Journal of the American Society of Hypertension 3(3) (2009) 221227

Research Article

Six-item self-administered questionnaires in the waiting room: an aid to explain uncontrolled hypertension in high-risk patients seen in general practice
Isabelle Mulazzi, MDa, Jean Pierre Cambou, MDb, Xavier Girerd, MDc, Robert Nicodeme, MDa, Bernard Chamontin, MDb,d, and Jacques Amar, MDb,d,*
De partement de Me decine Ge ne rale, Centre Hospitalier Universitaire de Toulouse, Toulouse, France; b INSERM 558, Toulouse, France; c Ho pital Pitie Salpe trie ` re, Paris, France; and d Service de Me decine Interne et dHypertension Arte rielle, Centre Hospitalier Universitaire de Toulouse, Toulouse, France Manuscript received August 22, 2008 and accepted December 13, 2008
a

Abstract
We designed a cross-sectional study to determine whether 6-item self-administered questionnaires addressing difculties in taking treatment provide independent and relevant information on uncontrolled hypertension in high-risk cardiovascular patients seen in general practice. Patients with both treated hypertension and a history of vascular diseasesmyocardial infarction, stroke, or peripheral artery diseasewere included. Risk factors, treatment, history of vascular diseases, blood pressure, and difculties in taking treatment were assessed by 6-item self-administered questionnaires and recorded. Each positive response to the questions was weighted by 1 and each negative response by 0. Individual item scores were added together to produce 1 composite score for all 6 questions. A total of 11,096 patients were analyzed. Among them, 5,288 (51.4%) were controlled at 140/90 mm Hg threshold. In multivariate analysis, in addition to age, male gender, treated diabetes, peripheral artery disease, treatment, and alcohol consumption, the adherence score was negatively and independently associated with hypertension control (odds ratio score ! 3, 0.73; [95% condence interval, 0.650.81; P < .0001]. This study overwhelmingly conrms on a very large scale the effectiveness of this self-administered questionnaire in identifying difculties in taking treatment in general practice. This questionnaire constitutes an inexpensive and timesaving tool capable of helping general practitioners to understand why hypertension is not controlled in patients at high cardiovascular risk. Whether the use of this questionnaire will improve hypertension control remains to be established. J Am Soc Hypertens 2009;3(3):221227. 2009 American Society of Hypertension. All rights reserved.
Keywords: Adherence; secondary prevention; control; cardiovascular.

Introduction
Approximately two-thirds of cases of treated hypertension are not controlled at 140/90 mm Hg threshold. Furthermore, whereas the benet of blood pressure (BP) control is greater

The ESPERE Study and the analysis of these data were nancially supported by Sano Aventis. Dr. Amar received consultancy fees from Sano Aventis to design and analyze the study. Conict of interest: none. *Corresponding author: Jacques Amar, MD, Centre Hospitalier Universitaire de Toulouse, Hopital Rangueil, Department of Internal Medicine and Arterial Hypertension, Toulouse, France. Tel: 33 5 61 323 084; fax: 33 5 61 32 27 10. E-mail: amar.j@chu-toulouse.fr

in patients with a higher baseline cardiovascular risk1 in primary prevention populations, the presence of associated cardiovascular risk factors such as age, obesity, and diabetes is negatively correlated with both an increased incidence of hypertension and poor hypertension control.2,3 In secondary prevention, the difculty in controlling hypertension has been demonstrated.46 Cross-sectional surveys conducted in Europe show that some 50% of patients admitted for unstable angina or myocardial infarction have high BP 6 months after hospitalization.7 Among the factors underlying these disappointing results, lack of adherence to BP lowering medication is a major reason for poor control of hypertension worldwide. At least 40% to 50% of patients do not adhere to long-term therapies for hypertension8 and half of those thought to have refractory hypertension are

1933-1711/09/$ see front matter 2009 American Society of Hypertension. All rights reserved. doi:10.1016/j.jash.2008.12.004

222

I. Mulazzi et al. / Journal of the American Society of Hypertension 3(3) (2009) 221227

Figure. Adherence evaluation test score and prevalence of uncontrolled hypertension.

nonadherent.9,10 Signicantly, DiMatteo et al,11 using a metaanalysis of research correlating adherence with objective measures of treatment outcomes such as glycosylated hemoglobin in diabetics or self-reported pain in knee osteoarthritis, state that the difference in outcome between highand low-adherence is approximately 26%. Furthermore, research has consistently failed to identify demographic variables that predict whether patients will follow the advice of physicians.1215 In other words, nonadherence occurs regardless of class, ethnicity, or level of education. In line with these data, clinicians have been shown to make inaccurate and biased judgments concerning their patients level of compliance, even among patients they feel they know well.16 Therefore, it is of paramount importance to provide practitioners in charge of hypertensive patients with accurate and reproducible measures of adherence to the prescribed regime. However, self-reported compliance has repeatedly been found to be poorly correlated with real compliance.16 In addition, electronic monitoring of compliance is not yet available in general practice. Franc The Comite ais de Lutte Contre Lhypertension rielle has therefore devised a simple self-administered Arte questionnaire with 6 items addressing difculties in taking antihypertensive treatment (Figure). The reliability of this evaluation test to detect nonadherence among hypertensive-treated patients has been reported.17 However, whether this approach is feasible in clinical practice and genuinely helps the practitioner to determine the reasons why hypertension is not controlled remains to be seen. The aim of this study was to assess, on a large scale, whether this questionnaire constitutes an inexpensive and timesaving tool providing caregivers with valuable and relevant information on resistant hypertension in high-risk cardiovascular patients recruited in general practice.

in France. The survey was completed anonymously. All patients seen between April and June 2003 were eligible. The rst 5 patients with a documented history of ischemic stroke, myocardial infarction, or peripheral artery disease per doctor were included. Patients who had had an acute coronary event or any revascularization therapy in the previous 3 months were excluded. First, in the waiting room and without the help of the GP, patients completed the adherence test (Figure). However, patients knew that the doctor would see their questionnaire. Practitioners then reported gender, age, smoking habits, alcohol consumption, diabetes mellitus, drug regimens, and cardiovascular history. Height, weight, and BP were measured. Smokers were dened as patients who currently smoked (cigarettes smoked >1/day) or who had stopped smoking less than 1 year ago. The last measurement of low-density lipoprotein (LDL) cholesterol available in the medical records was noted.

BP Measurement
The investigators were instructed to measure BP using their usual devices after a minimum 5-minute rest period. The average of 3 BP recordings was used for the statistical analysis. Systolic blood pressure (SBP) was determined by the occurrence of repetitive sounds, and diastolic blood pressure (DBP) by Korotkoff phase V sounds.

Screening Questionnaire to Estimate Adherence to Antihypertensive Treatment

Patients were asked to complete a 6-item questionnaire to assess factors that could affect medication adherence. The questions were as follows: 1) Did you omit to take your treatment this morning? 2) Since your last visit, have you run out of treatment? 3) Have you ever taken your treatment later than instructed? 4) Have you ever forgotten to take your treatment? 5) Have you ever decided not to take your treatment because of its side effects? 6) Do you feel that the number of pills you have to take daily is too high? A score between 0 and 6 was calculated by adding the positive answers together. This questionnaire has previously been validated

Methods
ESPERE was a cross-sectional study conducted in all French regions in a geographically stratied and random sample of 4,000 general practitioners (GP). The number of GP by region was closely related to the distribution of GP

I. Mulazzi et al. / Journal of the American Society of Hypertension 3(3) (2009) 221227

223

in hypertensive patients,17,18 and a positive relationship between nonadherence (with the drug being taken as prescribed on less than 80% of days) estimated by electronic monitoring of adherence using medication event monitoring system (MEMS) devices and the score given by the questionnaire (c2, P < .02) has been demonstrated in 140 hypertensive patients treated with at least 1 antihypertensive medication. A receiver operating characteristic curve analysis showed that a score of 3 or above was the best cut-off to estimate nonadherence. Where the questionnaire score was 3 or above, the positive predictive value to detect nonadherence was 80.3%.

Data Analysis
The analysis was performed in patients with treated hypertension. Controlled hypertension was dened as BP <140/90 mm Hg on the day of consultation1 and uncontrolled hypertension as BP !140/90 mm Hg. Adherence to antihypertensive treatments was assessed using the adherence test (Figure). Each positive response was weighted by 1, and each negative response by 0. Individual item scores were added together to produce a composite score for all 6 questions ranging from 0 to 6. The higher the score, the greater the difculties in taking treatment. The relationships between demographic characteristics, cardiovascular history, treatments, cardiovascular risk factors, adherence test, and hypertension control were assessed rst by bivariate analysis and then using logistic regression. We introduced the variables found to be significantly associated with hypertension control in the bivariate analysis to the model. Because a score of 3 or above has been shown to be the best cut-off to estimate nonadherence and to identify a threshold, we analyzed the adherence test score as a dichotomous variable (<3 or !3) in a logistic regression model. To avoid overadjustment, we also introduced treated dyslipidemia and treated diabetes instead of statins and antidiabetic drugs to the model; for obvious reasons, we did not introduce BP level. Statistical analysis was performed using SAS statistical software (SAS/STAT users guide 1997, release 6.12; SAS Institute Inc, Cary, NC). P < .05 was considered as statistically signicant.

current smoking, alcohol consumption, diuretics, calciumchannel blockers (CCB), angiotensin-receptor blockers (ARB), central acting agents, or alpha-blockers were associated with uncontrolled hypertension (Table 1). On the contrary, beta-blockers, antiplatelet agents, statins, and LDL cholesterol <1 g/L were positively correlated with hypertension control (Table 2). All screening questions from the adherence test were also correlated with BP control (Table 3). As shown in the Figure, the percentage of uncontrolled hypertensive patients increased gradually alongside the adherence test score from 43% to 60%. In multivariate analysis (Table 4), age, obesity, treated diabetes, alcohol consumption (>2 drinks per day), peripheral artery disease, prescribing a CCB, ARB, alpha-blocker and central-acting agents, and the score assessing the difculties in taking treatment were negatively and independently associated with hypertension control. Conversely, male gender, treated dyslipidemia, and LDL cholesterol <1 g/L were positively and independently correlated with hypertension control (Table 4).

Discussion
The main nding of this study conducted in patients with a previous history of cardiovascular diseases (CVD) was that a self-administered questionnaire addressing difculties in taking treatment provides signicant information on hypertension control. Poor compliance with antihypertensive drug regimens is one recognized cause of inadequate BP control.18 Compliance is difcult to measure, so poor adherence to treatment remains largely undiagnosed in clinical practice. When the therapeutic response to a drug is not as expected, it is a major challenge for many physicians to decide whether the patient is a nonresponder or a noncomplier.19 Indeed, many clinicians nd it difcult to differentiate between suboptimal adherence and nonresponse to drug treatment if a patient is not controlled despite being prescribed effective drugs. Poor compliance is therefore often wrongly interpreted as a lack of response to treatment. Not detecting noncompliance can lead to the wrong measures being taken. Providing practitioners with an inexpensive and practical tool to estimate adherence to treatment is therefore a major issue. However, it has repeatedly been shown that self-reported adherence to treatment may often be inaccurate because of difculties with patient recall or attempts to please the practitioner. In addition, whereas electronic monitoring of compliance provides important longitudinal information about drug-intake behavior,20,21 this method cannot be used easily in daily practice. In an attempt to develop a simple and inexpensive tool to Franc estimate adherence to treatment, the Comite ais de rielle has therefore Lutte Contre Lhypertension Arte devised a simple 6-item self-reported questionnaire to identify difculties in taking antihypertensive treatment rather than adherence to treatment. The interest of this questionnaire in detecting nonadherence among treated

Results
From the 16,835 patients included in the ESPERE study, 11,284 (68.90%) had been treated for hypertension. Among them, information on demographic characteristics, medical history, treatment, or BP was missing in 649 (5.75%) patients and the questionnaire was not completed correctly by 342 (3.03%) patients. All these patients were removed from the analysis. Finally, 10,293 treated hypertensive patients formed the basis of this report (Table 1). Among them, 5,288 (51.4%) were controlled above the 140/90 mm Hg threshold. Age, male gender, obesity, peripheral artery disease, treated diabetes,

224

I. Mulazzi et al. / Journal of the American Society of Hypertension 3(3) (2009) 221227

Table 1 Characteristics of the study population Controlled hypertensives n 5,288 Age Male Body mass index Obesity (body mass index !30 kg/m2) Cardiovascular history Coronary disease Stroke Peripheral artery disease Metabolic risk factors Treated dyslipidemia Treated diabetes Smoking status Current smoker Ex-smoker Never smoker Alcohol consumption 2 drinks per day 34 drinks per day 57 drinks per day !8 drinks per day Systolic blood pressure Diastolic blood pressure LDL cholesterol (g/L)a <1 g/L !1 g/L Not available
a

Uncontrolled hypertensives n 5,005 68.5 10.2 3662 (72.0) 27.8 4.3 1313 (26.2) 3535 (70.6) 1013 (20.2) 1799 (35.9) 3425 (68.4) 1559 (31.1) 706 (14.1) 2464 (49.2) 1835 (36.7)

P .001 .003 <.0001 <.0001 <.0001 .21 <.0001 <.0001 <.0001 .002

67.8 10.4 3995 (74.5) 26.9 4.1 1075 (20.3) 3894 (75.3) 1019 (19.3) 1466 (27.7) 3868 (73.1) 1249 (23.6) 634 (12.0) 2727 (51.6) 1927 (36.4) 3959 (74.9) 1070 (20.2) 214 (4.0) 45 (0.8) 129.4 7.2 75.77 6.4 1165 (22.0) 2937 (55.5) 1186 (22.4)

<.0001 3330 (66.5) 1261 (25.2) 325 (6.5) 89 (1.8) 148.2 9.7 83.10 7.4 836 (16.7) 2897 (57.8) 1272 (25.4)

<.0001 <.0001 <.0001

LDL, low-density lipoprotein. LDL cholesterol was available in 3,733 patients with uncontrolled hypertension and in 4,102 patients with controlled hypertension.

hypertensive patients has been established.18 In line with this approach, we assessed the relationship between this compliance evaluation test and hypertension control. Importantly, we showed that this questionnaire brings relevant and not redundant information on hypertension control. These results suggest that this self-administered questionnaire could be a valuable, timesaving, and inexpensive tool in helping general practitioners to understand why hypertension is uncontrolled in high-risk cardiovascular patients. Signicantly, the low number of subjects who answered yes to the rst question Did you omit to take your treatment this morning? may support the idea that patients are more likely to take their medications on the day of their visit to the doctor to ensure that their measured BP is controlled and suggests that the hypertension control rate is overestimated in the population. In this respect, it is likely that the use of this questionnaire may help practitioners to start a discussion on the difculties in taking treatment. Other such subjective evaluation, including the 4-item Morisky Medication-Taking Behavior Scale,22 was found to be associated with hypertension control. However, to the best of our knowledge, this study is the rst to

validate on a very large scale the feasibility and interest of this approach in high-risk cardiovascular patients seen in general practice. Obviously, whether the use of this questionnaire results in improvement in BP control remained to be established by a prospective randomized trial. Finally, among factors found to be independently correlated with hypertension control, our study underscores the role of alcohol consumption. Epidemiological and experimental investigations have established a close association between alcohol consumption and hypertension.23 Among the >50 epidemiological studies that have addressed this question, some have recorded a linear dose-response relationship, sometimes starting with a consumption threshold of 3 drinks per day (30 g of ethanol).24,25 In line with these data, we showed that the likelihood of control decreased continuously with increasing alcohol consumption. Furthermore, this large secondary prevention population provided the opportunity to estimate the prevalence of excessive drinkers among patients with uncontrolled hypertension. Importantly, regarding BP level, we showed that one-third of these patients with uncontrolled hypertension had excessive alcohol consumption. This nding may account for

I. Mulazzi et al. / Journal of the American Society of Hypertension 3(3) (2009) 221227 Table 2 Antihypertensive treatment and adherence evaluation test score according to hypertension control Controlled hypertensives n 5,005 Treatment Statins Antiplatelet agents Beta-blockers Angiotensin-converting enzyme inhibitors Calcium-channel blockers Diuretics Angiotensin-receptor blocker Central acting agents or alpha-blockers Antihypertensive strategy 1-drug therapy 2-drug therapy 3 or more drug therapy 3 or more drug therapy with diuretic Adherence evaluation test score Adherence evaluation test score !3 (%)
a

225

Uncontrolled hypertensives n 5,288 3021 (60.4) 3700 (73.9) 2169 (43.3) 2545 (50.8) 2039 (40.7) 2203 (44.0) 1142 (22.8) 466 (9.3) 1497 (29.9) 1929 (38.5) 1579 (31.5) 1303 (26.0) 1.5 1.3 1011 (20.2)

3478 (65.8) 4095 (77.4) 2643 (49.9) 2772 (52.4) 1971 (37.3) 2042 (38.6) 941 (17.8) 236 (4.5) 1648 (31.2) 2241 (42.4) 1399 (26.5) 1175 (22.2) 1.2 1.2 721 (13.6)

<.0001 <.001 <.0001 .11 .0003 <.0001 <.0001 <.0001 <.0001a

<.0001 <.00001 <.0001

Comparison performed between 1-drug, 2-drug, and 3-drug combination therapy.

a large proportion of resistant hypertension in patients with a previous history of CVD. Finally, as previously reported, we identied the uncontrolled hypertensive patients as having more associated risk factors. Pathophysiological mechanisms involved in the clustering of hypertension with other cardiovascular risk factors may inuence BP control. For example, obesity is associated with apnea syndrome, a cause of resistance to antihypertensive treatment.26 Neuropathy and nephropathy associated with diabetes27,28 negatively inuence BP control. With aging, smoking,29 and the extent of atherosclerosis,30 the arterial wall stiffens, resulting in increased pulse pressure and in SBP. Therefore, the presence of additional risk factors may per se contribute to resistance to antihypertensive treatment.
Table 3 Screening questions and blood pressure control

Limitations and Strengths of the Study

The results of this study should be interpreted in light of some potential limitations. First, we studied a secondary prevention population. Whether the interest of the questionnaire could be generalized to the primary prevention population is questionable since poor adherence to treatment because of a large number of prescribed drugs is likely to be less frequent in patients with no history of CVD. However, the greater the risk, the greater the benet of BP control; therefore, it is of paramount importance to estimate adherence to treatment in secondary prevention patients. Second, this compliance evaluation test focused on the difculties in taking antihypertensive treatment. Obviously,

Controlled hypertensives (n 5,288)

Uncontrolled hypertensives (n 5,005)

Yes answer Did you omit to take your treatment this morning? Since the last visit, have you run out of treatment? Have you ever taken your treatment with delay in comparison with usual schedule? Have you ever missed your treatment because of bad memory? Have you ever decided not to take your treatment because of side effects? Do you feel that the number of pills you have to take daily is too high? Positive answer to 3 or more questions (%) Percentages are in parentheses. 166 (3.1) 494 (9. 3) 2480 (46.9) 824 (15.6) 375 (7.1) 2033 (38.4) 721 (13.6) 243 (4.9) 722 (14.4) 2680 (53.5) 981 (19.6) 534 (10.7) 2203 (44.0) 1011 (20.2) <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001

226

I. Mulazzi et al. / Journal of the American Society of Hypertension 3(3) (2009) 221227

Table 4 Determinants of controlled blood pressure (logistic regression analysis) Odds ratio Age (per 1-year increase) Male vs. female Obesity (yes vs. no) Treated dyslipidemia (yes vs. no) LDL cholesterol !1 g/L vs. <1 g/L Not available vs. <1 g/L Treated diabetes (yes vs. no) Alcohol consumption 34 drinks vs. 02 drinks 57 drinks vs. 02 drinks !8 drinks/day Peripheral artery disease (yes vs. no) 2 hypotensive drug combination vs. monotherapy (yes vs. no) Calcium-channel blocker (yes vs. no) Angiotensin-receptor blocker (yes vs. no) Alpha-blockers or central acting agents (yes vs. no) Adherence evaluation test with score !3 (yes vs. no) LDL, low-density lipoprotein. .99 1.19 .79 1.12 .75 .72 .75 .70 .59 .48 .77 1.13 .90 .79 .51 .73 95% Condence intervals .99 1.06 .71 1.02 .67 .64 .69 .64 .48 .33 .70 1.00 .80 .71 .43 .65 .99 1.32 .87 1.23 .83 .82 .83 .78 .71 .70 .85 1.27 .99 .87 .61 .81 P .009 .002 <.0001 .02 <.0001 <.0001 <.0001 <.0001 <.0001 .0001 <.0001 .04 .03 <.0001 <.0001 <.0001

many other factors such as psychosocial, depressive symptoms have been associated with noncompliance. However, only a very simple and self-administered questionnaire could be widely used in general practice. In this respect, the feasibility of this questionnaire has been established by this study: indeed, of the 11,284 treated hypertensive patients included, less than 5% did not complete this questionnaire in full. In addition, as patients were aware that their doctor would see their individual results, the design may undermine the usefulness of this questionnaire in daily practice. However, its reliability, if it is used each time the patient comes to see the physician, remains to be explored.

2.

3.

4.

Conclusion
5. In conclusion, this study validates on a very large scale and in general practice a self-administered questionnaire addressing difculties in taking treatment. This questionnaire provides relevant information on uncontrolled hypertension and constitutes an inexpensive and timesaving tool helping clinicians to understand why hypertension is not controlled in patients at high cardiovascular risk. Whether the use of this compliance evaluation test results in improvement in hypertension control remains to be established by prospective randomized trials.

6.

7.

References
1. European Society of Hypertension-European Society of Cardiology Guidelines Committee. 2003 European Society of Hypertension-European Society of 8.

Cardiology guidelines for the management of arterial hypertension. J Hypertens 2003;21:101153. Shankar A, Wang JJ, Rochtchina E, Mitchell P. Positive association between plasma brinogen level and incident hypertension among men. Population-based cohort study. Hypertension 2006;48:10439. Amar J, Vaur L, Perret M, Bailleau C, Etienne S, Chamontin B. PRATIK Study Investigators. Hypertension in high-risk patients: beware of the underuse of effective combination therapy (results of the PRATIK Study). J Hypertens 2002;20:77984. Amar J, Chamontin B, Genes N, Cantet C, Salvador M, Cambou JP. Why is hypertension so frequently uncontrolled in secondary prevention? J Hypertens 2003;21: 1199205. Amar J, Chamontin B, Ferrieres J, Danchin N, Grenier O, Cantet C, et al. Hypertension control at hospital discharge after acute coronary event: inuence on cardiovascular prognosisthe PREVENIR Study. Heart 2002; 88:58791. Amar J, Cambou JP, Quentzel S, Amelineau E, Danchin N. Controlled diastolic blood pressure, previous stroke and associated risk factors are obstacles to improving systolic blood pressure. J Hum Hypertens. Forthcoming. EUROASPIRE II Study Group. Lifestyle and risk factor management and use of drug therapies in coronary patients from 15 countries; principal results from EUROASPIRE II Euro Heart Survey Programme. Eur Heart J 2001;22:55472. Caro JJ, Salas M, Speckman JL, Raggio G, Jackson JD. Persistence with treatment for hypertension in actual practice. CMAJ 1999;160:317.

I. Mulazzi et al. / Journal of the American Society of Hypertension 3(3) (2009) 221227

227

9. Setaro JF, Black HR. Refractory hypertension. N Engl J Med 1992;327:5436. 10. Yakovlevitch M, Black HR. Resistant hypertension in a tertiary care clinic. Arch Intern Med 1991;151:178692. 11. DiMatteo MR, Giordani PJ, Lepper HS, Croghan TW. Patient adherence and medical treatment outcomes: a meta-analysis. Med Care 2002;40:794811. 12. Stewart M. The validity of an interview to assess a patients drug taking. Am J Prev Med 1987;3:95100. 13. Kruse W, Weber E. Dynamics of drug regimen compliance. Its assessment by microprocessor-based monitoring. Eur J Clin Pharmacol 1990;38:5615. 14. Mallion JM, Meilhac B, Tremel F, Calvez R, Bertholom N. Use of a microprocessor-equipped tablet box in monitoring compliance with antihypertensive treatment. J Cardiovasc Pharmacol 1992;19:S418. 15. Vaur L, Vaisse B, Genes N, Elrik F, Legrand C, Poggi L. Use of electronic pill boxes to assess risk of poor treatment compliance. Results of a large-scale trial. Am J Hypertens 1999;12:37480. 16. Waeber B, Burnier M, Brunner HR. Compliance with antihypertensive therapy. Clin Exp Hypertens 1999; 21:97385. 17. Girerd X, Radauceanu A, Achard JM, Fourcade J, Tournier B, Brillet G, et al. The compliance evaluation test: a validated tool for detection of nonadherence among hypertensive treated patients. Arch Mal Cur 2001;94:83942. 18. Girerd X, Hanon O, Anagnostopoulos K, Ciupek C, Mourad JJ, Consoli S. Assessment of antihypertensive compliance using a self-administered questionnaire: development and use in a hypertension clinic. Presse Med 2001;30:10448. 19. Burnier M. Medication adherence and persistence as the cornerstone of effective antihypertensive therapy. Am J Hypertens 2006;19:11906. 20. Burnier M, Santschi V, Favrat B, Brunner HR. Monitoring compliance in resistant hypertension: an important step in patient management. J Hypertens Suppl 2003;21:S3742.

21. Burnier M, Schneider MP, Chiolero A, Stubi CL, Brunner HR. Electronic compliance monitoring in resistant hypertension: the basis for rational therapeutic decisions. J Hypertens 2001;19:33541. 22. Morisky DE, Green LW, Levine DM. Concurrent and predictive validity of a self-reported measure of medication adherence. Med Care 1986;24:6774. 23. Fuchs FD, Chambless LE, Whelton PK, Nieto FJ, Heiss G. Alcohol consumption and the incidence of hypertension: the Atherosclerosis Risk in Communities Study. Hypertension 2001;37:124250. 24. MacMahon SW, Blacket RB, Macdonald GJ, Hall W. Obesity, alcohol consumption and blood pressure in Australian men and women: the National Heart Foundation of Australia Risk Factor Prevalence Study. J Hypertens 1984;2:8591. 25. Marques-Vidal P, Arveiler D, Evans A, Amouyel P, Ferrieres J, Ducimetiere P. Different alcohol drinking and blood pressure relationships in France and Northern Ireland: the PRIME Study. Hypertension 2001;38:13616. 26. Grote L, Hedner J, Peter JH. Sleep-related breathing disorder is an independent risk factor for uncontrolled hypertension. J Hypertens 2000;18:67985. 27. Schrier RW. Treating high-risk diabetic hypertensive patients with comorbid conditions. Am J Kidney Dis 2000;36:S107. 28. Franklin SS, Gustin W 4th, Wong ND, Larson MG, Weber MA, Kannel WB, et al. Hemodynamic patterns of age-related changes in blood pressure. The Framingham Heart Study. Circulation 1997;96: 30815. 29. Amar J, Ruidavets JB, Chamontin B, Drouet L, Ferrieres J. Arterial stiffness and cardiovascular risk factors in a population-based study. J Hypertens 2001;19:3817. 30. Blacher J, Asmar R, Djane S, London GM, Safar ME. Aortic pulse wave velocity as a marker of cardiovascular risk in hypertensive patients. Hypertension 1999;33:11117.