S171

Factors Affecting Adherence to Antiretroviral Therapy

Margaret A. Chesney From the School of Medicine,
University of California San Francisco
In both clinical trials and clinical practice, nonadherence to medications is widespread
among patients with chronic diseases. The shift to combination therapies for treating human
immunodeciency virus (HIV)infected individuals has increased adherence challenges for
both patients and health-care providers. Estimates of average rates of nonadherence to an-
tiretroviral therapy range from 50% to 70%. Adherence rates of !80% are associated with
detectable viremia in a majority of patients. The principal factors associated with nonadher-
ence appear to be patient-related, including substance and alcohol abuse. However, other
factors may also contribute, such as inconvenient dosing frequency, dietary restrictions, pill
burden, and side effects; patienthealth-care provider relationships; and the system of care.
We discuss the major reasons reported by HIV-infected individuals for not taking their med-
ications. Improving adherence probably requires clarifying the treatment regimenandtailoring
it to patient lifestyles.
Measurement of Adherence
The shift to the use of highly active antiretroviral therapy
(HAART) for treating human immunodeciency virus (HIV)
disease has led to increasingly complex drug regimens. These
present signicant challenges to both patients and health-care
providers with respect to adherence. Without adequate adher-
ence, antiretroviral agents are not maintained at sufcient con-
centrations to suppress HIV replication in infected cells and to
lower the plasma viral load. In addition to being associated
with poor short-term virological response, poor adherence to
antiviral medication accelerates development of drug-resistant
HIV. Therefore, identifying and overcoming the factors that
reduce adherence to combination antiretroviral agents is of ut-
most importance for prolonged viral load suppression.
There are a number of key issues in the study of adherence
to antiretroviral therapy, including accurate measurement of
adherence, assessment of the impact of adherence on viral load
and clinical outcome, determination of the factors that affect
adherence, and the development of interventions. Addressing
these issues may provide valuable information about which
patients are most at risk for nonadherence and about how
adherence might be improved. The critical factors that inuence
adherence fall into 4 main groups: (1) patient factors, such as
drug use, alcohol use, age, sex, or ethnicity; (2) medication
regimen, such as dosing complexity, number of pills, or food
Financial support for this research was provided by National Institutes
of Health grants to the Center for AIDS Prevention Studies (MH42459)
and the Center for AIDS Research (MH59037).
Reprints or correspondence: Dr. Margaret A. Chesney, University of Cal-
ifornia San Francisco, Center for AIDS Prevention Studies, 74 New Mont-
gomery St., Suite 600, San Francisco, CA 94105 (mchesney@psg.ucsf.edu).
Clinical Infectious Diseases 2000; 30(Suppl 2):S1716
2000 by the Infectious Diseases Society of America. All rights reserved.
1058-4838/2000/3006S2-0010$03.00
requirements; (3) the patienthealth-care provider relationship;
and (4) the system of care.
Adherence to therapy is difcult to measure accurately. Four
basic techniques have been developed for quantifying adher-
ence, all of which have limitations. First and most common
are patient self-reports. These have the advantages of low cost
and exibility of design (questionnaires suit individual language
abilities). The data are easily collected and can help to deter-
mine the reasons why patients are nonadherent. They assume,
however, that patients can accurately recall their behavior and
are providing honest answers. A major limitation of self-reports
is that they reect only short-term or average adherence and
may often overestimate it. Nevertheless, some studies show sig-
nicant relationships between data from self-reports and viral
load [1, 2]. Other studies that compare data from self-reports
to pill counts or electronic measurements found differences,
suggesting that self-reports provide inated estimates of ad-
herence behavior [3, 4].
Second are patients reports of missing pills, which are almost
always reliable [5], so self-reports can be helpful for under-
standing the dynamics surrounding missed medication. Pill
counts have been widely used. The return of excess pills pro-
vides tangible evidence of nonadherence. However, pill counts
require patients to return the medication packaging to the cli-
nician. Even in clinical trial situations, patients tend to forget
the packages or inadvertently discard them. There have also
been reports that patients other than those with HIV, aware
that pill counts are being conducted, engage in pill dumping
to appear adherent. As a result, pill counts typically overesti-
mate adherence.
Third, assays of drug levels have been used in clinical trials
to measure the last dose taken; however, these assays are often
impractical because of their expense and lack of general avail-
ability. In addition, serum concentrations of nucleoside ana-
logues may not reect intracellular concentration of the active

a
t

U
n
i
v
e
r
s
i
t
a
s

D
i
p
o
n
e
g
o
r
o

o
n

J
u
l
y

3
1
,

2
0
1
3
h
t
t
p
:
/
/
c
i
d
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m

S172 Chesney CID 2000;30 (Suppl 2)
Table 1. Poor medication adherence correlates with virologic failure:
a study of 45 HIV-infected patients.
Medication adherence, % of patients Virologic failure, % of patients
!80 87
8090 47
195 10
NOTE. Seventy-two percent of patients were treated with a single protease
inhibitor (nelnavir, 65%; indinavir, 26%; other inhibitors, 9%); 28% received a
combination of saquinavir and ritonavir. Adapted from [6].
Table 2. Factors reported to negatively affect adherence in HIV-
infected patients.
Patient factors
Active substance abuse (drugs and/or alcohol)
Male sex
Youth
Active depression
Lower level of education
Lack of self-efcacy
a
Extreme anxiety
Extreme pain
No change in health status
Non-white
Medication factors
Dose frequency of more than twice a day
Pill burden
Type of drug
Inability to take medication when away from home
Food requirement
Side effects
Poor doctorhealth-care provider relationship
System of care
Dissatisfaction with past experience of health-care system, leading
to avoidance
NOTE. Based on data from [2, 68, 12, 13, 1519].
a
Self-efcacy refers to patients belief in their ability to take medication
as prescribed.
triphosphates. Furthermore, these assays typically measure
only recent doses and thus provide limited data. Adherence
may be overestimated if patients are more conscientious about
taking their medication before a clinic visit.
Fourth, electronic monitoring systems, such as the Medi-
cation Event Monitoring System (MEMS), are inserted into
medication bottle caps; they contain a computer chip that re-
cords the date and time of opening and closing of the bottle.
Interpretation of these data assumes that a single dose is taken
each time the bottle is opened, and may lead to inaccuracies if
multiple doses are removed at once. Despite the limitations of
these measurement techniques, adherence data are providing
valuable insight into the association between drug taking and
viral load, as well as approaches that may be useful for im-
proving adherence.
Adherence to HIV Therapy
Large-scale studies of the impact of adherence on viral load
and clinical outcome in HIV therapy are underway within the
AIDS Clinical Trials Group (ACTG) and at clinical sites in the
United States and abroad. However, ndings have not yet been
reported. A small study by Paterson et al. [6] examined patient
adherence to protease inhibitor therapy by use of MEMS. The
investigators found that poor adherence correlated with clinical
and virological failure at 3 months of follow-up (table 1). These
results suggest that a high degree of adherence is necessary for
maintenance of drug efcacy. Further similar studies are needed
to corroborate these data and to establish whether adherence
requirements vary for different drugs.
Although very little published information is available on
medication adherence of HIV-infected patients, new data from
a number of studies were presented at numerous conferences,
including the 12th World AIDS Conference in Geneva and the
38th Annual Interscience Conference on Antimicrobial Agents
and Chemotherapy (ICAAC) in San Diego. Because the study
of adherence is in its infancy, study designs and end points vary
widely, making study comparisons difcult. Self-reported ad-
herence, as dened in research studies, has been reported to
range from 0% to 100% [613]. Although the results of some
of these studies appear to conict, important information is
emerging about the extent of and factors associated with ad-
herence. The earliest reports of nonadherence suggested that
slightly 110% of patients missed 1 dose of medication each
day [1, 14]. Rates of nonadherence may be as high as 50% when
averaged over time and with an arbitrary cutoff point of !80%
of medication taken [8]. This gure is supported by observations
from 2 larger studies: an international multicenter study of 235
HIV-infected patients and a United States study of 244 HIV-
infected Medicaid-insured patients. The reported rates of pa-
tients who take !80% of doses in these studies were 46% and
40%, respectively [9, 10]. It is interesting to note that these
results are consistent with estimates of adherence in other
chronic diseases and support the view that nonadherence is a
common behavior that should be expected, even with a serious
disease such as HIV infection.
In order to implement measures to improve adherence, it is
rst essential to identify the principal factors that contribute
to the inability of patients to take their medication. Those fac-
tors identied to date are summarized in table 2.
A published study of adherence by Eldred et al. [10] required
244 patients who were receiving antiretroviral therapy and
Pneumocystis carinii pneumonia (PCP) prophylaxis to report
the medications they were taking, their pattern of use, and their
knowledge and attitudes about HIV therapies [10]. Most of the
patients in the study were receiving monotherapy. Eldred et al.
found that 60% of patients reported 180% adherence to treat-
ment regimens in the previous 7 days. When PCP prophylaxis
was analyzed by assaying urinary sulfamethoxazole levels, a
correlation of 80% was found between self-report and levels
of sulfamethoxazole in the urine. This leads to the conclusion
that at least 40% of patients were getting !80% of their antiviral
medication. Good adherence was associated with dosing twice
a day or less, the likelihood that patients take medication when
away from home, and self-efcacy (i.e., patients belief in their

a
t

U
n
i
v
e
r
s
i
t
a
s

D
i
p
o
n
e
g
o
r
o

o
n

J
u
l
y

3
1
,

2
0
1
3
h
t
t
p
:
/
/
c
i
d
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m

CID 2000;30 (Suppl 2) Adherence to Antiretroviral Therapy S173
Table 3. Frequent causes for medication nonadherence by HIV-
infected patients who are receiving highly active antiretroviral therapy.
Reason for nonadherence Prevalence, % of patients
Forgot or busy 3452
Away from home 2742
Change in daily routine 40
Side effects 1925
Depression or illness 923
Lack of interest or drug holidays 419
NOTE. Based on data from [4, 11, 19].
ability to take their medication). The major factor associated
with nonadherence was active illicit drug use. In particular,
patients who had used crack cocaine were signicantly less ad-
herent. These ndings, although not directly applicable to com-
bination therapy, may shed light on adherence issues for pa-
tients who are receiving HAART.
Preliminary data from a number of studies of adherence to
antiretroviral agents have been presented in abstract form [4,
6, 12, 15, 1922]. Specic ndings vary from study to study,
due at least in part to the different samples and measures used.
In general, it appears that the most important factors that affect
adherence are patient-related. The most common reasons given
by patients for nonadherence are summarized in table 3; the
principal reasons are that they forgot or were busy. Data from
a large cohort of HIV-infected patients (1322 persons living
with HIV/AIDS) who were receiving antiviral combination
therapies demonstrated that characteristics that predict non-
adherence included youth, government-subsidized health in-
surance, extreme anxiety or pain, and no perceived change in
health status as a result of drug therapy [15]. These data suggest
that vulnerable subgroups of HIV-infected patients need to be
identied and targeted [15].
Reduction of the administration frequency of current drug
regimens may be limited by the relatively short half-lives of the
nonnucleoside reverse transcriptase inhibitors (NNRTIs) and
protease inhibitors. Although it has been assumed that reducing
dosing frequency or pill burden will increase adherence, it is
important to note that this may not be the case. In the above-
mentioned trial of 179 patients and in another smaller trial of
45 patients, there was no association between dose frequency
or pill burden and nonadherence [6, 10]. There are indications
from a number of other studies that meal restrictions and other
factors, and not simply pill burden, are predictors of nonad-
herence [2, 4, 10, 13, 19]. It is worth noting that the pharma-
ceutical industry is investing considerable effort to develop new
compounds or regimens with longer half-lives in order to reduce
pill count and dosing frequency and is also trying to nd other
ways to simplify drug regimens. The simplication of existing
regimens may prove worthwhile as therapies continue to fail
in nonadherent patients and even more medications are added
to regimens.
A good patienthealth-care provider relationship may be an
important motivating factor for taking and adhering to com-
plex combination drug therapies [20]. A qualitative study of
homosexual youths showed that primary-care providers exhib-
ited judgmental behavior, stereotyping, homophobia, and fail-
ure to address cultural issues when administering care [17]. Such
experiences are likely to lead some people with HIV infection
to avoid the health care system. On the other hand, factors
that have been identied as strengthening patienthealth-care
provider relationships include perceptions of health-care pro-
vider competence, communication quality and clarity, com-
passion, willingness to include patients in treatment decisions,
adequacy of referrals, and convenience of visiting the doctor
[21]. Conversely, frustration for health-care providers is asso-
ciated with lack of patient adherence to treatment, miscom-
munication, missed appointments, complexity of treatment reg-
imens, and medication side effects [21, 23]. In light of these
problems, it is heartening to nd that initiatives are underway
to encourage health-care providers to work with patients as
partners in care and to involve representatives fromthe entire
HIV community [24].
Other important adherence issues have arisen, particularly
with regard to economically disadvantaged patients with mul-
tiple social problems. Many clinicians feel that lifestyle factors,
such as homelessness, substance abuse, lack of education, and
mental illness, are predictors of nonadherence and therefore are
withholding HAART from these patients. A review of these
studies indicates that some, but not all, have found an asso-
ciation between nonadherence and youth, female sex, less edu-
cation, or a current or past history of substance abuse [25].
Therefore, health-care providers should be cautious in making
assumptions about patients likelihood to encounter problems.
It is of interest that results of trials show that health-care pro-
viders are not very good at predicting which patients will be
adherent to medication [6, 22, 26]. For example, a comparison
of health-care providers opinions and self-reports from 193
HIV-infected
patients revealed that the health-care providers overestimated
the inuence of social factors on adherence [22]. In this study,
social factors made no signicant difference. Results of another
study of 45 patients indicated that health-care providers pre-
dicted adherence of patients poorly: 33% of nonadherent pa-
tients were identied by health-care providers as adherent, and
36% of patients with better than 95% adherence (measured by
MEMS Caps) were identied as being poorly adherent [3]. In
addition, the results of a recent study of 31 HIV-infected youths
found that homelessness, current living situation, years of ed-
ucation, clinical depression, and substance abuse did not predict
adherence to combination therapy [26]. These data do not sup-
port the routine withholding of HAART from specic social
groups, since adherence cannot reliably be predicted on the
basis of patient characteristics [27].
Special issues with adherence exist for HIV-infected children
and adolescents. Infants and young children are dependent on
adults for administration of their medications, which means

a
t

U
n
i
v
e
r
s
i
t
a
s

D
i
p
o
n
e
g
o
r
o

o
n

J
u
l
y

3
1
,

2
0
1
3
h
t
t
p
:
/
/
c
i
d
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m

S174 Chesney CID 2000;30 (Suppl 2)
Table 4. Strategies for enhancing adherence to antiretroviral therapy.
Strategy Remarks
Directly observe therapy Ensures adherence
Labor intensive
Expensive
Encourage family member to observe
Clarify instructions using personal treatment plan Helps patients to be better organized
Helps patients understand dosing schedules
Tailor drug regimen to suit patients lifestyles Identify daily activities that patients can link with taking
medication
Show patients how to keep medication diary Helps establish a routine
Encourage pill sorting and convenient placement
of pills
Helps establish a routine of thinking about the practicalities
of taking medication
Encourage planning ahead for changes in daily
routine (weekends and holidays)
Helps patients to avoid missing doses
Make clinic appointments convenient and pleas-
ant, provide educational materials, etc.
Encourages patients to keep appointments
Promotes self-efcacy
a
Encourages patients to learn about medications
Refer patients for social support, treatment of
substance abuse, and for stress
Counseling may help patients cope with drug and/or alcohol
abuse or with medication-related food requirements
Social support promotes self-efcacy
a
a
Self-efcacy refers to patients beliefs in their ability to take medication as prescribed.
that their adherence is only as good as that which their care-
givers are able to achieve. Unfortunately, liquid formulations
are often not particularly palatable, and food requirements for
some antiretroviral agents make therapies difcult toadminister
to infants who require frequent formula feeding. These factors
can affect the willingness of the caregiver to administer the
medication and the willingness of the child to take it. Another
barrier to adherence for children and adolescents with HIV
may be their families desire for secrecy about the condition.
For example, parents may be unwilling to ll prescriptions at
local pharmacies and/or may send their child to school without
their medication to hide the fact that the child is HIV-infected.
Adolescents nd adherence particularly challenging as they en-
ter a stage of life when they are particularly self-conscious and
do not want to be different from their peers. A detailed as-
sessment of the barriers to adherence should be implemented
for all minors who require antiretroviral therapy. Case man-
agers and counselors may often be able to work with families
to resolve specic issues.
The Way Ahead
A variety of methods for improving adherence have been
suggested, and descriptions from clinicians, pilot studies, and
1 trial [28] provide encouraging positive results. A study by
Workman et al. [18] describes a practice that implemented a
number of patient management interventions aimed at im-
proving adherence. These included extended consultation time
to explain and to reinforce medication instructions, tailoring
the choice of drug regimen to patient lifestyle, frequent follow-
up when initiating or changing drug regimens, rapid viral load
feedback, and the use of reminder calls and alarms [18]. Ad-
herence rates of 80%100% (with 73% reporting 100% adher-
ence) were reported by 99% of 77 HIV-infected patients [18].
The authors concluded that the interventions signicantly en-
hanced adherence, compared with mean adherence rates re-
ported in the literature [6, 810]. The effectiveness of an alarm
set to alert patients when their drugs should be taken was tested
in another study of 49 HIV-infected patients [29]. After 3
months, the frequency of 100% pill adherence was 89%; 80%
adherence was observed in 99% of patients. The patients also
exhibited good adherence to timing and diet requirements. All
the patients were enthusiastic about the alarm device; however,
concerns were raised as to its durability, since malfunctions
occurred in the devices issued to half of the patients. Managed
social support and perceived health-care provider support for
promoting adherence have also been investigated. In 2 separate
studies, patient support systems were found to enhance adher-
ence and were particularly helpful with regard to enabling pa-
tients to better follow advice and instructions [30, 31].
As the previous review demonstrates, the major factors as-
sociated with nonadherence are related to patient behavior, and
a variety of strategies by health care professionals are beginning
to yield improved adherence. Some of the interventions that
have been studied are listed in table 4. Of particular importance
is the tailoring of medication to the patients lifestyle [18, 25,
28]. This can be illustrated by consideration of dosing require-
ments and the need for special instructions that take into ac-
count the life patterns of patients. For some patients, certain
dosing intervals might be easier to adhere to; for others, lack
of ready access to refrigeration or water may prove problematic.
Moreover, if a patients treatment fails, the patient should be
approached to review adherence before the regimen is changed.
This should help ensure that patients begin a new regimen with
optimal adherence.
Future clinical trials that study safety and efcacy of anti-
retroviral agents with respect to virologic and immunologic
endpoints may be considered decient if they do not include

a
t

U
n
i
v
e
r
s
i
t
a
s

D
i
p
o
n
e
g
o
r
o

o
n

J
u
l
y

3
1
,

2
0
1
3
h
t
t
p
:
/
/
c
i
d
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m

CID 2000;30 (Suppl 2) Adherence to Antiretroviral Therapy S175
at least 1 acceptable measure of adherence to therapy. Indeed,
simultaneous measurement of adherence by 2 different methods
may be the best approach. In addition, there is a need for studies
that directly assess strategies that are designed to increase
adherence.
Finally, important lessons can be drawn from studies of ad-
herence among patients with other chronic diseases, where there
is a larger body of published literature. Both the factors related
to adherence and many of the interventions to improve ad-
herence to other medications may largely overlap with those
important to antiretrovirals [27]. For example, a recent review
of hypertension stated the following: Once-daily dosing should
be coupled with selection of a drug with long duration of action
to overcome problems of missed doses. Widespread adoption
of simple compliance enhancement methods could lead to de-
creased morbidity and mortality [32]. Incomplete compliance
with any long-term medication is a multifactorial problem, but
it is a problem that can and must be addressed.
Summary
It is important to recognize that some degree of nonadher-
ence is common and should be expected in all patients who are
receiving antiretroviral therapy. The rst step toward addressing
the problem of medication nonadherence is to accurately iden-
tify patients whose risk of nonadherence is sufcient to un-
dermine clinical outcomes. However, a number of studies have
demonstrated that health-care providers cannot accurately
identify those patients likely to be nonadherent. Studies of large
samples suggest that substance abuse is associated with non-
adherence. Therefore, health-care providers may want to be
careful to ask patients with histories of substance abuse about
adherence. It does not follow, however, that all patients who
abuse substances will be nonadherent. Similarly, homelessness,
lack of education, and mental illness are not necessarily pre-
dictors of nonadherence, but might warrant extra attention and
support. Furthermore, the absence of alcohol and drug abuse
does not predict good adherence. Steps to maximize adherence,
therefore, should be reviewed with all patients. Although self-
reports tend to overestimate adherence, they are inexpensive
and fairly accurate for providing an indication of problems. In
particular, reports of nonadherence are reliable and call for
action. Such reports also help to determine why HIV-infected
patients are nonadherent. Of the available methods, self-reports
are the most practical for routine use in the clinic. Once non-
adherent patients are identied, health-care providers may want
to implement a variety of interventions to enhance adherence.
Strategies for intervention are likely to be based on tailoring
the drug regimen to the lifestyle of the patient and assessing
adherence as part of a follow-up program.
References
1. Hecht FM, Colfax G, Swanson M, Chesney M. Adherence and effectiveness
of protease inhibitors in clinical practice [abstract 15]. In: Program and
abstracts of the 5th Conference on Retroviruses and Opportunistic In-
fections (Chicago). Alexandria, VA: Foundation for Retrovirology and
Human Health, 1998.
2. Niewkerk P, Gisolf EH, Van Leeuwen R, Danner SA, DeBoer JB, Sprangers
MA. Self-reported adherence to ritonavir/saquinavir and ritonavir/saqui-
navir/stavudine in a randomized clinical trial: preliminary results [abstract
32362]. In: Program and abstracts of the 12th World AIDS Conference.
Geneva: Marathon Multimedia, 1998.
3. Melbourne K, Geletko S, Brown S, Willey C, Chase S, Fisher A. Electronic
adherence assessment versus self-report in HIV-infected individuals [ab-
stract I-175]. In: Program and abstracts of the 38th Interscience Confer-
ence on Antimicrobial Agents and Chemotherapy (San Diego). Washing-
ton, DC: American Society for Microbiology, 1998:420.
4. Hales G, Mitchell J, Smith DE, Kippex S. Validity of patient questioning
versus pill count as an assessment of compliance [abstract 32373]. In:
Program and abstracts of the 12th World AIDS Conference. Geneva:
Marathon Multimedia, 1998.
5. Berg J, Dunbar-Jacobs J, Rohay JM. Compliance with inhaled medications:
the relationship between diary and electronic monitoring. Ann Behav Med
1998; 20:368.
6. Paterson DL, Swindels S, Mohr JA, et al. Adherence with protease inhibitor
therapy for human immunodeciency virus infection [abstract I-172]. In:
Program and abstracts of the 38th Interscience Conference on Antimi-
crobial Agents and Chemotherapy (San Diego). Washington, DC: Amer-
ican Society for Microbiology, 1998:419.
7. Williams A, Wolf H, Yu C, Singh M. Adherence to antiretroviral therapy
among HIV positive women [abstract 32374]. In: Program and abstracts
of the 12th World AIDS Conference. Geneva: Marathon Multimedia,
1998.
8. Bennett M, Indyk D, Golub S. Adherence re-framed in the BIG picture: a
qualitative ecological perspective on HIV

patients and protease inhibitors

[abstract 32365]. In: Program and abstracts of the 12th World AIDS
Conference. Geneva: Marathon Multimedia, 1998.
9. Gir E, Pratt R, Bunch EH, Holzemer WL. Adherence to anti-retroviral ther-
apy: a four country comparison [abstract 60129]. In: Program and ab-
stracts of the 12th World AIDS Conference. Geneva: Marathon Multi-
media, 1998.
10. Eldred LJ, Wu AW, Chaisson RE, Moore RD. Adherence to antiretroviral
and pneumocystis prophylaxis in HIV disease. J Acquir Immune Dec
Syndr Hum Retrovirol 1998; 18:11725.
11. Tuldra A, Ferrer MJ, Rodriguez C, Bayes R, Burger D, Clotet B. A proposed
model to predict compliance with antiretroviral therapy [abstract I-177].
In: Program and abstracts of the 38th Interscience Conference on Anti-
microbial Agents and Chemotherapy (San Diego). Washington, DC:
American Society for Microbiology, 1998:421.
12. Walsh JC, Dalton M, Gill J, Burgess AP, Gazzard BG. Adherence to highly
active antiretroviral therapy (HAART) [abstract I-171]. In: Program and
Abstracts of the 38th Interscience Conference on Antimicrobial Agents
and Chemotherapy (San Diego). Washington, DC: American Society for
Microbiology, 1998:419.
13. Stewart KE, Greene P, Call S, et al. Patterns of self-reported adherence to
ART in a prospective clinical cohort [abstract. I-176]. In: Program and
abstracts of the 38th Interscience Conference on Antimicrobial Agents
and Chemotherapy (San Diego). Washington, DC: American Society for
Microbiology, 1998:420.
14. Ickovics J, Chesney M. Issues regarding anti-retroviral treatment for patients
with HIV-1 infection. JAMA 1997; 278:12334.
15. Klosinski LE, Brooks RN. Predictors of non-adherence to HIV combination
therapies [abstract 32375]. In: Program and abstracts of the 12th World
AIDS Conference. Geneva: Marathon Multimedia, 1998.

a
t

U
n
i
v
e
r
s
i
t
a
s

D
i
p
o
n
e
g
o
r
o

o
n

J
u
l
y

3
1
,

2
0
1
3
h
t
t
p
:
/
/
c
i
d
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m

S176 Chesney CID 2000;30 (Suppl 2)
16. Haubrich R, Little S, Dube M, Forthal D, Beall G, Kemper L. Self reported
treatment adherence and drug/alcohol use are associated with virologic
outcomes in CCTG 570: a clinical strategy trial of HIV RNA and anti-
retroviral monitoring [abstract 32379]. In: Program and abstracts of the
12th World AIDS Conference. Geneva: Marathon Multimedia, 1998.
17. Schilder AJ, Buchner C, Hogg RS. Poor experiences of health care leads to
avoidance by HIV-positive gay men in youth, ages 18 to 27, and can
inuence access and adherence [abstract 60308]. In: Programand abstracts
of the 12th World AIDS Conference. Geneva: Marathon Multimedia,
1998.
18. Workman C, Musson C, Workman C. Impact of interventions on adherence
rates [abstract 60588]. In: Program and abstracts of the 12th World AIDS
Conference. Geneva: Marathon Multimedia, 1998.
19. Gifford AL, Shively MJ, Bormann JE, Timberlake D, Bozzette SA. Self-
reported adherence to combination antiretroviral medication regimens in
a community-based sample of HIV-infected adults [abstract 32338]. In:
Program and abstracts of the 12th World AIDS Conference. Geneva:
Marathon Multimedia, 1998.
20. Stone VE, Clarke J, Lovell J, et al. HIV/AIDS patients perspectives on
adhering to regimens containing protease inhibitors. J Gen Intern Med
1998; 13:58693.
21. Scott-Lennox J, Braun JF, Morrow JE, Lawson K, Tirelle R, Dietrih D.
Development of the HIV treatment satisfaction survey (HTSS) to improve
adherence and quality of outpatient care [abstract 12445]. In: Program
and abstracts of the 12th World AIDS Conference. Geneva: Marathon
Multimedia, 1998.
22. Salicru R, Winter C, Tanowski M, et al. Comparison between physicians
and patients advises about adherence to antiretroviral therapies [abstract
60481]. In: Program and abstracts of the 12th World AIDS Conference.
Geneva: Marathon Multimedia, 1998.
23. Kristofco R, Howell L, Squires KE, Casebeer LL, Carillo AS, Shapiro J.
Treatment guidelines in HIV: self reported physician practices [abstract
24120]. In: Program and abstracts of the 12th World AIDS Conference.
Geneva: Marathon Multimedia, 1998.
24. Lubin B, Linsk N, Sherer R, Schechtman B. The Midwest AIDS training
and education partners (MATEP) adherence initiative: rationale and goals
[abstract 32381]. In: Program and abstracts of the 12th World AIDS
Conference. Geneva: Marathon Multimedia, 1998.
25. Chesney MA, Ickovics J, Hecht FM, Skipa C, Rabkin J. Adherence: a ne-
cessity for successful HIV combination therapy. AIDS 1999; 13:S18.
26. Belzer M, Fuchs D, Tucker D, Slonimsky G. High risk behaviors are not
predictive of antiretroviral non-adherence in HIV

youth [abstract 32372].

In: Program and abstracts of the 12th World AIDS Conference. Geneva:
Marathon Multimedia, 1998.
27. Lerner BH, Gulick RM, Dubler NN. Rethinking non-adherence: historical
perspectives on triple-drug therapy for HIV disease. Ann Intern Med
1998; 129:5738.
28. Knoebel H, Carmona A, Grau S, Saballs P, Gimeno JL, Lopez Colomes JL.
Strategies to optimise adherence to highly active antiretroviral treatment
[abstract 32322]. In: Program and abstracts of the 12th World AIDS
Conference. Geneva: Marathon Multimedia, 1998.
29. Mannheimer S, Hirsch Y, El-Sadr W. The impact of the ALR alarm device
on antiretroviral adherence among HIV-infected outpatients in Harlem
[abstract 32325]. In: Program and abstracts of the 12th World AIDS
Conference. Geneva: Marathon Multimedia, 1998.
30. Brown AM, Powell-Cope GM, Inouye J. Social support and adherence in
HIV

persons [abstract 32346]. In: Program and abstracts of the 12th

World AIDS Conference. Geneva: Marathon Multimedia, 1998.
31. Powell-Cope GM, Brown MA, Holzemer WL, et al. Perceived health care
providers support and HIV adherence [abstract 32354]. In: Program and
abstracts of the 12th World AIDS Conference. Geneva: Marathon Mul-
timedia, 1998.
32. Cramer JA. Consequences of intermittent treatment for hypertension: the
case for medication compliance and persistence. Am J Manag Care 1998;
4:15638.

a
t

U
n
i
v
e
r
s
i
t
a
s

D
i
p
o
n
e
g
o
r
o

o
n

J
u
l
y

3
1
,

2
0
1
3
h
t
t
p
:
/
/
c
i
d
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m