Original article
Compliance with therapy in hypertensive patients Alfonso Lagi, Antonio Rossi, Maria Teresa Passaleva, Alessandro Cartei, Simone Cencetti Department of Internal Medicine, S. Maria Nuova Hospital, Florence, Italy
Objective. This study investigated pharmacological compliance in hypertensive patients recruited during a 24-month period at the Hypertension Clinic of the S. Maria Nuova Hospital. Methods. The compliance profile scale was developed in order to classify compliance as sufficient, or poor. Compliance was evaluated according to patients’ 1) global cardiovascular risk (hypertensive) and 2) complexity of prescribed treatment at enrolment (drug, number of pills). Results. A total of 367 patients were recruited; 82 were excluded owing to changes in therapeutic schedule or medication necessitated by other medical or surgical disease. Compliance was evaluated in 285 patients, 66% were male; the mean age was 72 ± 12 years. Compliance was evaluated at a series of follow-ups at 3-month intervals and at 36 months of observation. Compliance at 36 months was good or sufficient in 56% of patients, but differed according to hypertensive and
drug status. Compliance was better in patients with a higher level vs lower level of hypertension severity. Specifically, compliance was better in hypertensives with co-morbidity (76%) than in hypertensives without co-morbidity (37%) or hypertensives with risk factors (65%). Moreover, compliance was better in patients with a more complex therapeutic regimen (drug 1 through 3: 47, 59 and 66%). Conclusions. These data demonstrate a relationship between disease severity and compliance, independent of the prescribed therapeutic regimen. This suggests that the most important factor for therapy adherence is disease awareness and severity, and patients with a higher level of disease are more likely to be compliant. (Intern Emerg Med 2006; 1 (3): 204-208)
Introduction
reported compliance, pill counts and measurement of urine or plasma drug concentrations12,13. Thus far, pillboxes that electronically record every opening have been regarded as the most useful and accurate method for tracking compliance, although this device can be used only with joint work of medical industries and is difficult to impossible to enforce in general clinical practice14. The aim of the present study was to investigate the compliance with pharmacological therapy of hypertensive patients treated in the Hypertension Clinic of the S. Maria Nuova Hospital. Self-reported compliance and institutional control were used to monitor the patients and evaluate patient conduct with respect to the prescribed pharmacological treatment. We predicted that patient compliance would vary according to both the complexity of the therapeutic regimen and hypertension severity.
Compliance with therapeutic treatment is difficult to measure, and few attempts have been made to do so because of the many obstacles present in everyday practice. The current opinion is that poor adherence to medication regimens remains largely underdiagnosed, and may contribute to a lack of response to hypertension treatment1-3. Thus, patients thought to have resistant hypertension may in fact be non-compliant4,5. In recent studies investigating insufficient response to antihypertensive treatment, 17-50% of patients were determined to be noncompliant with the therapeutic drug regimen6,7. Non-compliance is a major concern considering that control of arterial hypertension is achieved in only approximately 25% of patients, and only 60% of patients take their drugs as prescribed8,9. Furthermore, based on clinical research studies, it has been calculated that 1 year after diagnosis, only 15% of treated patients strictly follow their drug regimen, and up to 50% discontinue medication during that time10,11. The extent of non-compliance is difficult to estimate and different methods have been proposed, including self-
Received 12 June 2006; revised 27 June 2006; accepted 27 June 2006. Address for correspondence: Dr. Alfonso Lagi, Department of Internal Medicine, S. Maria Nuova Hospital, Via G. Mameli 44, 50131 Florence, Italy. E-mail:
[email protected] © 2006 CEPI Srl
204
Key words: clinical management, compliance, hypertension
Methods Patients Three hundred and sixty-seven consecutive outpatients referred to the Hypertension Clinic of the S. Maria Nuova Hospital were recruited for the study between January 2001 and December 2002. Informed consent was obtained for each patient. At the first visit, patients were informed that they had been evaluated for their compliance and that they had been included in an interventional trial. The patients were evaluated with respect to two factors: level of global cardiovascular risk (GCVR) (hypertensive,
IM Alfonso Lagi et al.
hypertensive with risk factors, hypertensive with comorbidity) and number of pills prescribed15. Each factor included three levels, designated 1, 2, and 3 and representing progressively worse prognosis founded on GCVR or progressively increased prescription complexity (Fig. 1). Drug 1 corresponded to 1-2 pills/day, drug 2 corresponded to 3 pills/day, and drug 3 corresponded to ≥ 4 pills/day, independent of the number of pharmacological active molecules in each pill. By grouping patients according to hypertensive and drug factors for purposes of data analysis, patient compliance was evaluated with respect to both severity of prognosis and complexity of therapy. Patients were excluded if during the course of the study they underwent a change in the number of pills prescribed, thus shifting from one drug category to another. Patients were still included in the follow-up if their medication changed but not the total number of tablets per day prescribed. Procedures At enrolment, each patient was provided with an individualised daily therapeutic plan including a pharmacological prescription (one or more pills), a dietary plan (number of calories per day), and an exercise program (≥ 20 min 3 times a week). Only the pharmacological plan was considered in this analysis. The therapeutic plan was established according to the needs of each patient by the doctors responsible for outpatients of the hypertension clinic (AR and MTP).
367 patients recruited
Groups
Hypertensive 367 patients
Subgroups
123
Drug 367 patients
123
82 excluded
285 Statistical analysis applied
Compliant
Treatment adherence was established according to the therapeutic program. The same physicians performed a follow-up every 3 months for 36 months. All patients were consulted for the same amount of time per visit. At each visit, the physician verified the patient’s lifestyle changes (dietary and physical exercise) and pharmacological compliance. Hypertension severity was also calculated (see patients section), and patient compliance was described independent of the results of the therapy. Each patient was treated with the goal of hypertension control, and the therapeutic plan (definite drug or lifestyle) was altered as necessary to obtain the therapeutic target. During follow-up visits, the assigned physicians could, based on their best judgement, vary the prescribed medication to obtain better hypertension control. Table 1 shows the compliance profile scale, developed at the beginning of the study in order to define the degree of pharmacological adherence. Patient compliance was categorised as good, sufficient, or poor according to the number of violations to the therapeutic plan. During the scheduled office visits, the physician performed a structured interview. The patient was asked to describe his or her behaviour and adherence to the assigned therapeutic plan (patients were conscious that their declaration would not exclude them from the hypertension clinic). Adherence was verified in two ways, patient self-report (i.e. “I followed or I did not follow the therapeutic plan”) and by monitoring patient drug acquisition with respect to pre-established quantitative goals. The latter was accomplished by requiring the acquisition of medication through the hospital pharmacy. Each patient was provided with the prescribed number of tablets for an allotted time and was required to return the empty blister packs in order to receive more medication. In this way, investigators could compare the number of pills prescribed to the number of pills consumed. The judgement of compliance was made by the doctors and the pharmacist, both of whom had to be in agreement. Patients lost to follow-up were considered dropouts and considered in the subsequent statistical analysis. Results are expressed as the total number and percentage of patients in each drug and hypertensive category and subcategory according to level of compliance. Patients with ratings of sufficient or good were designated compliant, while patients with ratings of poor were designated non-compliant. A χ2 test for paired data was Table 1. Compliance profile scale defining the degrees of pharmacological adherence.
Non-compliant
No. protocol violations/no. days Excluded for protocol violation
Figure 1. Study profile.
Dropouts
Good Sufficient Poor
< 1/7 or < 4/30 > 2/7 or < 8/30 > 3/7 or > 12/30
205
Intern Emerg Med 2006, Vol 1 No 3
used to compare group data. Statistical significance was set at p < 0.05. Logistic regression was used to determine the extent to which patient compliance differed by age and sex, controlling for other covariates. STATA statistical software (STATA Corp., College Station, TX, USA) was used for these analyses.
Results Of the 367 patients recruited for the study, 82 were excluded because they changed drug category as a result of a change in medical needs independent of compliance. That is, these patients had other medical or surgical diseases that required a change of medication and schedule. Statistical analysis was performed on a total of 285 patients (Fig. 1). The mean age was 72 ± 15 years (range 18-83 years), and 189 were men. The remaining results are presented according to GCVR (hypertensive 1-3) and pharmacological complexity (drug 1-3) status. Patient distributions across these categories are presented in Table 2. Patient medications are shown in Table 3. Forty of the 285 recruited patients were lost to follow-up at 36 months of observation and were considered noncompliant. Their allocation to the different subgroups is shown in Table 4. No significant difference in dropout rate was found between the drug and hypertensive designations and their subgroups (1-3). Follow-up results are shown in Tables 5 and 6. There were 161 (56%) and 126 (44%) compliant and non-compliant patients, respectively. Compliance was significantly less frequent and non-compliance more frequent for hypertensive 1 (36 compliant vs 64% non-compliant) than for hypertensive 2 (65 vs 35%) and 3 (79 vs 21%). For the
Table 2. Patient distribution across hypertensive and drug categories. Group Hypertensive 1 Hypertensive 2 Hypertensive 3 Drug 1 Drug 2 Drug 3
No. patients
Mean age
Male %
115 (40%) 106 (37%) 64 (23%) 89 (31%) 146 (51%) 50 (18%)
68 70 72 70 74 75
76 69 41 59 96 32
Table 3. Medication use in the patient population. Medication
No. patients
Diuretic Beta-blockers Calcium channel blockers ACE-inhibitors ARB (sartanic) Other
128 (45%) 35 (12%) 43 (15%) 114 (40%) 35 (12%) 14 (5%)
ACE, angiotensin-converting enzyme; ARB, angiotensin II receptor blockers.
Table 4. Patients lost at follow-up. Category
No. dropouts
Hypertensive 1 Hypertensive 2 Hypertensive 3 Total Drug 1 Drug 2 Drug 3 Total
18 (15%)* 13 (12%) 9 (14%) 40 12 (13%)* 22 (15%) 6 (12%) 40
* no significant difference between subgroup 1 and subgroups.
Table 5. Patient compliance according to global cardiovascular risk. No. patients Hypertensive 1 Hypertensive 2 Hypertensive 3
115 106 64
Compliance Good
Sufficient
Good or sufficient
Poor
20 35 **45**
21 34 6
*41 (36%)* 69 (65%) 51 (79%)
74 (64%) 37 (35%) 13 (21%)
* p < 0.01 hypertensive 1 vs hypertensive 2 and 3; ** p < 0.01 good vs sufficient.
Table 6. Patient compliance according to prescription complexity. No. patients Drug1 Drug2 Drug3
89 146 50
* p < 0.05 drug 1 vs drug 2 and drug 3.
206
Compliance Good
Sufficient
Good or sufficient
Poor
15 46 19
27 40 14
*42 (47%)* 86 (59%) 33 (66%)
47 (53%) 60 (41%) 17 (3%)
IM Alfonso Lagi et al.
hypertensive 3 category, for which compliance was highest, there was a significant difference between good and sufficient compliance (Table 5). Patient compliance according to complexity of the drug regimen is presented in Table 6. Compliance was significantly better with regimens of more than 2 pills/day, the difference between drug 1 and 2 and drug 1 and 3 were both significant The crossing of the two analysis categories is shown in Table 7. Increased level of GCVR (from hypertensive 1 through 3) was significantly related to complexity of drug therapy. Controlling for covariates by logistic regression, individuals with hypertension associated with age, sex, risk factors or with co-morbidity (hypertensive 2 and 3) were more likely to be compliant (odds ratio [OR] 1.94 and 1.12, respectively) than hypertensive 1 individuals (OR 2.11). The same pattern was observed for the drug parameters; drug 2 and 3 patients were more likely to be compliant (OR 1.99 and 1.97) than were drug 1 individuals (OR 2.68). All patients had office hypertension control at the end of the study without significant differences.
Table 7. Distribution of patients in the groups. Hypertensive 1
Hypertensive 2
Hypertensive 3
74 41 0
11 85 10
4* 20* 40*
Drug 1 Drug 2 Drug 3
* p < 0.01 hypertensive 3 vs hypertensive 2 and 1.
Discussion Determination compliance with pharmacological regimens is difficult in the absence of a reliable method for establishing whether a patient does or does not take the medication, particularly when the patient does not know how much medication he or she consumed. A recent study demonstrated that electronic monitoring of drug consumption provides different information than does patient report or a pharmacy system16. Our results do not speak to which of these systems is better. Rather, the aim of this study was to analyse individual compliance in patients differing in cardiovascular risk and complexity of drug therapy. Our results show that in a homogeneous population of hypertensive patients, compliance increased in parallel to GCVR. The distribution of patients across drug and hypertensive categories and subcategories was not homogeneous, and may reflect levels of GCVR and therapeutic need in the study population. Data from European studies show that fewer than 50% of patients with hypertension are free from risk factors and co-morbidity17,18. In the Hypertension Optimal Treatment (HOT) study, a
therapeutic trial designed to define the optimal treatment for hypertension, 75% of patients were prescribed more than 2 pills19. In our study, 69% of patients took ≥ 3 pills. We believe that the sample size recruited for the present study is representative of the general hypertensive population in GCVR, age range, and type of therapy. The difference between dropouts in the subgroups (from 12 through 15%) is not significant (Table 4). This pattern demonstrates that the complexity of therapy does not affect the dropout rate. Other possible factors contributing to dropout were not included in the study design. However, we can assume that dropout might primarily be the consequence of intolerance to pharmacological management, in terms of the schedule or type of therapy or failure of the doctor-patient relationship. We computed the number of dropouts because that status is an index of non-compliance, independent of the causes of dropout. Epidemiological studies carried out on pharmacological adherence of outpatients show that 63-66% of patients referred to primary care physicians are non-compliant20-22. In the hospital service poor compliance is less frequent, observed in 8-53% of the cases23,28. We find that 46% of patients are non-compliant. These conflicting data require further investigation. We think that these differences reflect non-homogeneous patient populations. In the subset of hypertensive patients from our study the percentage of non-compliant patients decreases across designations 1 to 3 (64, 35 and 21% respectively), and the number of compliant patients significantly increases. Furthermore, there are significantly more patients with good vs. sufficient compliance characterised as hypertensive 3. These data suggest that the level of cardiovascular risk may be related to grade of compliance. The number of non-compliant patients in the hypertensive 1 subgroup is similar to the number of outpatients of family physicians from the European studies, while the number of non-compliant patients in the hypertensive 3 subgroup is similar to the number referred to the hospital service in the observational studies. These results regarding compliance to pharmacological prescription could signify that the key factor driving good compliance is disease awareness and severity. In fact, the outcomes for the hypertensive 2 and 3 patients demonstrate that patients with a higher cardiovascular risk have better pharmacological compliance. This observation is in line with findings reported for patients suffering from life threatening or invalidating disease, such as stroke or myocardial infarction. There are fewer dropouts (approximately 4%) and excellent pharmacological compliance (rate between 85 and 87%) among those who are in secondary prevention24,25. These results are also supported by a comparison study of compliance among hypertensive patients with and without diabetes26. Consciousness of the disease, awareness of its effects and the experience of being an invalid seem to be the best deter207
Intern Emerg Med 2006, Vol 1 No 3
rent to good or sufficient compliance. This means that physicians should spend more time with their patients, with the objective of achieving the highest possible degree of therapeutic compliance27. The pattern of results according to drug status, with drug 3 and drug 2 showing greater compliance than drug 1, might be considered unexpected because it seems plausible that it would be easier to be non-compliant when prescribed more versus fewer pills. However, the patients who were prescribed ≥ 3 pills (drug 2 and drug 3) were also hypertensive 2 or 3. Thus, the pattern of findings for drug status appears to be linked to compliance associated with hypertensive status. The lack of difference between good and sufficient compliance in the drug 3 subcategory demonstrates greater attention to prescription, differently from the other two subgroups. In conclusion, the overall results show that the higher the GCVR, the better the patient compliance. Compliance increases in parallel to the level of disease severity, while no correlation is found with age or sex. These data seem to suggest a relationship between disease awareness and compliance and provide a scientific rationale for implementing multicomponent behavioural lifestyle intervention programmes in order to control blood pressure and prevent cardiovascular risk.
References 1. Burnier M, Santschi V, Favrat B, Brunner HR. Monitoring compliance in resistant hypertension: an important step in patient management. J Hypertens Suppl 2003; 21: S37-S42. 2. Tykarski A. Resistant hypertension. Blood Press Suppl 2005; 2: 42-5. 3. Mar J, Rodriguez Alonso F. Which is more important for the efficiency of hypertension treatment: hypertension stage, type of drug or therapeutic compliance? J Hypertens 2001; 19: 149-55. 4. Setaro J, Black H. Refractory hypertension. N Engl J Med 1992; 327: 534-47. 5. O’Rorke JE, Richardson WS. Evidence based management of hypertension: what to do when blood pressure is difficult to control. BMJ 2001; 322: 1229-32. 6. Nuesch R, Schroeder K, Dieterle T, Martina B, Battegay E. Relation between insufficient response to antihypertensive treatment and poor compliance with treatment: a prospective case-control study. BMJ 2001; 323: 142-6. 7. World Health Organization. Adherence to long-term therapies: evidence of action. Geneva: WHO, 2003. 8. Eraker S, Kirscht J, Becker M. Understanding and improving patient compliance. Ann Intern Med 1984; 100: 258-68. 9. Mallion J, Baquet J, Siche J, Tremel F, de Gaudemaris M. Compliance, electronic monitoring antihypertensive drugs. J Hypertens Suppl 1998; 16: S75-S79. 10. Costa FV. Compliance with antihypertensive treatment. Clin Exp Hypertens 1996; 18: 463-72. 11. Cohen JS. Adverse drug effects, compliance and initial doses of antihypertensive drugs recommended by the Joint
208
National Committee vs. Physicians’ Desk Reference. Arch Intern Med 2001; 161: 880-5. 12. Melnikow J, Kiefe K. Patient compliance and medical research: issues in methodology. J Gen Intern Med 1994; 9: 96-104. 13. Urquhart J. Role of patient compliance in clinical pharmacokinetic. A review of recent research. Clin Pharmacokinet 1994; 27: 202-15. 14. Mallion J, Meilhac B, Tremel F, Calvez R, Bertholom M. Use of a microprocessor equipped tablet box in monitoring compliance with antihypertensive treatment. J Cardiovasc Pharmacol 1992; 19 (Suppl 2): S41-S48. 15. 1999 World Health Organization-International Society of Hypertension Guidelines for the Management of Hypertension. Guidelines Subcommittee. J Hypertens 1999; 17: 151-83. 16. Wetzels GE, Nelemans PJ, Schouten JS, van Wijk BL, Prins MH. All the glisters is not gold: a comparison of electronic monitoring versus filled prescriptions – an observational study. BMC Health Serv Res 2006; 6: 8. 17. Menotti A, Blackburn H, Kromhout D, Nissinen A, Adachi H, Lanti M. Cardiovascular risk factors as determinants of 25-year all-cause mortality in the seven countries study. Eur J Epidemiol 2001; 17: 337-46. 18. Menotti A, Mulder I, Nissinen A, et al. Cardiovascular risk factors and 10-year all-cause mortality in elderly European male populations; the FINE study. Finland, Italy, Netherlands, Elderly. Eur Heart J 2001; 22: 573-9. 19. Hansson L, Zanchetti A, Carruthers SG, et al. Effects of intensive blood pressure lowering and low dose aspirin in patients with hypertension: principal results of the Hypertension Optimal Treatment randomised trial. Lancet 1998; 150: 341-53. 20. Degli Esposti E, Sturani A, Degli Esposti L, et al. Pharmacoutilization of antihypertensive drugs: a model of analysis. J Clin Pharmacol Ther 2001; 39: 251-8. 21. Ambrosioni E, Leonetti G, Pessina AC, Rappelli A, Trimarco B, Zanchetti A. Patterns of hypertension management in Italy: results of pharmacoepidemiological survey on antihypertensive therapy. J Hypertens 2000; 18: 1691-9. 22. Sarquis LM, Dellacqua MC, Gallani MC, et al. Compliance in antihypertensive therapy: analysis in scientific articles. Rev Esc Enferm USP 1998; 32: 335-53. 23. Girerd X, Radauceanu A, Achard JM, et al. Evaluation of patients compliance among hypertensive patients treated by specialists. Arch Mal Coeur Vaiss 2001; 94: 839-42. 24. Sappok T, Faulstich A, Stuckert E, Kruck H, Marx P, Koennecke HC. Compliance with secondary prevention of ischemic stroke. Stroke 2001; 32: 1884-9. 25. GISSI Prevenzione Investigators. Results of the low-dose (20 mg) pravastatin GISSI Prevenzione Trial in 4271 patients with recent myocardial infarction: do stopped trials contribute to overall knowledge? Ital Heart J 2000; 1: 810-20. 26. Egede LE. Lifestyle modification to improve blood pressure control in individuals with diabetes: is physician advice effective? Diabetes Care 2003; 26: 602-7. 27. Kennedy JG. What information do patients need about medicines? BMJ 2003; 327: 862-3. 28. Almas A, Hameed A, Ahmed B, Islam M. Compliance to antihypertensive therapy. J Coll Physicians Surg Pak 2006; 16: 23-6.
