patients had to wait to assess whether it might be possible to reduce waiting times further (ie, if increases in patient demands increased waiting times, is it possible to anticipate and plan for these increases?). If the target was not met, use the information gathered which explains fluctuations to assess what changes need to be made so that the target can be met. Also consider whether the target is realistic.

Measure 5 Patient access to a quality consulation with GPs Measure 1= Percentage of patients whose appointment with a GP is minutes duration in a 3 month period. Much of the discussion in the PHT design team concerned how to most effectively use the staff resources within the team to meet patient needs. The aim was to achieve this by having mechanisms in place which ensured that patients saw the health professional in the team most appropriate for their needs and as a consequence, reduce the number of patients who needed/wanted to see a GP. This would enable GPs to have longer (10 minute) appointments with those patients whose health needs require a GP consultation. GPs being able to have a longer appointments with patients was judged by the team to be a measure of quality of care because it is seen as an indication that patients needs are being met by the appropriate health professional in the team. It can also a measure effective use of resources. In addition, ithe measure is an indication that the mechanisms put in place to ensure that patients see the most appropriate health professional are effective (on the condition that the longer GP appointments do not increase the workloads of the other health professionals in the team).

Steps to clarifying the measure Over a month monitor the number of patients who have 10 minute appointment with a GP (this is a booked appointment, not a shorter appointment which over runs). Calculate the average number or % of patients in a week who can be offered a 10 minute appointment. Decide what is the target number or % of patients who can be offered a 10 minute appointments. When setting the target it may also be useful to consider the types of patients on the practice list who might benefit from longer appointment so as to establish the level of possible demand. It will also be useful to consider other factors which might affect the demand. For example, will there be seasonal variations?

Using the measure Over a 3 month period monitor the number or % of patients each week who are able to have a 10 minute appointment. Note the types of patients seen and, if there are weekly variations, the factors which might account for these variations Also note whether there are any unforeseen consequences (eg, increased work load for other team members, administrative difficulties).

Next steps Compare the average number or % of patients who could be offered a 10 minute appointment with the target which was set. If the target is achieved consider whether (a) any of the unforeseen consequences need to be taken into account, (b) whether there are ways that this target can be improved. If the target is not achieved consider whether any changes to the factors which accounted for variations might help the team top achieve the target. Also consider any impact of the unforeseen consequences.

Measure 6 Use of out of hours services by patient Measure = Percentage reduction in the use of private out of hours services by patients in a 6 month

period.

The PHC design team believed that using out of hours services did not provide patients which the best quality service (eg, they are seen by a health professional who does not know their medical history). Thus reducing the number of patients using out of hours services would reduce number receiving poor quality treatment. A reduction would also lead to more effective use of resources in the team the savings made from the reduction in the use of out of hours services could be used to employ an additional health profession in the team, eg a nurse practitioner. Steps to clarifying the measure Over a 3 month period monitor the use of the out of hours services by patients. Note the types of patients who use the service (is it a few frequent users or widely spread). Are the reasons for the reasons for using the out of hours services different for frequent users and occasional users? If the reasons are different, would it be possible to reduce the out of hour usage of these two groups? If there are fluctuations in usage? Note factors which account for these variations. Decide the acceptable level of out of hours service usage and the target amount of reduction. Set this target taking account of the factors which are associated with increases and decreases in usage (eg, seasonal variations, public holidays etc). It may be necessary to have separate targets for frequent users and occasional users.

Using the measure

Over a six month period monitor the use of the out of hours service. If there are weekly of monthly variations in usage, note the factors which might account for these. Next steps Calculate the average number of times the out of hours service has been used each month over the six month period and compare this with the target set. If the target has been met, using the other information collected, consider whether the most appropriate patients have been using out of hours service, and whether there are ways that (a) the usage could be reduced further, and (b)

whether steps could be taken to ensure that the out of hours service is used by the most appropriate patients. If the target was not met, use the information collected to consider changes which need to made to help ensure that the target is met in the future. Also use the information collected to consider whether the target is realistic.

Measure 7 Patients have access to an appropriate health professional Measure = Percentage of patients who have contact with a health professional from the team at a time and location most appropriate to them and to the professional in a 6 month period. This measure emerged from discussions the PHC design team had about how to achieve quality of care by ensuring that patients needs were met by the health professional most qualified to meet those needs. This is a complex measure and more work is required to develop a measure which can be used to assess effectiveness. Agreement needs to be reached on the following: which health needs can be most effectively met by which health professional. how health needs are assessed which health needs can be most effectively met in which location (eg, home, oneto-one consultation, booked appointment, drop-in, clinic etc.) This needs to take account of both patients and the health care professionals views. what % of patients it might be possible for each health care professional in the team to see at a time and location most appropriate to them and the patient.

Measure 8 Patients have access to a home visit from an appropriate health professional. Measure = Percentage of patients who have a home visit from the most appropriate health professional in a six month period.

This measure emerged from a discussion of the use of staff resources within the team. The PHC design team were considering which team members carried out home visits, the time of day when it was most convenient to carry out home visits and how to determine whether home visits were appropriate (ie, some home visits meet social rather than health needs). The aim is to ensure that only patients who need a home visit receive one, and that they are visited by the health professional (DN, GP, pharmacist, HV etc) who has the expertise to meet their needs. This is a complex measure. The following needs to be determined before it can be developed in a measure of effectiveness. which health (social) needs can only be met by a home visit how to assess these needs which of these health needs can be most effectively met by which health professional in the team what % of patients can realistically be seen at home by the most appropriate health care professional.

Objective: Enable personal and community responsibility for individual health

Measure 9 - Patients understand the role and function of the PHCT.

Measure = Number of patient requests, use health professionals time and PHCT services which are inappropriate in a 3 month period. Steps to clarify the measure: Patient understanding is demonstrated by appropriate use of the health professionals and other staff in the team, PHCT services, and appropriate requests for information. Define what are judged to be inappropriate uses of: health professionals time, and other staff in the team; PHCT services; inappropriate requests for information. Develop a checklist of the above and circulate to team members. Decide what is an acceptable level of inappropriate uses of: health professionals time, and other staff in the team; PHCT services; inappropriate requests for information.

Using the measure: Over a two week period all members of the PHCT record the number of inappropriate uses of health professionals and other team members time, and inappropriate use of PHCT services and requests for information. Note type of inappropriate use/request, and type of patient. After two weeks collate the data from all team members and calculate the number of (a) inappropriate uses of health professionals time, (b) number of inappropriate uses of other staff members time, (c) number of inappropriate uses of PHCT services by patients, and (d) number of inappropriate requests for information.

Next steps: If the number of inappropriate uses of staff time, PHCT resources and/or requests for information are unacceptable, develop interventions to reduce the number. Use information on the type of inappropriate use of time/services, and type of patients to target the information. After the interventions have been put in place repeat the measuring process to assess progress.

Objective: Efficient Use of Resources Measure 10 - Patients able to manage minor illness

Measure = Percentage of patients seen by health professionals in the team who had a minor illness which could have been managed themselves. Steps to clarify the measure: Define what is meant by minor illness. Develop a checklist of minor illnesses and circulate to all health professionals in the team. Decide what is an acceptable level of patients to see with a minor illness 10% or 40%? Decide whether some groups of patients should be excluded.

Using the measure: Over a two week period the health professionals in the team log each patient seen, and record which patients attend for minor illness. Note the type of minor illness, type of client. After two weeks collate the data from team members and calculate (a) total number of patients seen, (b) total number attending with minor illness. It may be useful to look at % of patients with a minor illness seen by each type of health professional, and to note which types of minor illness patients attended with, and the types of patients presenting with a minor illness.

Next steps: If the measure indicates that an unacceptable % of patients are seen who have minor illnesses decide on interventions to reduce the %. The data collected on types of minor illness, which health professionals are seeing these patients, the types of illnesses and types of patients can all be used to target the intervention/s. After interventions have been put in place repeat the measuring processes to assess progress.

Instructions for Record Sheet The data are being collected over 5 working days, starting on XXX. You can continue to collect data in w/c XXX, if you miss any days in the previous week. Please record on the form information about every patient you have contact with on each of the 5 days. Codes for each column are also printed on the bottom of the form. Type of illness Column one MA = minor illness, acute MC = minor illness, chronic Column two Please describe all types of minor illnesses you have recorded in addition to those in your leaflet, using medical terms. If the illness conforms to the definitions of minor illness in your leaflet, no further information is required. Type of contact Column three - 1 = phone Type of consultation Column four - 1 = routine Type of Patient Column five - 1 = female Column six - Patients age in years Seen before in last 7 days has the patient seen another health care professional in the last 7 days for the same illness as recorded in column 1? Column seven - record which health professional has seen the patient. 1 = GP 2 = PN 3 = DN 4 = HV 5 = CPN 6 = other 2 = male 2 = emergency 2 = home visit 3 = consultation C = chronic illness A = acute illness

Other comments Column eight - Please write down any other important information, and, if relevant, note if the patient has been referred inappropriately by other agencies such as secondary care/A&E/social services/dentist, as well as inappropriate internal referrals. Please give ALL completed recording forms to the Practice Manager.

Date:

Day of week: Type of illness 2) Description of 3) Type of contact 4) Type of consultatio n

Name: Type of patient 5) Gender 6) Age 7) Seen before in last week

Job title: 8) Comments

1) Code illness 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

1) Minor/Acute = MA information, and Minor/Chronic = MC referrals from social services and

3) 1 = phone 2 = home visit

5) 1 = female

7) 1 = GP 2 = male

8) Please include other 2 = PN

Chronic = C team members. Acute = A 2) Please describe illness if not minor according to the agreed definitions.

3 = consultation 6) Age in years 4) 1 = routine 2 = emergency

3 = DN 4 = HV 5 = Other

secondary care, other

Measure 11 - Patients/clients who do not attend for an appointment

Measure = Average percentage of total patients appointments not kept in a week (calculated over a 3 month period).

Steps to clarifying the measure: Monitor the DNAs for a one month period for each professional group (GP, DN, SN) to establish the current level in a one week period. Collect information on DNA levels for other comparable practices (i.e. have a similar type of practice population). Decide what is an acceptable DNA level for each professional group.

Using the measure: Monitor the DNAs for a two month period for each professional group. Monitor, where possible, the following: which patients DNA (persistent or across a wide range?) characteristics of DNAs (age/gender/ethnicity) when patients DNA (i.e. time of day/a particular GP, PN etc,/regular appointments booked in advance) whether DNAs make another appointment whether DNAs use other services (e.g. out of hours, home visit) Calculate the average % of patients/clients who DNA for each professional group in a one week period (this is the number of DNAs as a % of the total number of appointments made in each week). Compare the DNA average with the acceptable level for each occupational group and with other practices. Compare the % DNAs for each week in the two month period - does it vary from week to week? If so, can these differences be explained? Use the additional information collected to assess whether discernible patterns in the DNAs. Does it happen at certain times of the day, and/or do certain types of patients DNA more than others? This information can be used to make decisions about how to reduce DNAs. The information about whether DNA patients make another appointment and/or use other services will provide additional evidence about the cost of DNA to the team.

Next steps: If the DNA levels for the team as a whole and/or for specific occupational groups are unacceptably high introduce initiatives to reduce the level, and monitor progress using the measure.

The additional information collected will help to establish what types of initiatives might help to reduce DNAs. It may also be necessary to gather information from patients and clients about the reasons for DNA (is it because they forgot to attend? Because it is difficult to cancel an appointment? Because they are unaware of the implications to the team of DNA?).

Measure 12 Efficient use of administrative systems

Measure = Percentage of patients not attending appointments with health professionals in the team which result from errors in the administrative system. The additional work carried out at in the PHCT to explore the reasons why patients DNA revealed some problems with the administration systems. The result was that patients cancelled their booked appointment, but this was not entered on the system. In addition some patients reported that as they had attended the surgery close to a booked appointment they assumed that this had been cancelled. These findings suggest that some reduce in DNA rates would result from developing and improving current systems. Developing a measure requires the following: - further work to identify the range of system problems which could be improved by the team - calculate the current number of DNAs which result from system problems - improve and develop the system - continue to monitor the reasons for DNA and assess whether the number resulting from system problems decreases / or monitor the DNA rate and if this decreases attribute this to the improvements and developments in the system.

Measure 13 - Efficient use of GP resources in the team Measure = Average number of patients seen by a GPs in a week The PHCT had introduced a sit and wait session; all patients who went to the surgey between 10.00 qnd 11.00 were seen by a GP. Introducing the sit and wait session enabled the resources of the GPs to be used differently. Two GPs held patient consultations while a third carried out administrative tasks, dealt with telephone queries and carried out home visits. The aim was to offer the same number of faceto-face consultations with patients, but with a reduced GP resource. It was decided, therefore, that maintaining the number face-to-face consultations with patients was an indication of the success of the sit and wait. Data from the practice computer showed that there had been a substantial reduction in the number of patients seen by GPs (comparisons were made between a 1 week period in 1999 and the same week in 1998). Discussion of the reasons why GPs might be seeing fewer patients revealed that this reduction might indicate that the practice was infact using resources more efficiently. Reasons proposed were: Less use of locum doctors for home visits (therefore fewer patients being advised to see their own GP after the home visit). Nurses seeing more patients (diabetics, blood clinic, hypertensive), nurse practitioner (available on Fridays). Patients can call the surgery and talk to a GP and get advice. The pharmacist visits some patients at home to discuss medication.

This work emphasises the importance of looking at the activities of the team as a whole when assesses the effectiveness of specific aspects.

Objective: Continuous personal and professional development

Measure 14 Team member access to training

Measure = Percentage of who are satisfied with the extent to which their training needs are assessed and met in the previous year.

Steps to clarifying the measure

q

Agree what is an acceptable level of satisfaction within the team. Using the measure

q q

Each member of the PHCT completes the measure of satisfaction with training. Calculate the mean satisfaction with training score for each person (total the responses from each question .. and divide by the number of questions). Then calculate what % of staff report a satisfaction level at, above and below the ageed acceptable level. Next steps

Analysis of the responses to the individual questions in the training satisfaction measure can be used to determine the steps which need to be taken to improve access to training within the team. For example, these responses can indictate whether there are concerns about funding available to support training, or if there are issues relating to the identification of training needs. The former could be resolved by identifying additional sources of funding, while the latter could be tackled via the appraisal system.

Objective: High team member commitment, stress and satisfaction.

Measure 15- Team member commitment and satisfaction

Measure = Percentage of staff in the team who feel committed and satisfied

Steps to clarify the measure: Agree what is an acceptable level of commitment within the team Agree what is an acceptable level of job satisfaction

Using the measure: Each member of the PHCT completes the measures of commitment and satisfaction. It is important that confidentiality is maintained and that it is not possible for individual responses to be identified Calculate the mean job satisfaction for each person (total the responses from each question, extremely dissatisfied = 1 to extremely satisfied = 7 and divide by the number of questions, 16). Then calculate what % of staff report a satisfaction level at and above the agreed acceptable level for the team Calculate the mean commitment for each person (total the responses from each question, strongly agree = 5, strongly agree = 1, and divide by the number of questions, 6). Then calculate what % of staff report a level of commitment at and above the agreed acceptable level for the team

Next steps
Analysis of the responses to the individual questions in the commitment and job satisfaction measure can be used to determine the steps which need to be taken to improve the overall levels of commitment and satisfaction within the team.

Measure 16 - Team members use each others skills, knowledge and expertise appropriately

Measure = Percentage of team members who report that skills, knowledge and expertise within the team are used appropriately in 3 month period.
Steps to clarifying the measure
q

Agree what is an acceptable level of appropriate use of skills, knowledge and expertise.

Using the measure: Each member of the team completes the questions on the use of each others skills, knowledge and expertise. For each question, calculate the extent to which skills, knowledge and expertise are used appropriately (total the responses on each dimension = 1, = 5 and then calculate the mean (divide the total by the number of dimensions). Next steps If the levels of awareness and appropriate use of skill, knowledge and expertise are below the acceptable level the team could improve this by holding more effective meetings (when all members are encouraged to contribute to decision-making), by getting involved in joint projects together, and carrying out work shadowing.

Measure 17 - Effective team working

Measure = Percentage of requests for help and information and referrals from other team members which are inappropriate in a 3 month period. Steps to clarifying the measure: Agree what is an acceptable % inappropriate requests for help and information, and level of inappropriate referrals from other team members

Using the measure: Over a specified period (one or two weeks) each member of the team logs each time another team member requests help and information, and refers a patient. Against each, each team member notes whether this was an appropriate or inappropriate request for help/information or patient referral. It may also be useful to note which team member made the request for help/information or made the referral At the end of the specified period calculate what percentage of the total number of requests for information and patient referrals were inappropriate. Compare this with the acceptable levels agreed.

Objective: Responsiveness to client and community

Measure 18 - Patients Experiences of the PHCT service (1) Measure = Percentage of patients who report that their experiences of the PHCT services match the standard agreed by the PHCT. This can be assessed by measuring patients experiences of the PHCT. Steps to developing a measure: Identify all the aspects of the PHCTs work and how it is delivered which are known to be associated with patients satisfaction, e.g. not having to wait, getting repeat prescriptions, phone answered quickly, being able to get advice etc. Develop a checklist for patient asking if they have experienced each of the aspects associated with satisfaction. Either ask about experiences of the PHCT in general, e.g. Do you get your repeat prescription within 48 hours? always never

sometimes

Or ask about the contact with PHCT the patient has just had, e.g. Did you have to wait more than 2 days to get an appointment with the GP? yes no Identify other information you would like to collect from patients which might help you to use or understand the information you collect on patients experiences (e.g. age, gender, number of visits to the surgery in the previous month). It might be useful to ask patients to write their own comments. Consider what would be acceptable and unacceptable responses to the patients experiences questions, e.g. would you expect 90% of patients to report they got a repeat prescription within 48 hours or 10%? Also consider whether some areas are more important than others.

Using the measure: Over a one week period ask all patients attending the surgery to complete a checklist. Send a % of questionnaires to home addresses, and distribute via DN, HV etc. Collate the information from patients. Calculate a total score for each patient and the mean.

It may be useful to look at responses on each item separately (particularly if you considered some patient experiences to be more important), and to identify whether the views of different types of patients vary.

Next steps: If patients experiences of some aspects of the PHCT work are not as positive as the team had anticipated develop interventions to improve these. Or you may find experiences vary across different types of patients and the team want to take steps to remedy this.

After interventions have been put in place repeat the patient survey to assess progress.

Measure 19 - Patients experiences of the PHCT services (2) (Using the existing measure)

Measure = Percentage of patients whose experiences of the PHCT services meet the standard set by the team. Steps to clarifying the measure: On each of the questions in the patient Opinion survey agree the ideal standard the PHCT wants to achieve for example question 1, the length of time patients wait to get an appointment with a GP, what % of patients does the team aim to see on the same day/next day/after 2 days/3 days? On each of the questions in the Patient Opinion survey agree the expected standard that the PHCT currently achieves.

Using the measure: Over a one or two week period distribute questionnaires to patients attending the health centre/surgery, attending clinics, and those seen by the HV, DN, CPN and by other professionals carrying out domicillary care. Also survey a sample of patients, selected at random from the practice list, who have not been seen during the week. On each question calculate % of patients whose experiences of the PHCT services meet the ideal standards set by the teams. % of patients whose experience of the PHCT services meet the expected standard. Calculate a total score for % of patients whose experiences meet the ideal standard (total number of questions where patients experiences met the ideal standard and calculate this number as a % of the total number of questions). Calculate a total score for % of patients whose experiences meet the expected standard (total the number of questions where patients experiences met the expected standards and calculate this number as a % of the total number of questions).

Date ___________________

Patient Opinion Survey

Please could you answer the questions listed below. Your answers will help us to improve the service we provide for patients.

Age __________ years

Male

Female

1. The last time you wanted an appointment with any of the GPs, how soon did you get one? same day next day after 2 days longer ___________ 2. The last time you wanted an appointment with the GP of your choice, how soon did you get one? same day next day after 2 days after 3 days longer ___________ 3. How could the appointment service be improved? __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ __________________________________________________________________

4. The last two times that you phoned the surgery, how long did you wait for the phone to be answered? 1st time _________ mins 2nd time _________ mins Time of day: Time of day: morning morning afternoon afternoon

5. The last two times that you asked for a repeat prescription, how long did you have to wait to get it? 1st time ______________ days 2nd time ______________ days

Not applicable

6. Have you ever experienced problems/delays with getting a repeat prescriptions? Yes No

7. If yes, please give details of where the delay occurred e.g. at the Health Centre or at the chemist __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ __________________________________________________________________

8. How could the repeat prescription service be improved? __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ 9. Have you ever used the services of the Practice Nurse? Yes No

If No, please go to Question 13.

10. Did you have an appointment? Yes No

11. The last two times you had an appointment with the Practice Nurse, how long after the appointment time did you have to wait to see her? 1st time ____________ mins 2nd time ____________ mins

12. If your GP referred you to the Practice Nurse, how long did you sit in the waiting room until the practice nurse was available? _____________ mins

Did you know that you could make an appointment to see the Practice Nurse? Yes No

13. Have you required a doctors appointment and had to wait for the Practice Nurse to phone? Yes No

If No, please go to Question 15. 14. The last two times you used this service how long after 2pm did you have to wait for the Practice Nurse to phone you? 1st time _____________ mins 2nd time _____________ mins 15. How could the Practice Nurse services be improved? __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ __________________________________________________________________

16. Which of the following services do you think are offered by Health Visitors? Which have you used? offered Ante natal care Post natal advice/support Child development assessment Behaviour management Childcare advice Continence advice Adult support/advice Elderly support/advice used

If you do not use the services of the Health Visitor, please go to question 20. 17. Who is your named Health Visitor? _________________________________

18. In the last month: How many morning clinics did you attend? How many afternoon clinics did you attend? How long did you (your child) wait to be seen at each clinic? morning __________ __________ __________ __________ afternoon __________ __________ __________ __________

mins mins mins mins

mins mins mins mins

19. How could the services offered by Health Visitors be improved? __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ 20. Have you ever used of the services of District Nurses? Yes No

If No, please go to Question 27. 21. Have you got a named District Nurse? Yes No

22. If yes, who is your named District Nurse? __________________________

23. How long did you have to wait for a visit from a District Nurse for the following: (i) Urgent condition ________ hours days Not applicable ________

(ii) Discharge from hospital ________ hours ________ days (iii) Routine referral from GP________ hours days (iv) Nursing home assessment________ hours days

Not applicable

Not applicable

________

Not applicable

________

24. Did your GP tell you that they would arrange for the District Nurse to call? Yes No

If yes, did this happen within the time period given by the GP? Yes No Not applicable

25. Hospital discharge (if you have not been discharged from hospital in the past month, please ignore this section) Did the hospital tell you that the District Nurse would call to see you? Yes No

Did you have to contact the Health Centre before the District Nurse made a visit? Yes No

How long did you wait following discharge to see the District Nurse? ________days 26. How could the District Nursing services be improved? __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ 27. Please write below any other comments you would like to make about the Practice. __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ __________________________________________________________________ __________________________________________________________________

__________________________________________________________________ __________________________________________________________________ __________________________________________________________________ ____________________________________________________________ Many thanks for your help. Other measures developed in training workshop Quality % of patient records which are: relevant; concise; contemporaneous; legible; dated; signed and actioned. % of previouly agreed conditions for which there is an evidence based protocol which is reviewed annually. % of evidence based protocols which are audited. % of time spent on patient and non-patient contact. Number of complaints about access to services. Number of inappropriate experiences of access.

q q q q

q q

Team working Number of suggestions which are agreed and acted upon. % of time in a month when GPs are available for consultation with other members of the PHCT. Administrative efficiency % of time spent looking for case notes % of time spent preparing repeat prescriptions

q q

Appendix IV Training Programme Tools and Techniques for Assessing Performance

The Productivity Measurement and Enhancement System (ProMES) was used in the research to develop effectiveness measures from primary care. Objectives were developed in workshops with domain relevant experts from primary care (see Appendix III) and effectiveness measures were developed with representatives two primary health care teams (see Chapter 3 and Appendix II). ProMES is a theoretically grounded approach, based on the NPI theory of motivation (Naylor, Pritchard & Ilgen (1980), that has very practical applications. A critical feature of the ProMES process is that those people who performance is being assessed are involved in developing their own measurement and feedback systems. Participants learn to set clear objectives, to identify ways of measuring whether they are acheiveing these objectives, and to collect and use information to provide feedback on performance.

Two key lessons were learnt from the Workshops and the work carried out with primary health care team members developing measures. Firstly, the ProMES process provides a valuable learning experience, and secondly, it is possible to identify common objectives for primary care and to develop measures to assess performance against these objectives.

Having demonstrated the utility and value of ProMES, and encouraged by the exepriences of the representatives from primary care who had experienced the ProMES process, the research team developed a ProMES training programme. The programme was designed to train primary health care team members, and trust and health authority staff working with primary care teams, in the ProMES technique.

Letters were sent to all the primary health care teams involved in the research inviting representatives to attend the training programme. Responses were received from only two teams and four team members attended the training. Letters of invitation were also sent to representatives working with primary health care teams in the local community trust and health authority and four people attended from these organisations.

Details of the training programme are provided below.

Tools and Techniques for Assessing Performance Training Plan

Day 1 11.00 - 11.15 - Overall objectives of the training programme: Introductions OHP - Aims and objectives To develop skills which enable the participants to develop performance measures with primary health care teams. To develop skills which the participants can use to identify and use evidence to assess performance in primary health care teams. To teach participants how to use the measures produced by the research team. What else would the participants want to include? What specifically do they want to get out of the training? 11.15 - 12.00 - Introduction The purpose of this session is to: (a) clarify what is meant by performance (b) provide participants with an understanding of the contribution that measuring performance can make to team effectiveness in primary care, (c) raise awareness of the complexities of measuring performance in primary care. What is performance? It is important to be clear about what it is we are trying to measure. Discussion session - What do they understand by the performance of a PHCT? The definition to use: The outcomes from a team, what they produce, what they are trying to achieve. Introduce the basic principles of ProMES (Handout 1) - productivity - motivation - feedback - prioritising effort. Basic approach: - Design team - Clarify objectives - Develop measures of the objectives - Gather information with the measures and use this to assess extent to which meeting objectives, and to identify changes which can be made to improve performance. (Handout 2, Handout 3) Measuring performance and team effectiveness. (i) The importance of effective team working in multidisciplinary groups.

(ii) The principles of effective team working (clarity of objectives and feedback on whether achieving these objectives, participation, task style, support for innovation, reflexivity). Feedback is only useful in the context of clear goals. To measure performance PHCTs need to have clear objectives. Using measures can provide feedback on performance. In addition the process of developing and using measures requires participation and reflexivity. Thus measuring performance can make a considerable contribution to effective team working. Developing and using measures also enables PHCTs to use evidence based practice. (Handout 4, Handout 5) The difficulties of measuring performance in primary care Given what we are trying to achieve (measure the outcomes from PHCTs) Why might this be difficult? Discussion with the group. - multiple stakeholders, therefore, lack of agreement re. objectives, outcomes, desirable outcomes - outcomes difficult to measure (eg quality of care, patient satisfaction) - lack of sound evidence. - aspects not within the teams control. 12.00 - 1.00 - Basic Principles for Assessing Performance The purpose of this session is to: (a) Agree the objectives for primary care. (b) Identify potential sources of information which could be used to evaluate performance against these objectives. (c) Consider what are good and bad measures of performance. (Handout 6)

Step 1 - Setting objectives Outline the process used to develop the Objective for Primary Care (workshop, work with teams). Present the group with the Objectives. Rate the objectives in terms of importance (purpose of this is to get them thinking critically). Do they disagree with any of the objectives? (Handout7, Handout 8) Critically evaluate the objectives using the criteria for good objectives. (Handout 9) Step 2 - Identifying sources of feedback information Work in pairs, each discuss 2 objectives, and identify sources of information available which could provide feedback on performance on these objectives. (the aim at this stage is to raise awareness about all the sources of information available, not necessarily to identify the best sources). Step 3 - What is a good measure? Discussion of good and bad measures. Introduce the idea that what it is easy to measure, is not necessarily the best measure. Discuss the criteria for a good measure. (Handout 10) 2.00 - 3.30 - Measuring Use of Resources

The purpose of this session is to give participants the opportunity to work through the process of developing a measure. The task is for the group to develop a measure / measures which will indicate whether resources are being used effectively in a team. An example of a measure. In manufacturing organisations wastage of raw materials can add considerable amount to costs. A measure of performance in relation to the objective Efficient use of Resources could be, therefore, % reduction in the waste of raw materials. Exercise The group/s consider the following questions and develop a measure of use of resources. What are the resources used / available to a PHCT? (eg financial, skills, knowledge, equipment, time, rooms)? What evidence is available to indicate that resources are being used efficiently? What evidence is available to indicate that resources are being use inefficiently? Which resources is it most critical to use efficiently? (Ie which potentially have the greatest impact on the performance of the team?) Decide on one aspect of resources and develop a measure. Critically appraise the measure using the criteria for measures.

The groups work on their own with support from the trainers. They need to select a scribe and someone willing to feedback in the plenary session. The group give feedback on the process they went through to develop a measure, and the measure developed (Handout 12) 4.00 - 5.30 Plenary Session Feedback. What aspects of the process did they find easy/difficult? What problems did they encounter? Did they develop a good measure? Review learning (from the whole day) Day 2 9.00 - 10.45 - Measuring Quality of Care The purpose of this session is to give participants more experience of working through the process of developing a measure. iReview the learning points from the session on developing measures on use of resources (ie what have they learnt and will do differently ?) The task is to develop a measure / measures of in relation to the objective Quality of Care. This is a difficult exercise so we are providing some materials which might help. Introduce the 5 dimensions of quality. (Handout 13, Handout 14)

Exercise. The group/s consider the following questions. What is meant by quality of care in primary care? Whose perspective should be taken into account? What evidence is available to indicate that good quality of care is being provided by the PHCT? What evidence is available to indicate that quality of care is not good?

Which aspects of the PHCT services / types of conditions is it most critical to focus on? (ie which would improvements in service / care have the greatest impact on team performance?) Decide on one aspect of quality of care and develop a measure. Critically appraise the measure using the criteria for measures.

The groups work on their own with support from the trainers. They need to select a scribe and someone willing to feedback in the plenary session. It is a difficult task, but we will be there to help. We want them to feedback on the process they went through to develop a measure, and the measure developed 11.00 - 11.30 - Plenary session Feedback. What aspects of the process did they find easy/difficult? What problems did they encounter? Did they develop a good measure? Review learning. Important to emphasise that they are not now experts in developing measures, but are more aware of the process and the difficulties. With support from me they can do this with teams. 11.30 - 12.30 - Developing measures with teams / PCGs The purpose of this session is to provide participants with practical skills in running measurement development sessions with PHCTs Run this as a how to do it information giving session. Hightlight pitfalls and problems. Emphasis thet MUST follow the correct process. MUST allow everyone to have a voice. Must NOT impose their own objectives / agenda. In addition , given the nature of primary care, the process can generate conflict - because it starts to make the implicit explicit and hightlight differences in persepctives and values. Plus give the group the opportunity to discuss concerns, problems, obstacles that they foresee. etc. (Handouts 15 - 22) For example - setting-up the design team, explaining the purpose, agreeing objectives, starting to develop measures, logging progress, action planning, gathering information, assessing the value of information, using information 1.30 - 3.00 - Using Performance Measures The purpose of this session is to: (a) Familarise the participants with the measures developed by the research team so that they understand how and when ( to use them. (how to gather evidence) (b) Develop an understanding of how to use the feedback from measures to prioritise activities within the PHCT, and to change existing working practices/services. (how to use evidence) Explain how (the process) and why (to help teams to get feedback on aspects of their performance which were considered important to improving overall effectiveness) the measures were developed and used by the researchers. Provide each participant with a pack of measures (need to think on an interesting way of doing this). Discuss when and how the measures might be useful. (Handouts 23, Handout 24) Present some feedback data from the measures to the participants and discuss:

(a) how they would interpret this, what else they might need to know? (b) what actions might be taken as a result of the feedback? 3.15 - 4.00 - Plenary Session Review of learning Next steps and action plans Contacting the research team for support Support network/further meetings/next training.

Handout 1 Measuring and Enhancing Effectiveness in Primary Health Care Teams

Measuring and enhancing system effectiveness is an important element in any organisational system, and Pritchard (Pritchard, Jones, Roth, Stuebing & Ekeberg, 1988, 1989; Pritchard, 1990) has developed a sophisticated and widely applicable approach to this - the productivity measurement and enhancement system (ProMES). The ProMES approach is based on the theory of motivation presented by Naylor, Pritchard, and Ilgen (1980). In this theory, motivation is maximised when people see clear connections between their efforts and the behavioural products or results of these efforts, there are clear perceived connections between a persons products and their evaluations, and there are clear connections between these evaluations and valued outcomes. When these conditions are met, motivation is high. In addition, motivation is maximised when the different evaluators and controllers of rewards in

the persons environment such as the person himself/herself, peers, different supervisors, top management, and union personnel agree as much as possible on what should be done in the work and how it should be evaluated. When such agreement exists, the efforts of the person are more clearly directed and the same amount of effort results in greater productivity. In addition, stress and wasted effort are reduced. The ProMES system develops a formal method to measure productivity and uses these measurements as feedback to people doing the work to help them increase their productivity through maximising motivation. The idea is to maximise the variables indicated in the theory so that motivation will also be maximised. People are given the tools to do the work better and at the same time help them feel a sense of ownership in the resulting system and empowerment in determining important aspects of their work. One of the key elements in ProMES is feedback. People doing the work get regularly occurring, high quality feedback about how the work unit is doing. The personnel in the work unit then use this feedback to develop plans for improving productivity. Feedback after this time tells them how well the plans they developed have actually improved productivity. Furthermore, since they are heavily involved in the design of the measurement system and resulting feedback system, they have more confidence in its validity and accept it more than systems imposed from above. The ProMES approach has been applied and evaluated in a wide range of settings (Pritchard, 1995) and substantial improvements in performance have been shown as a result of using ProMES.

Handout 2

Practical tips and guidelines Why use this system? it makes you think! it gives you useful information it indicates where you should focus your resources it allows you to define your own measurement system - it puts you in control it increases your participation you are the first to know about any problems people report less stress people know how they are being evaluated everyone has to agree the priorities - they are not imposed you get valuable feedback

Handout 3

Practical tips and guidelines Key implementation principles Measurement is the foundation It takes a lot of work to measure well What you measure is what you get Measures for decision making are different from those for motivation Good measurement makes feedback easy Good feedback leads to productivity improvements People want to do a good job The key is to give them the tools People want control over their lives Acceptance of the system is essential for success Participation leads to acceptance

Handout 4

Teamworking in Primary Care

The idea that teams are important to modern organisations was established about 70 years ago. However, in only the past 15 years has the idea been seized and widely acted on by large numbers of organisations in the public and private sectors (Guzzo, 1996). But how effective are teams within organisations generally?

Macy and Izumi (1993) conducted an analysis of 131 organisational change studies in order to determine their effectiveness. Those interventions with the greatest effects on financially-related measures of organisational performance were teamrelated interventions. These also reduced turnover and absenteeism more than did other interventions, showing that team-oriented practices can have broad positive effects in organisations. Abblebaum and Batt (1994) offer convergent evidence. They reviewed the results of a dozen surveys of organisational practices as well as 185 case studies of innovation in management practices. They too found compelling evidence that teams contribute to improving organisational effectiveness, particularly increasing efficiency and quality. Other researchers provide evidence of the impact of team-based work practices on organisational performance. Kalleburg and Moody (1994) studied over 700 work establishments and found that those in which teamwork was developed were more effective in their performance than those in which were not used.

The importance of teamworking has been emphasised in numerous reports and policy documents on the National Health Service. One recent document (NHSME 1993) particularly emphasised the importance of teamworking if health and social care for people in local communities were going to be of the highest quality and efficiency.

The best and most cost-effective outcomes for patient and clients are achieved when professionals work together, learn together, engage in clinical audit of outcomes together, and generate innovation to ensure progress in practice and service.(para 4.3)

Overall, research based evidence of teamworking in primary health care in the UK is consistent with research in other sectors in suggesting the value of this way of working for effectiveness and efficiency. Primary care team working has been reported to improve health delivery and staff motivation (Wood, Farrow and Elliot, 1994) and to have led to better detection, treatment, follow-up and outcome in hypertension (Adorian, Silverberg, Tomer and Wamosher, 1990). In a longitudinal study of 68 primary health care teams, Poulton (1995) found a clear relationship between teamwork and effectiveness. Those teams with high levels of clarity of team objectives and team members commitment to those objectives were more effective than those with unclear objectives.

However, despite these encouraging research studies, there is considerable evidence that the context of primary health care is such that there are substantial barriers to co-operation and collaboration in the delivery of primary health care. Bond et al (1985) found little interprofessional collaboration in primary health care teams in their study of 309 paired professionals. West and Poulton (1995) examined primary health care team functioning in 68 practice teams and found that on all 4 dimensions of team functioning primary health care teams scored significantly lower than the other team types. West, Poulton and Hardy (1994) in a study of 9 primary health care teams identified structural, managerial and employment patterns in primary care as crucial in undermining the effectiveness of teamworking. These barriers to co-operation and collaboration need to be removed or reduced for teamworking to be effective in primary heath care.

There are a number of key elements to effective teamwork (Guzzo and Shea, 1992):

First, Individuals should feel that they are important to the success of the team. When individuals feel that their work is not essential in a team, they are less likely to work effectively with others or to make strong efforts towards achieving team effectiveness. Roles should be developed in ways which make them indispensable and essential.

Individuals roles in the team should be meaningful and intrinsically rewarding. Individuals tend to be more committed and creative if the tasks they are performing are engaging and challenging.

Teams should also have intrinsically interesting tasks to perform. Just as people work hard if the tasks they are asked to perform are intrinsically engaging and challenging, when teams have important and interesting tasks to perform, they are committed, motivated and co-operative (Hackman, 1990).

Individual contributions should be identifiable and subject to evaluation. People have to feel not only that their work is indispensable, but also that their performance is visible to other team members.

Above all there should be clear, shared team goals with built-in performance feedback. Research evidence shows consistently that where people are set clear targets at which to aim, their performance is generally improved. For the same reasons it is important for the team as a whole to have clear team goals with performance feedback.

In primary health care, by and large, the first three conditions for effective teamworking hold true. However, in primary health care teams it is rare for individual contributions to be measured and feedback on performance given. Moreover, primary health care teams tend not to have clear, specific objectives and goals and

feedback on performance against those objectives is rarely available. The development of teamworking in primary healthcare, therefore, needs to focus on developing clear, shared objectives and on providing feedback on performance.

Handout 5 The difficulties of measuring performance in primary care

There are multiple stakeholders in primary care (different professional groups, and organisations), therefore, lack of agreement about objectives, what are the outcomes from primary care teams, and what are desirable outcomes. Many of the outcomes are difficult to measure (e.g. quality of care, patient satisfaction). There is a lack of sound evidence. Many of the factors which influence outcomes are not within the teams control (e.g. other agencies, characteristics of practice population).

Handout 6

Practical tips and guidelines Setting Objectives

ask the team what it is they are trying to accomplish for their organisation focus on larger objectives - give the group examples which are as similar to their work as possible this stage is typically not difficult consensus should be easy to reach at this stage the discussion at this point sets the tone for the future - there needs to be balance between the facilitator saving the group time and taking control

Handout 7 Core Objectives for Primary Health Care

To promote, maintain and improve health Provide high quality health care Accurate identification of individual and population health and care needs Review and improve the effectiveness of health care provision

Manage illness, injury and disease taking account of agreed standards and evidence based practice Enable personal and community responsibility for individual health Enable patients/clients to make informed decisions about their own health Proactively encourage positive health behaviour Implementation of health education and preventative care programmes Efficient use of resources Human resources - skills, knowledge, expertise, time Physical resources - budgets, equipment, premises Continuous personal and professional development Individual annual training plans which take account of the plans for the PHCT Equal access to training/development resources High team member commitment, stress and satisfaction

Teamworking

Mechanisms for reviewing and acting upon staff dissatifactions, conflicts and complaints Responsiveness to clients and community

Mechanisms for gathering information and feedback from clients/community stakeholders/opinion leaders. Collaboration and partnership with other relevant organisations

Handout 8 Core Objectives for Primary Health Care Teams

To what extent do you think your team effectively meets the following objectives and sub-objectives?

To promote, maintain and improve health importance important 1 2 3 4 5 6 7 - Provide high quality health care

Of no

Very

1 5 6 7

- Accurate identification of individual and population health and care needs 7 - Review and improve the effectiveness of health care provision 5 6 7 - Managing illness, injury and disease taking account of agreed standards and evidence 5 6 7 based practice

Enable personal and community responsibility for individual health

- Enable patients/clients to make informed decisions about their own health 7 - Proactively encourage positive health behaviour 1 2 3 4 5 6 7 - Implemention of health education and preventative care programmes 7

Efficient use of resources importance important 1 2 3 4 5 6 7 -Human resources (skills, knowledge, expertise time) 7

Of no

Very

- Physical resources (budgets, equipment, premises) 1 2 3 4 5 6 7

Continuous personal and professional development - Individual annual training plans which take account of the plans for the PHCT 7 - Equal access to training/development resources 7

High team member commitment, stress and satisfaction 7 - Teamworking 1 2 3 4 5

- Mechanisms for reviewing and acting upon staff dissatisfactions, conflicts and complaints 7

Responsiveness to clients and community 1 2 3 4 5 6 7 - Gather information and feedback from clients/ community stakeholders/opinion leaders 7

Collaboration and partnership with other relevant organisations 7

Handout 9

Criteria for Objectives

stated in clear terms if exactly that objective was done, the organisation would benefit the set of objectives must cover all important aspects of the work objectives must be consistent with the broader organisation higher management must be committed to each objective keep the number of objectives manageable, normally 3 to 8

Handout 10

Practical tips and guidelines Criteria for measures

the measure must be consistent with the objectives of the broader organisation if the measure was maximised would the organisation benefit all important aspects of each objective must be covered by the set of measures higher management must be committed to the measures measures must be under control of the staff measures must be understandable and meaningful to staff it must be possible to provide information on the measure in a timely manner the data must be cost effective to collect the information provided by the measure must neither be too general or too specific

Handout 11 Measuring Use of Resources

What are the resources used / available to a PHCT? (eg financial, skills, knowledge, equipment, time, rooms)? What evidence is available to indicate that resources are being used efficiently? What evidence is available to indicate that resources are being used inefficiently? Which resources is it most critical to use efficiently? (i.e. which potentially have the greatest impact on the performance of the team?) Decide on one aspect of resources and develop a measure. Critically appraise the measure using the criteria for measures.

Select a scribe and someone willing to feedback in the plenary session. . We want them to feedback on the process they went through to develop a measure, and on the measure/s developed

Handout 12

Practical tips and guidelines Developing Measures

ask how they would show that the stated objectives were being met this is a difficult step for the design team to do it is frustrating - tell the group they will feel this you must train the design team to develop and evaluate measures the design team may not know the answer but they can find out if someone tells you that you cannot measure what they do then ask them how they think they are doing?, is it different to last year?. If they have an idea of their performance then it can be measured.

Handout 13 Dimensions of Quality

Effectiveness: Is the treatment given the best available in a technical sense, according to those best equipped to judge? What is their evidence? What is the overall result of the treatment?

Acceptability: How humanely and considerately is this treatment/service

delivered? What does the patient think of it? What would/does an observant third party think of it (How would I feel if it were my nearest and dearest?) What is the setting like? Are privacy and confidentiality safeguarded?

Efficiency:

Is the output maximised for a given input or (conversely) is the

input minimised for a given level of output? How does the unit cost compare with the unit cost elsewhere for the same treatment/service?

Access:

Can people get this treatment/service when they need it? Are

there any identifiable barriers to service - for example, distance, inability to pay, waiting lists, and waiting times - or straightforward breakdowns in supply?

Equity:

Is this patient or group of patients being fairly treated relative to

others? Are there any identifiable failings in equity - for example, are some people being dealt with less favourably or less appropriately in their own eyes than others?

Relevance:

Is the overall pattern and balance of services the best that could be achieved, taking account of the needs and wants of the population as a whole?

Handout 14 Measuring Quality of Care

What is meant by quality of care in primary care? Whose perspective should be taken into account? What evidence is available to indicate that good quality of care is being provided by the PHCT? What evidence is available to indicate that quality of care is not good? Which aspects of the PHCT services / types of conditions is it most critical to focus on? (ie which would improvements in service / care have the greatest impact on team performance?) Decide on one aspect of quality of care and develop a measure. Critically appraise the measure using the criteria for measures.

Select a scribe and someone willing to feedback in the plenary session. Feedback on the process they went through to develop a measure, and the on the measure/s developed

Handout 15 Basic ProMES Approach

Gaining management and staff support Set-up a design team Identifying objectives Develop measures for the objectives Gather information with the measures Feedback from the measures Identify changes which can be made to improve performance.

Handout 16

Practical tips and guidelines Conditions for success

Management support - this means public support of the project on a regular basis, providing the resources of the project regularly, solving project problems and continuing the project until a clear evaluation can be made Trust between management and staff - process needs to be fully explained including advantages and costs, explain why us?, make participation voluntary, explain how the whole team will be involved even though only part will be on the design team, explain how the design team will be chosen All interested parties approve the project - consider stakeholders, partners, unions Values match between management and staff - all see potential improvement as valuable and have a long range perspective, all see participation/acceptance as essential Stable personnel and group structure - are there any major upheavals in staff premises or technology

Handout 17

Practical tips and guidelines Resources Needed for Development

Time for design team meetings Uninterrupted meeting setting Full attendance by design team Access to existing data Meeting with management for approval of the system

Resources needed for implementation

Collecting of existing and new data for feedback Preparation and distribution of feedback results Regular meetings of the group Regular meetings of the group members to discuss feedback reports

Handout 18

Practical tips and guidelines Checklist for starting a project

All interested constituencies have been involved Benefits and costs clearly explained to all Have assessed trust and common values Have management support (see number 2) Have staff support (see number 2)

Handout 19

Practical tips and guidelines Selecting the design team

people who are respected must be a cross section of the larger group e.g. practice nurse, senior receptionist, general practitioner, district nurse, practice manager, health visitor no-one occupational group should dominate if someone is unable to attend they must nominate a deputy to represent them individuals should feel confident about themselves and who/what they represent and feel able to represent the views of their group

Handout 20

Practical tips and guidelines Logging progress

keep an accurate record of the progress made at each meeting provide the design team with an update of their progress at regular intervals you may find that between meetings the design team do not any homework and you will need to review previous sessions at the beginning of new sessions provide the team with the list of objectives when completed along with their list of measures the design team must also report progress to the other members of the larger group and your updates is a useful way of achieving this the team will need reassurance that progress is being made even if it is minor to spur them on to the next stage

Handout 21 Practical tips and guidelines Action planning

at the end of each session you will need to plan for the actions to be carried out by the next session this may include collecting information and specific data it must be clear from the outset that this is not the responsibility of the facilitator members of the team will need to volunteer to carry out actions it is easy to walk away from a session believing that everyone knows what they are soposed to do - it is more than likely someone does not check and check until you are satisfied that everyone understands their responsibilities if you do this regularly the team will start to check you! it is a good idea to have a session plan, you must be clear about what you want to achieve at each session also record the feedback the team give you at the beginning of the session, it often proves useful later

Handout 22

Practical tips and guidelines Possible problems

lack of commitment from the design team - not present at meetings or lack of response getting stuck - cant see how to move forward design team going off on a tangent - avoiding what can be hard thinking work disagreement between members - possible personal issues domineering members - other members passively agreeing feeling lost with it all - do not know what to do next not been given the required time and space for the sessions or interrupted larger team losing interest management support withheld no-one willing to collect data got the data dont know what to do next

Handout 23

Practical tips and guidelines Gathering information the information gathered must be of good quality the team must think about the most effective ways to gather information you will need to discuss their expectations, what do they expect to be the outcome they may need to design record sheets or charts or questionnaires which must be piloted to evaluate their effectiveness who will distribute, collect, collate, evaluate and present the data to the design team these may all be new skills but is an excellent development opportunity the team will need support in this, from you and from each other

Assessing the value of the information does the data answer the question are you satisfied that the data is good quality did you discuss how to asses the data before you began do you have the resources to assess the data, people, time, technology

Handout 24

Practical tips and guidelines Using information

does this data provide you with useful information you will need to discuss dissemination with the team you must feedback the results of data gathering if you want people to co-operate in the future what next - if you have some useful information you can then use this to plan ahead and use the information to make your case for change

_______________________________________________________________ _____

BIBLIOGRAPHY
______________________________________________________ _____
Adorian, D., Silverberg, D.S., Tomer, D. & Wamasher, Z. (1990). Group discussion with the health care team: A method of improving care of hypertension in general practice. Journal of Human Hypertension, 4 (3), 265 - 268. Alderfer, C.P. (1977). Group and intergroup relations in J.R. Hackman and J.L. Suttle (eds) Improving the quality of work life. Pallisades, C.A.: Good year. pp 277 - 296. Alexander, J. A., Lichtenstein, R. & DAunno, T. A. (1996). The effects of treatment team diversity and size on assessments of team functioning. Hospital & Health Services Administration, 41, 37-53. Allen, N. J. (1996). Affective reactions to the group and organisation. In M. A. West (Ed.), Handbook of Work Group Psychology (pp. 371-396). Chichester: Wiley. Amabile, T.M. (1983). The social psychology of creativity: A componential conceptualization. Journal of Personality and Social Psychology, 45, 357-376. Ancona, D.F. & Caldwell, D.F. (1988). Bridging the boundary: External activity and performance in organisational teams. Administrative Science Quarterly, 37, 634-665. Anderson, N.R. & King, N. (1993). Innovation in organisations. In C.L.Cooper & I.T. Robertson (eds). International Review of Industrial and Organizational Psychology, Vol 8, Chichester: Wiley. Anderson, N, & West, M.A. (1994). The Team Climate Inventory: Manual and Users Guide. Windsor, England: NFER-Nelson. Anderson, N. & West, M.A. (1998). Measuring climate for work group innovation: development and validation of the team climate inventory: Journal of Organizational Behaviour, Vol 19, 235 - 258. Applebaum, E. & Batt, R. (1994). The New American Workplace. Ithaca, NY: ILR Press. Audit Commission (1992). Homeward Bound: A New Course for Community Health. London: HMSO. Bales, R.F., Strodtbeck, F.L., Mills, T.M. & Roseborough, M.E. (1951). Channels of communication in small groups. American Sociological Review, 16, 461-468. Baumeister, R.F. & Leary M.R., (1995). The need to belong desire for interpersonal attachments as a fundemental motivator. Psychological Bulletin, 117, 497-529.

Berger, J., Fisek, M.H., Norman, R.Z., & Zelditch Jr, M. (1977). Status characteristics and social interaction. NY: Elsevier.

Berger, J., Rosenholtz, S.J., & Zelditch Jr., M. (1980). Status organizing processes. Annual Review of Sociology, 6, 479-508. Bhugra, D., Bridges, K., & Thompson, C. (1995). Caring for a community: the community care policy of the Royal College of Psychiatrists. London: Royal College of Psychiatrists. Billings, R.S., Milburn, T.W. & Schaalman, M.L. (1980). A model of crisis perception: A theoretical and empirical analysis. Administrative Science Quarterly, 25, 300-316. Blakar, R.M. (1985). Towards a theory of communication in terms of precondition: A conceptual framework and some empirical explorations. In H. Giles and R.N. St Clair eds), Recent Advances in Language, Communication and Social Psychology. London: Lawrence Erlbaum. Blau, Peter M. (1977). Inequality and Heterogeneity. New York: Free Press. Bobko, P., & Colella, A. (1994). Employee reactions to performance standards: A review and research proposition. Personnel Psychology, 47, 1-29. Bond, J., Cartilidge, A.M., Gregson, B.A., Philips, P.R., Bolam, F., & Gill, K.M. (1985). A study of interprofessional collaboration in primary health care organisations. Report No 27 (2), Newcastle-upon-Tyne, Health Care Research Unit, University of Newcastle-upon-Tyne. Borrill, C.S. & West, M.A. (1997). Effectiveness in primary health care. CAIPE bulletin. No 14. Borrill, C.S. & West, M.A. (1998). Strain in primary health care. Unpublished report. Institute of Work Psychology, University of Sheffield, England. Borrill, C.S., Wall. T.D., West, M.A., Hardy, G.E., Shapiro, D.A., Haynes, C.E., Stride, C.B., Woods, D. and Carter, A.J. (1998) Stress among staff in NHS Trusts. Institute of Work Psychology, University of Sheffield, Psychological Therapies Research Centre, University of Leeds. Bottger, P.C. & Yetton, P.W. (1987). Improving group performance by training in individual problem solving. Journal of Applied Psychology, 72, 651-657. Bowers, D.G. & Seashore, S.E. (1966). Predicting organisational effectiveness with a four-factor theory of leadership. Administrative Science Quarterly, 11, 238-263. Brewer, N., Wilson, C. & Beck, K. (1994). Supervisory behavior and team performance amongst police patrol sergeants. Journal of Occupational and Organizational Psychology, 67, 69-78. Brown, R.J. (1988). Group Processes: Dynamics Within and Between Groups. London: Blackwell.

Bryk, A. & Raudenbush, S. (1992) Hierarchical Linear Models: Applications and data analysis methods. In J. Deleeuw (Ed.) Advanced quantitative techniques in the social sciences series. Newbury, CA: Sage Publications. Burns, T., & Stalker, G.M. (1966). The Management of Innovation. London: Tavistock Publications. Carletta, J., Garrod, S., & Fraser-Krauss, H. (1998). Communication and Placement of Authority in Workplace Groups The Consequences for Innovation. Small Group Research, 29(5), 531-559. Carter, M.F., Evans, K.E., Crosby, C., Prendeergast, L.A., & De Sousa Butterworth, K.A. (1997). The all-Wales community mental health team survey. Bangor: Health Services Research Unit. Campion, M.A., Medsker, G.J. & Higgs, A.C. (1993). Relations between work group characteristics and effectiveness: Implications for designing effective work groups. Personnel Psychology, 46, 823-850. Campion, M.A., Papper, E.M. & Medsker, G.J. (1996). Relations between work team characteristics and effectiveness: A replication and extension. Personnel Psychology, 49, 429-689. Cant, S. & Killoran, A. (1995) Team tactics: a study of nurse collaboration in general practice, Health Education Journal, 52, 203 - 8. Coch, L. & French, J. R. (1948). Overcoming resistance to change. Human Relations, 1, 512-532. Cohen, S.G. & Bailey, D.E. (1997) What makes teams work: Group effectiveness research from the shop floor to the executive suite. Journal of Management, No 3. 239 - 290. Connolly, T., Conlon, E.J. & Deutsch, S.J. (1980). Organizational effectiveness: a multi-disciple-constituency approach, 98, 310 - 357. Cott, C. (1997). We decide, you carry it out: A social network analysis of multidisciplinary long-term care teams. Social Science & Medicine, 45 (9), 14111421. Cowan, D.A. (1986). Developing a process model of problem recognition. Academy of Management Review, 11, 763-776. D'Zumla, T.J. & Goldfried, M.R. (1971) Problem solving and behaviour modifications. Journal of Abnormal Psychology. 78, 107 - 126. Davenport, T.H. (1993). Process Innovation: Re-engineering Work Through Information Technology. Cambridge, MA: Harvard Business School Press. Deming, W.E. (1986). Out of the Crisis. Cambridge, MA: Center for Advanced Engineering Study, Massachusetts Institute of Technology.

Department of Health (1990). The Care Programme Approach for people with a mental illness referred to the specialist psychiatric services. London, Department of Health, HC (90) 23/LASSL (90) 11. Department of Health (1995). Building bridges: a guide to arrangements for interagency working for the care and protection of severely mentally ill people. London: HMSO. Department of Health (1997). The Patient's Charter: mental health services. London, Department of Health. Department of Health (1999). National Service Framework for Mental Health: Modern standards and service models. London: Department of Health. Dreachslin, J.L., Hunt, P.L. & Sprainer, E. (2000). Workforce diversity: Implications for the effectiveness of health care delivery teams. Social Science & Medicine, 50, 1403-1414. Dreachslin, J.L., Hunt, P.L. & Sprainer, E. (1999a). Communication patterns and group composition: Implications for patient-centred care team effectiveness. Journal of Healthcare Management, 44, 252-268. Dreachslin, J.L., Hunt, P.L. & Sprainer, E. (1999b). Key indicators of nursing team performance: Insights from the front line. The Health Care Supervisor, 17, 70-76. Drolen, C.S. (1990). Current community mental health center operations: Entrepreneurship or business as usual? Community Mental Health Journal, 26, 547558. Drory, A. & Shamir, B. (1988). Effects of organizational and life variables on job satisfaction and burnout. Group and Organization Studies, 13 (4), 441-455. Eden, D. (1990). Pygmalion without interpersonal contrast effects: Whole groups gain from raising manager expectations. Journal of Applied Psychology, 75, 394-398. Eggert, G.M., Zimmer, J.G., Hall, W.J. & Friedman, B. (1991). Case management: A randomised controlled study comparing a neighbourhood team and a centralized individual model. Health Services Research, 26 (4), 471-507. Faulkner, A. (1997). Knowing our own minds: A survey of how people in emotional distress take control of their lives. London: Mental Health Foundation. Fay, N., Garrod, S., & Carletta, J. (2000). Group discussion as interactive dialogue or serial monologue: The influence of group size. Psychological Science, 11(6), 487492. Field, R. & West, M.A. (1995). Teamwork in primary health care. Two Perspectives from practices. Journal of Interprofessional Care, 9, 2, 123-130. Freeman, M., Miller, C. & Ross, N. (2000). The impact of individual philosophies of teamwork on multi-professional practice and the implications for education. Journal of Interprofessional Care, 14 (3), 237-247.

Galbraith, J. R. (1993). The business unit of the future. In J. R. Galbraith, E.E. Lawler III & Associates (Eds), Organizing for the Future: The New Logic for Managing Complex Organizations. San Francisco: Jossey-Bass. Galbraith, J. R. (1994). Competing with Flexible Lateral Organisations (2nd edn). Reading, MA: Addison-Wesley. Galbraith, J.R., Lawler, E.E. III & Associates (1993). Organizing for the Future: The New Logic for Managing Complex Organizations. San Francisco: Jossey-Bass. George, J.M. (1989). Mood and absence. Journal of Applied Psychology, 74, 317324. George, J.M. (1990). Personality, affect, and behavior in groups. Journal of Applied Psychology, 75, 107-166. George, J.M. (1995). Leader positive mood and group performance: The case of customer service. Journal of Applied Social Psychology, 25, 778-794. George, J.M. (1996). Group affective tone. In M.A. West (Ed.), Handbook of Work Group Psychology (pp. 77-94). Chichester: Wiley. George, J.M. & Bettenhausen, K. (1990). Understanding pro-social behaviour, sales performance and turnover. A group-level analysis in a service context. Journal of Applied Psychology, 75, 698-709. Gladstein, D. (1984). Groups in context: A model of task group effectiveness. Administrative Science Quarterly, 29, 499-517. Goldberg, D.P. & Williams, P. (1991). A user's guide to the General Health Questionnaire. Windsor: NFER-Nelson. Goldberg. D.P. (1972). The detection of minor psychiatric illness by questionnaire. Oxford: Oxford University Press Goni. S. (1999). An analysis of the effectiveness of Spanish primary health care teams. Health Policy, 48, 107-117. Grusky, O. (1995). The organization and effectiveness of community mental health systems. Administration & Policy in Mental Health, 22, 361-388. Guzzo, R.A. (1996). Fundamental considerations and about work groups. In M.A. West (ed) Handbook of Work Group Psychology. Chichester: John Wiley. Guzzo, R. A. & Dickson, M. W. (1996). Teams in organisations: Recent research on performance and effectiveness. Annual Review of Psychology, 46, 307-338. Guzzo, R.A., Jette, R.D., & Katzell, R.A. (1985). The effects of psychologically based intervention programs on worker productivity: A meta-analysis. Personnel Psychology, 38, 275-291. Guzzo, R. A. & Salas, E. (Eds) (1995). Team Effectiveness and Decision Making in Organisations. San Francisco: Jossey-Bass.

Guzzo, R.A. & Shea, G.P. (1992). Group performance and intergroup relations in organisations. In M. D. Dunnette and L. M. Hough (Eds), Handbook of Industrial and Organizational Psychology, (Vol 3, pp. 269-313). Palo Alto, CA: Consulting Psychologists Press.

Hackman, J.R. (1987). The design of work teams. In J.W. Lorsch (ed) Handbook of Organisational Behaviour. Englewood Cliffs, NJ: Prentice-Hall. Pp 315 - 342. Hackman, J.R. (1990). (Ed), Groups That Work (and Those That Don't): Creating Conditions for Effective Teamwork. San Francisco: Jossey-Bass. Hackman, J.R., Brousseau, K.R. & Weiss, J.A. (1976). The interaction of task design and group performance strategies in determining group effectiveness. Organizational Behavior and Human Performance, 16, 350-365. Hackman, J.R. & Morris, C.G. (1975). Group task, group interaction process, and group performance effectiveness: A review and proposed integration. In L. Berkowitz (Ed.), Advances in Experimental Social Psychology, (Vol. 8). New York: Academic Press. Hannigan, B. (1999). Joint working in community mental health: prospects and challenges. Health and Social Care in the Community, 7, 25-31. Hardy, G. E., Shapiro, D.A., Hayes, C.E. & Rick, J.E. (1999). Validation of the General Health Questionnaire using a sample of employees from the health care services. Submitted. Haynes, C.E., Wall, T.D., Bolden, R.I. & Rick, J.E. (1998). Measures of perceived work characteristics for health service research: test of a measurement model and normative data. Submitted. Hedburgh, B.L.T., Nystrom, P.C. and Starbuck, W.H. (1976) Company on seesaws: prescriptions for a self designing organisation. Administrative Science Quarterley, 21, 41 - 65. Hill, F. (1998). Trying to catch a cloud: organizational climate in the NHS. Unpublished PhD Thesis, Institute of Work Psychology, University of Sheffield. Hirokawa, R.Y. (1990). The role of communication in group decision-making efficacy: A task-contingency perspective. Small Group Research, 21, 190-204. Hoffman, L.R. & Maier, N.R.F. (1961). Sex differences, sex composition, and group problem-solving. Journal of Abnormal and Social Psychology, 63, 453-456. Hogg, M. & Abrams, D. (1988). Social Identifications: A Social Psychology of Intergroup Relations and Group Processes. London: Routledge. Hughes, S.L., Cummings, J., Weaver, F., Manheim, L., Brawn, B. & Conrad, K. (1992).

A randomised trial of the cost effectiveness of VA hospital-based home care for the terminally ill. Health Services Research, 26 (6), 801-817. Jackson, G., Gater, R., Goldberg, D., Tantam, D. Loftus, L. & Taylor, H. (1993). A new community mental health team based in primary care: A description of the service and its effect on service use in the first year. British Journal of Psychiatry, 162, 375-384. Jackson, L.A., Sullivan, L.A. & Hodge, L.N. (1993). Stereotype effects on attributions, predictions and evaluations: No two social judgements are quite alike. Journal of Personality and Social Psychology, 65 (1), 69-84. Jackson, S.E. (1996). The consequences of diversity in multidisciplinary work teams. In M.A. West (Ed.), Handbook of Work Group Psychology, pp. 53-76, Chichester: Wiley. Jacobs, D. & Singell, L. (1993). Leadership and organizational performance: Isolating links between managers and collective success. Social Science Research, 22, 165-189. Janis, I.L. (1982). Groupthink: A Study of Foreign Policy Decisions and Fiascos, 2nd ed. Boston: Houghton Mifflin. Jansson, A., Isacsson, A. & Lindhom, L.H. (1992). Organization of health care teams and the populations contacts with primary care. Scandinavian Journal of Health Care, 10, 257-265. Jervis, I.L. (1976). Perception and Misperception in International Politics. Princeton, NJ: Princeton University Press. Jones, R.V.H. (1992). Teamworking in primary care:how do we know about it? Journal of Interprofessional Care, Vol 6, p25-29. Juran, J.M. (1989). Juran on Leadership for Quality. New York: Free Press. Jussim, L. (1986). Self-fulfilling prophecies: A theoretical and integrative review. Psychological Review, 93 (1), 429-445. Jussim, L., Coleman, L.M. & Lerch (1987). The nature of stereotypes: A comparison and integration of three theories. Journal of Personality and Social Psychology, 52 (3), 536-546. Kalleburg, A.L. & Moody, J.W. (1994). Human Resource Management and Organisational Performance. American Behaviourist Scientist, 37, 948 - 962. Kanter, R.M. (1983). The Change Masters: Corporate Entrepreneurs at work. New York: Simon & Schuster. Kiesler, S. & Sproull, L. (1982). Managerial responses to changing environments: perspectives in problem solving from social cognition. Administrative Science Quarterley, 27, 548 - 570. Kimble, C.E., Marsh, N.B. & Kiska, A.C. (1984). Sex, age and cultural differences in self-reported assertiveness. Psychological Reports, 55, 419-422.

King, N. (1990). Innovation at work: The research literature. In M. A. West & J. L. Farr (Eds), Innovation and Creativity at Work: Psychological and Organisational Strategies, pp. 15-59, Chichester: Wiley. King, N., Anderson, N.R. & West, M. (1991). Organizational innovation in the UK: A case study of perceptions and processes. Work and Stress, 5 (4), 331-339. King, R., Le Bas, J., & Spooner, D. (2000). The impact of caseload on the personal efficacy of mental health case managers. Psychiatric Services, 51, 364-368. Kinnunen, J. (1990). The importance of organizational culture on development activities in a primary health care organisation. International Journal of Health Planning and Management, 5, 65-71. Klimoski, R. & Mohammed, S. (1994). Team mental model: Construct or metaphor? Journal of Management, 20, 403-437. Komaki, J.L., Desselles, M.L. & Bowman, E.D. (1989). Definitely not a breeze: Extending an operant model of effective supervision to teams. Joumal of Applied Psychology, 74, 522-529. Koshuta, M. & McCuddy, M.K. (1989). Improving productivity in the health care industry: An argument and supporting evidence from one hospital. The Health Care Supervisor. LaFrance, M. & Mayo, C. (1978). Moving Bodies: Nonverbal Communication in Social Relationships. Monterey, C.A.: Brooks/Cole. Landsberger, H.A. (1955). Interaction process analysis of the mediation of labormanagement disputes. Journal of Abnormal and Social Psychology, 51, 522528. Latham, G.P., Erez, M. & Locke, E.A. (1988). Resolving scientific disputes by the joint design of crucial experiments by the antagonists: Application to the Erez Latham dispute regarding participation in goal setting. Journal of Applied Psychology, 73 (4), 753-772. Lawler, E.E. & Hackman, J.R. (1969). Impact of employee participation in development of pay incentive plans: A field experiment. Journal of Applied Psychology, 53, 467-471. Lawrence, P.R. & Lorsch, J. (1967). Organization and Environment. Cambridge, NIA:Harvard University Press. Macy, B.A. & lzumi, H. (1993). Organizational change, design and work innovation: A meta-analysis of 131 North American field studies-1961-1991. Research in Organizational Change and Design (Vol. 7). Greenwich, CT: JAI Press. Maier, N.R.F. (1963). Problem-solving Discussions and Conferences: Leadership Methods and Skills. New York: McGraw-Hill. Maier, N.R.F. (1970). Problem Solving and Creativity in Individuals and Groups. Monterey, CA: Brooks/Cole.

Maier, N.R.F. & Solem, A.R. (1962). Improving solutions by turning choice situations into problems. Personnel Psychology, 15, 151-157. Main, J. (1989). At last, software CE0s can use. Fortune, 13 March, 77 - 83. Markiewicz, L. & West, M.A. (1996). Team-based Organisation. Aberdeen: Grampian/ECITB. Markiewicz, L. & West, M.A. (1996). Team-based Organisation. Aberdeen: Grampian/ECITB. Mathison, D.L. & Tucker, R.K. (1982). Sex differences in assertive behaviour: A research extension. Psychological Reports, 51(3), 943-948. Maxwell, R.J. (1992). Dimensions of quality revisited: from thought to action. Quality Health Care 1, 171-177. Maznevski, M.L. (1994). Understanding our differences: Performance in decision making groups with diverse members. Human Relations, 47 (5), 531-552. McClure, L.M. (1984) Teamwork, myth or reality: community nurses effectiveness with general practice attachment. Journal of Epidemiology and Community Health, 21 (1), 68 - 74. McGrath, J.E. (1984). Groups: Interaction and Performance. Englewood Cliffs, NJ: Prentice-Hall. Miceli, M.P. & Near, J.P. (1985). Characteristics of organisational climate and perceived wrong-doing associated with whistle-blowing decisions. Personnel Psychology, 38, 525-544. Milliken, F.J. & Martins, L.L. (1996). Searching for common threads: Understanding the multiple effects of diversity in organizational groups. Academy of Management Review, 21(2), 402-433. Mistral, W., & Velleman, R. (1997). Community mental health teams: The professionals' choice? Journal of Mental Health, 6, 125-140. Mitrot, I.I. & Featheringham, T.R. (1974) On systematic problem solving and the error of the third kind. Behavioural Science, 19, 383 - 393. Mohman, S.A., Cohen, S.G. & Mohrman, A.M., Sr (1995). Designing Team-Based Organizations. San Francisco: Jossey-Bass. Mullarkey, S., Wall, T.D., Warr, P.B., Clegg, C.S. & Stride, C. (1999). Measures of job satisfaction, mental health and job-related well-being: A bench-marking manual. Sheffield, England: Sheffield Academic Press Ltd. Mumford, M.D. & Gustafson, S.B. (1988). Creativity syndrome: Integration, application and innovation. Psychological Bulletin, 103, 27-43.

Myer, C. (1993). How to Align Purpose, Strategy and Structure for Speed. New York: Free Press. Naylor, J.C., Pritchard, R.D., & Ilgen, D.R. (1980). A theory of behaviour in organisations. New York:Academic Press. Netten, A. & Dennett, J. (1997). Unit lists of health and social care. London: PSSRU. Nievaard, A.C. (1987). Communication climate and patient care: Causes and effects of nurses attitudes to patients. Social Science and Medicine, 24 (9), 777-784. Onyett, S. (1995). Responsibility and accountability in community mental health teams. Psychiatric Bulletin, 19, 281-285. Onyett, S. (1997). The challenge of managing community mental health teams. Health and Social Care in the Community, 5, 40-47. Onyett, S., Pillinger, T. & Muijen, M. (1995). Making community mental health teams work. London: Sainsbury Centre for Mental Health. Pearce, J.A. & Ravlin, E.C. (1987). The design and activation of self-regulating work groups. Human Relations, 40, 751-782. Peck, E., & Norman, I.J. (1999). Working together in adult community mental health services: Exploring inter-professional role relations. Journal of Mental Health, 8, 231-243. Peck, E., & Parker, E. (1998). Mental health in the NHS: Policy and practice 1979-98. Journal of Mental Health, 7, 241-259. Peiro, J.M., Gonzalez-Roma, V., & Romos, J. (1992) The infleince of work team climate on role stress, tension, satisfaction and leadership perceptions. European Review of Applied Psychology, 42 (1) 49-46. Pincus, H.A., Zarin, D.A. & West, J.C. (1996). Peering into the 'black box': measuring outcomes of managed care. Archives of General Psychiatry, 53, 870-877. Podsakolf, P.M. & Todor, W.D. (1985). Relationships between leader reward and punishment behavior and group processes and productivity. Journal of Management, 11, 55-73. Porac, J. F. & Howard, H. (1990). Taxonomic mental models in competitor definition. Academy of Management Review, 2, 224-240. Poulton, B. C. & West, M. A. (1993). Effective multidisciplinary teamwork in primary health care. Journal of Advanced Nursing, 18, 918-925. Poulton, B. C. & West, M. A. (1994). Primary health care team effectiveness: Developing a constituency approach. Health and Social Care, 2, 77-84. Poulton, B.C. & West, M.A. (1997). A failure of function: teamwork in primary health care. Journal of Interprofessional Care, 11, No 2, 1997.

Pritchard, R. D., Jones, S. D., Roth, P. L., Stuebing, K. K. & Ekeberg, S. E. (1988). Effects of group feedback, goal setting, and incentives on organizational productivity. Journal of Applied Psychology, 73, 337-358. Pritchard, R.D. (ed) (1995) Productivity Measurement and Improvement: Organisational Case Studies. New York: Praeger. Pritchard, R.D., Jones, S.D., Roth, P., Stuebing, K.K., & Ekeberg, S.E. (1989). The evaluation of an integrated approach to measuring organisational productivity. Personnel Psychology, 42 (1) 69-115. Pritchard, R.D. (Ed) (1995). Productivity Measurement and Improvement: Organisational Care Studies: New York: Praeger. Pritchard, R.D., Jones, S.D., Roth, P.L., Stuebing, K.K. & Ekeberg, S.E. (1988). The effects of feedback, goal setting, and incentives on organisational productivity. Journal of Applied Psychology Monograph Series, 73 (2), 337-358. Prosser, D., Johnson, S., Kuipers, E., Szmukler, G., Bebbington, P., & Thornicroft, G. (1996). Mental Health, 'Burnout' and job satisfaction among hospital and communitybased mental health staff. British Journal of Psychiatry, 169, 334-337. Quinn, R.E. & Rohrbaugh, J. (1983). Predicting sales success through handwriting analysis: an evaluation of the effects of training and handwriting sample content. Journal of Applied Psychology 68 (2), 212 - 17. Redmond, M.V. (1989). The functions of empathy (decentering) in human relations. Human Relations, 42, 593-605. Redmond. M.V. (1992). A multi-dimensional theory and measure of decentering. Unpublished manuscript. Reed J. (1995). Leadership in the mental health service: What role for doctors? Psychiatric Bulletin, 19, 67-72. Rees, A., Stride, C.B., Shapiro, D.A., Richards, A. and Borrill, C.S. (in press). Psychometric properties of the Community Health Team Questionnaire. (CMHTEQ) Journal of Mental Health. Richards, A. & Rees, A. (1998) Developing Criteria to measure the effectiveness of community mental health teams. Mental Health Care, 2, 14 - 17. Roberts, J.M. (1995) The History of the World. Hammondsworth, Middlesex: Penguin. Ross, F., Rink, E. & Furne, A. (2000). Integration or pragmatic coalition? An evaluation of nursing teams in primary care. Journal of Interprofessional Care, 14 (3), 259-267. Rousseau, D.M. (1985) Issues of level in organisational research: multi-level and cross-level perspectives. Research in Organisational Behaviour, 7, 1 - 37. Sainsbury Centre for Mental Health (1997). Pulling together: The future roles and training of mental health staff. London: Sainsbury Centre for Mental Health.

Schober, M.F., & Clark, H.H. (1989). Understanding by addressees and overhearers. Cognitive Psychology, 21, 211-232. Schwenk, C.R. (1988). The Essence of Strategic Decision-making. Cambridge, NIA: Heath. Senge, P. (1990). The Fifth Discipline: The Art and Practice of the Learning Organization. New York: Doubleday Currency. Shaw, M.E. (1976). Group Dynamics: The Psychology of Small Group Behavior. New York: McGraw-Hill. Shaw, M.E. (1981). Group Dynamics: The Psychology of Small Group Behavior. New York: McGraw-Hill. Sluyter, G.V. (1995). Mental health leadership training: A survey of state directors. Journal of Mental Health Administration, 22, 201-204. Smircich, L. (1983). Organization as shared meaning. In L.R. Pondy, P. Frost, G. Morgan & T. Dandridge (Eds), Organizational Symbolism (pp. 55-65). Greenwich, CT: JAI Press. Smith, K.G., Locke, E.A. & Barry, D. (1990). Goal setting, planning and organizational performance: An experimental simulation. Organizational Behavior and Human Decision Processes, 46, 118-134. Sommers, L.S., Marton, K.I., Barbaccia, J.C. & Randolph, J. (2000). Physician, nurse and social worker collaboration in primary care for chronically ill seniors, Archives of Internal Medicine, 160, 1825-1833. Stein, M. (1996). Unconscious phenomena in work groups. In M.A. West (Ed.), Handbook of Work Group Psychology, pp. 143-157 Chichester: Wiley. Stemberg, R.J. & Lubart, T.I. (1990) Defying the Crowd. Cultivating Creativity in a Culture of Conformity. New York: Free Press. Stevens, M.J., & Campion, M.A. (1994). The knowledge, skill and ability requirements for teamwork: Implications for human resource management. Journal of Management, 20, 503-530. Stevens, M.J. & Campion, M.A. (1999). Staffing Work Teams: Development and Validation of a Selection Test for Teamwork Settings. Journal of Management, 25, No 2 207 - 228. Sundstrom, E., De Meuse, K.P. & Futrell, D. (1990). Work teams: Applications and effectiveness. American Psychologist, 45, 120-133. Tajfel, H. (1978). Differentiation Between Social Groups: Studies in the Social Psychology of Intergroup Relations (European Monographs in Social Psychology, No. 14). London:Academic Press. Tajfel, H. & Turner, J.C. (1979). An integrative theory of intergroup conflict. In W. G. Austin and S. Worchel (Eds), The Social Psychology of Intergroup Relations. Monterey, CA: Brooks/Cole.

Tannenbaum, S.I., Beard, R.L. & Salas, E. (1992). Team building and its influence on team effectiveness: An examination of conceptual and empirical developments. In K. Kelley (Ed.), Issues, Theory and Research in Industrial/Organizational Psychology (pp. 117-153), London: North Holland. Tannenbaum, S.I., Salas, E. & Cannon-Bowers, J.A. (1996). Promoting team effectiveness. In M.A. West (Ed.), Handbook of Work Group Psychology, (pp 503529) Chichester: Wiley. Taylor, M.F. (Ed), with Brice, J., Buck, N. & Prentice, E. (1995). British Household Panel Survey User Manual. Colchester: University of Essex. Tjosvold, D. (1982). Effects of approach to controversy on superiors' incorporation of subordinates' information in decision making. Joumal of Applied Psychology, 67, 189-193. Tjosvold, D. (1985). Implications of controversy research for management. Journal of Management, 11, 21-37. Tjosvold, D. (1991). Team Organisation: An Enduring Competitive Advantage. Chichester: Wiley. Tjosvold, D. & Field, R.H.G. (1983). Effects of social context on consensus and majority vote decision making. Academy of Management Journal, 26, 500-506. Tjosvold, D. & Johnson, D.W. (1977). The effects of controversy on cognitive perspective-taking. Journal of Educational Psychology, 69, 679-685. Tjosvold, D., Wedley, W.C. & Field, R.H.G. (1986). Constructive controversy, the Vroom-Yetton model, and managerial decision making. Journal of Occupational Behavior, 7, 121 -138. Toon, P.D. (1994) What is Good General Practice? A Philosophical Study of the Concept of High Quality Medical Care. Tziner, A.E. (1988) Effects of team composition on ranked team effectiveness. Small Group Behaviour, 19, 363 - 378. Vroom, V.H. (1964). Work and Motivation. New York: Wiley. Wall, T.D. & Lischeron, J.H. (1977). Worker Participation: A Critique of the Literature and some Fresh Evidence. Maidenhead, UK: McGraw-Hill. Walsh, J.P. & Fahey, L. (1986). The role of negotiated belief structures in strategy making. Journal of Management, 12, 325-338. Walsh, J.P., Henderson, C.M. & Deighton, J. (1988). Negotiated belief structures and decision performance: An empirical investigation. Organizational Behavior and Human Decision Processes, 42, 194-216. Watson, W.E., Kumar, K. & Michaelsen, L.K. (1993). Cultural diversity's impact on interaction process and performance: Comparing homogeneous and diverse task

groups. Academy of Management Journal, 36, 590-602. Weldon, E. & Weingart, L.R. (1993). Group goals and group performance. British Journal of Social Psychology, 32, 307-334. West, M.A. (1990). The social psychology of innovation in groups. In M.A. West & J.L. Farr (Eds), Innovation and Creativity at Work (pp. 309-333). Chichester: Wiley. West, M.A. (1994). Effective Teamwork. Leicester: British Psychological Society. West, M.A. (1996a). The Handbook of Work Group Psychology. Chichester: Wiley. West, M.A. (1996b). Reflexivity and Work Group Effectiveness: A conceptual Integration in M.A. West (Ed) Handbook of Work Group Psychology (555 - 580), John Wiley & Sons Ltd. West, M.A. (1997). Developing Creativity in Organisations. Chichester: Wiley. West, M.A. (2000). State of the art: Creativity and Innovation at work. Psychologist, 13, 9 460 - 464. West, M.A. & Anderson, N.R. (1996). Innovation in top management teams. Journal of Applied Psychology, 81(6), 680-693. West, M.A., Borrill, C.S. and Stride, C.B. (1998). Strain as a moderator of the relationship between work characteristics and work attitudes. Journal of Occupational Health. Vol 4, No 1, 3 - 14. West, M.A., Borrill, C.S. & Unsworth, K. (1998). Team Effectiveness in Organisations. In C.L. Cooper & I.T. Robertson (eds) International Review of Industrial Organisational Psychology. Vol 13. Wiley & Sons: Chichester. West, M.A. & Field, R. (1995). Teamwork in primary health care. Perspectives from organisational psychology. Journal of Interprofessional Care, 9, 2, 117-122. West, M.A., & Poulton, B.C. (1995). Primary health care teams: Rhetoric versus reality. Paper submitted for publication. Institute of Work Psychology, University of Sheffield. West, M.A. & Slater,J.A. (1996). The Effectiveness of Team Working in Primary Health Care. London:Health Education Authority. West, M.A. & Wallace, M. (1991). Innovation in health care teams. European Journal of Social Psychology. Vol 21, 303 315. West, M.A. & Pillinger, T. (1995). Innovation in UK manufacturing (Research report). Institute of Work Psychology, University of Sheffield. Wood, N., Farrow, S., & Elliott, B. (1994). A review of primary health care organisation. Journal of Clinical Nursing, 3(4), 243-250.

Worchell, S., Wood, W. & Simpson, J.A. (eds) (1992). Group Processes and Productivity. Newbury Park, CA: Sage. Yeatts, D.E. & Seward, R.R. (2000). Reducing turnover and improving health care in nursing homes: The potential effects of self-managed work teams. The Gerontologist, 40 (3), 358-363. Zimmer, J.G., Eggert, G.M. & Chiverton, P. (1990). Individual versus team case management in optimising care for chronically ill patients with dementia. Journal of Aging and Health, 2 (3), 357-372.