International Journal of Cardiology 182 (2015) 503–508
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Duration of initial antihypertensive prescription and medication adherence: A cohort study among 203,259 newly diagnosed hypertensive patients☆ Martin C.S. Wong a, Wilson W.S. Tam a,b, Harry H.X. Wang a, Clement S.K. Cheung c, Ellen L.H. Tong c, N.T. Cheung c, Stephen R. Leeder d, Sian M. Griffiths a,⁎ a
JC School of Public Health and Primary Care, Faculty of Medicine, Chinese University of Hong Kong, Hong Kong, China Alice Lee Centre for Nursing Studies, National University of Singapore, Singapore c Hospital Authority Information Technology Services — Health Informatics Section, Hong Kong, China d Menzies Centre for Health Policy, University of Sydney, Australia b
a r t i c l e
i n f o
Article history: Received 25 August 2014 Received in revised form 25 November 2014 Accepted 21 December 2014 Available online 23 December 2014 Keywords: Duration Initial antihypertensive prescription Medication adherence Cohort study Chinese population
a b s t r a c t Background: Optimal adherence with antihypertensive medications is crucial to prevent hypertension-related complications. This study evaluated whether the duration of initial antihypertensive prescription is associated with better medication adherence in a large sample of Chinese hypertensive patients. Methods and results: From a validated clinical database which consists of all patients in the public healthcare sector in Hong Kong, all patients on their first-ever antihypertensive agent from 2001 to 2005 (N = 203,259) were included and followed-up for 12 months (and up to 5 years in separate analyses). The average age was 58.7 years (SD 17.3), and the overall rate of optimal adherence (as measured by having the Proportion of Days Covered ≥ 0.80) was 32.4%. The proportion of patients whose initial prescriptions lasted for ≤6 days; 7–14 days; 15–28 days and ≥29 days was 23.7%, 24.3%, 15.1% and 37.0%, respectively. The corresponding proportion of optimal adherence was 18.1%, 20.1%, 31.0% and 50.3%. The binary logistic regression analysis showed that after controlling for age, sex, socioeconomic status, service type, drug class, and district of residence, those whose initial prescription was 7–14 days (adjusted odds ratio [AOR] = 1.17, 95% C.I. 1.12–1.22); 15–28 days (AOR = 1.90, 95% C.I. 1.82–1.99) and ≥29 days (AOR = 4.13, 95% C.I. 3.96–4.31) were significantly more likely to be adherent than those who were prescribed for ≤6 days (all p b 0.001). These findings remained significant in separate analyses where the period of follow-up was extended to 5 years. Conclusions: Shorter duration of first antihypertensive prescriptions was associated with poorer medication adherence, and this practice should be avoided if possible. © 2014 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Hypertension is currently the most common chronic disease in clinical practice, and carries a substantial global burden to patients, physicians and the healthcare system [1]. Worldwide, around 54% of stroke and 47% of ischemic heart disease were due to hypertension alone [2]. Hypertensive patients are more prone to develop concomitant complications, including cardiovascular diseases, diabetes, retinopathy and renal failure [2,3]. Its prevalence is rising in different parts of the world, while its control rate remains low [4–7].
☆ Statement of authorship: All authors take responsibility for all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation. ⁎ Corresponding author at: 2/F, School of Public Health and Primary Care, Prince of Wales Hospital, Chinese University of Hong Kong, Shatin, NT, Hong Kong, China. E-mail address: siangriffi[email protected] (S.M. Griffiths).
http://dx.doi.org/10.1016/j.ijcard.2014.12.058 0167-5273/© 2014 Elsevier Ireland Ltd. All rights reserved.
The effectiveness of antihypertensive therapies to protect against stroke and heart attack has been extensively documented [8,9]. Mazzaglia and colleagues [10] showed that higher antihypertensive adherence was associated with a 38% lower risk of acute cardiovascular events, from a cohort of more than 18,800 newly diagnosed hypertensive patients. It demonstrated that even among those who had not suffered from any previous events, optimal antihypertensive adherence confers primary prevention to a substantial extent. However, it is widely recognized that a large number of hypertensive patients do not take antihypertensive therapies as prescribed [11], and adherence with prescriptions declined with time — especially in the first year of drug initiation [12–15]. Poor medication adherence may also lead to increased morbidity, mortality, hospitalization admissions and escalated healthcare costs [16–20]. Previous studies on the determinants of medication adherence mostly focused on patients' socio-demographic factors and the service settings [21–30]. The World Health Organization (WHO) has identified
504
M.C.S. Wong et al. / International Journal of Cardiology 182 (2015) 503–508
evidence for early action to enhance adherence with long-term therapies, and antihypertensive prescription is among one of them [11]. Three recent small studies investigated the association between the duration of physician prescription and long-term adherence with cardiovascular medications [31–33], and this raised a knowledge gap of whether the duration of initial antihypertensive drugs prescribed may have an effect on subsequent adherence. In light of these studies, we tested the a priori hypothesis that shorter duration of the initial antihypertensive prescription among newly diagnosed hypertensive patients was associated with poorer medication adherence in the subsequent 12 months of their followup, based on the findings of the existing literature which implied that such an association might exist [31–33]. 2. Methods 2.1. Ethics statement This study was approved by the Clinical Ethics Research Committee of the Hospital Authority, and the Survey and Behavioral Research Ethics Committee of The Chinese University of Hong Kong. The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki. Informed consent was not required as all subjects were anonymized with unique identifiers. 2.2. Data source This study used an electronic Clinical Management System (CMS) in Hong Kong, which has been previously described in details [21–30]. The electronic healthcare records capture information on patients' demographics, prescription details, and clinical diagnoses, which were coded by the International Classification of Diseases (ICD-9 and ICD-10) or the International Classification of Primary Care (ICPC-2). It is a computerized system adopted in all clinics in the public healthcare sector which serves as the major source of health care in Hong Kong. The CMS has been the only source of data entry, allowing the linkage of physician-entered information in each patient visit for all public clinical settings situated in all districts of Hong Kong. All prescriptions are entered by the consulting physicians, and are duly checked by the dispensers or pharmacists according to standardized procedures. Any amendments to the medications prescribed after the clinical consultations are also documented in the system. This database was found to be a validated research platform, and it has a high level of data completeness in terms of sociodemographic information (100%) and prescription details (99.8%) [21]. The present study extracted information from the entire Hong Kong population, which is more than 7 million in 2014. Hong Kong is divided into three distinct regions, namely Hong Kong Island (most urbanized); Kowloon; and the New Territories (most rural). 2.3. Patients We included all patients who had a clinical consultation at any public practices and who were prescribed their first-ever antihypertensive agents between the calendar years 2001 and 2005. These consisted of α-blockers, β-blockers, calcium channel blockers (CCBs), thiazide diuretics, and angiotensin converting enzyme inhibitors (ACEIs). Patients who received angiotensin receptor blockers (ARBs), polytherapy and fixed dose combination therapies were not analyzed due to the small number of such prescriptions. Patients who were prescribed any antihypertensive medications before index date or died 12 months after the index date were also excluded. All patients were followed up for 12 months and their medication adherence in the study period was evaluated. 2.4. Outcome variables and covariates The primary outcome variable was the adherence level 12 months after the initial prescription. We used the Proportion of Days Covered (PDC) as the proxy measure for medication adherence [34]. PDC is defined as the number of days where medication is supplied in a specified period divided by the total number of days within the period. Interval-based PDC is widely recognized as an internationally accepted metric for assessment of medication adherence in large database studies [34–36]. PDC was used as the measure for medication adherence since it was adopted extensively in published studies, and tends to be operationally defined more consistently than other adherence measures [34]. Its use also facilitates comparison of our study findings with existing literature. PDC ≥ 80% indicated that the patient is medication adherent [34,35], similar to an approach adopted by Mazzaglia and colleagues [10] to classify drug adherence. The variable tested for association with the outcome was the duration of the initial prescription, defined as the number of days supplied in the initial antihypertensive medication as recorded in the computerized system. We made reference to existing literature which compared the length of prescriptions [33]. Since less than 20% of the initial prescriptions among all patients lasted for more than 60 days, we divided the prescription durations into four distinct groups: ≤6 days; 7–14 days; 15–28 days; and ≥29 days. The covariates included patients' age, sex, payment status (recipients of comprehensive social security assistance did not need to pay; and for non-recipients each one-off consultation
costs approximately US$5.77, including both investigation and prescription fees regardless of the prescription duration), the service settings (in- or day-patient clinics; specialist outpatient clinics [SOPCs], accident and emergency departments (AEDs); general outpatient clinic [GOPC], and other clinic type), the district of residence (Hong Kong Island vs. Kowloon vs. the New Territories) and the first antihypertensive drug class received. 2.5. Statistical analyses The Statistical Package for Social Sciences version 16.0 (SPSS, Inc., Chicago, Illinois) was used for all data analyses. Student's t tests and chi-square tests of heterogeneity were used to compare continuous and categorical variables, respectively. The sociodemographic characteristics were compared among patients according to the duration of the initial prescription. A binary logistic regression model was constructed with the optimal medication adherence (PDC ≥ 80%) as the outcome variable. The explanatory variable (prescription length) and all the covariates listed above were entered into the regression analysis. We evaluated for multicollinearity and interactions among variables to ensure the robustness of the regression model. As sensitivity analyses, we (1) excluded patients whose initial prescriptions were N90 days and N180 days, respectively; (2) included the frequency of follow-up visits as an additional variable; (3) stratified our patients according to the district of residence (rural vs. urban regions);(4) included one more category of the prescription duration (≥60 days); and (5) used a different categorization for duration of prescriptions (1–6 days (less than 1 week), 7–13 days (1 week to less than 2 weeks), 14–27 days (2 weeks to less than 4 weeks), 28 days or more (4 weeks or more)), and similar analyses were performed to detect for any heterogeneity. In addition, medication adherence using interval PDC for 2, 3, 4 and 5-years was used, respectively, as the outcome variable while all the analyses were repeated. Since there may exist a possibility of residual confounding, stratified analyses according to each covariate were conducted in separate regression models to explore if the findings were different. All p values b 0.05 were regarded as statistically significant.
3. Results 3.1. Participant characteristics A total of 203,259 eligible patients were included in the study (Table 1). The average age was 58.7 years (SD 17.3), with 55.9% being female subjects. The majority were not recipients of public assistance (84.9%). The most common service settings where the antihypertensive medications were prescribed were the GOPCs (31.6%), followed by the SOPCs (31.2%) and in- or day-patient clinics (26.2%). Among them, βblockers (46.7%) and CCB (28.8%) were the most frequently prescribed, followed by ACEIs (10.2%), thiazide diuretics (8.1%) and the α-blockers Table 1 Demographic characteristics of patients (N = 203,259). n (%) Age (SD) Sex Male Female Public assistance Yes No Service types In-day-patient Special-out-patient Accident & emergency General-out-patient Others Types of first prescription ACEIs Alpha-blocker Beta-blocker CCB Thiazide diuretics District of residence Hong Kong Island Kowloon New Territories Duration of first prescription 1–6 days 7–14 days 15–28 days 29 days or more
58.65 (17.32) 89,725 (44.1%) 113,533 (55.9%) 28,891 (14.2%) 172,523 (84.9%) 53,230 (26.2%) 63,455 (31.2%) 16,692 (8.2%) 64,187 (31.6%) 5,687 (2.8%) 20,691 (10.2%) 12,554 (6.2%) 95,014 (46.7%) 58,530 (28.8%) 16,470 (8.1%) 35,576 (17.5%) 67,919 (33.4%) 99,764 (49.1%) 48,114 (23.7%) 49,329 (24.3%) 30,683 (15.1%) 75,133 (37.0%)
ACEI: angiotensin converting enzyme inhibitors; CCB: calcium channel blockers.
M.C.S. Wong et al. / International Journal of Cardiology 182 (2015) 503–508
(6.2%). Significantly more patients were residents in the most rural region (The New Territories; 49.1%). The proportion of patients who had follow-up visits after the initial prescription in the 4 patient subgroups was 74.3%, 67.4%, 75.6% and 79.0%, respectively. 3.2. The interval of initial antihypertensive prescription The proportion of patients whose initial prescriptions lasted for b6 days, 7–14 days, 15–28 days, and ≥ 29 days was 23.7%, 24.3%, 15.1%, and 37.0%, respectively (Table 1; Fig. 1). Table 2 shows the distribution of the period of initial prescriptions. The median duration of the initial prescription was 21 days, and 90% of the prescriptions lasted for 84 days or less (Table 2). 3.3. The association between the duration of initial prescription and medication adherence The overall 12-month adherence rate, as defined by PDC ≥ 80%, was 32.4%. The proportion of optimal adherence was 18.1%, 20.1%, 31.0% and 50.3% among patients prescribed for b6 days, 7–14 days, 15–28 days, and ≥29 days, respectively. The mean PDC in these patients was 0.34 (SD 0.44), 0.35 (SD 0.39), 0.52 (SD 0.42) and 0.73 (SD 0.41), respectively, showing a gradual increase in PDC with longer duration of antihypertensive prescriptions (Fig. 2). When a binary logistic regression analysis was modeled with optimal adherence as the outcome variable, patients prescribed for 7–14 days (adjusted odds ratio [AOR] = 1.17, 95% C.I. 1.12–1.22); 15–28 days (AOR = 1.90, 95% C.I. 1.82–1.99) and ≥29 days (AOR = 4.13, 95% C.I. 3.96–4.31) were significantly more likely to be adherent than those who were prescribed for ≤6 days (all p b 0.001) (Table 3). The other factors associated with optimal adherence included older age (AOR = 1.02, 95% C.I. 1.01–1.02, p b 0.001); male gender (AOR for female subjects = 0.87, 95% C.I. 0.85–0.89, p b 0.001); receiving public assistance (AOR = 1.12, 95% C.I. 1.09– 1.15); and drug class being ACEIs (AOR = 1.75, 95% C.I. 1.67–1.83) or CCBs (AOR = 1.18, 95% C.I. 1.13–1.23). Service setting at the Accident and Emergency Department (AOR = 0.25, 95% C.I. 0.23–0.27, p b 0.001) and prescription of α-blockers (AOR = 0.54, 95% C.I. 0.51–0.57, p b 0.001) were significantly associated with poorer medication adherence (Table 3). 3.4. Sensitivity analyses Additional analyses excluding patients whose antihypertensive prescription lasted for N 180 days and 90 days, respectively, generated similar conclusions with respect to the positive association between the duration of initial prescription and optimal adherence (Table 4). Also, this association was found to be significant even when the duration of observation was extended for up to 5 years (Table 5). Controlling for the frequency of follow-up visits as an additional variable in sensitivity
Fig. 1. The proportion of participants according to the interval of the first prescription.
505
Table 2 Percentile table on the duration of first prescription among patients (N = 203,259). Percentiles (%)
Duration of the first prescription (days)
10 20 30 40 50 60 70 80 90
2 4 7 14 21 28 42 56 84
analysis showed that longer duration of initial prescription was associated with higher levels of optimal adherence with antihypertensive agents. The key factors affecting medication adherence did not differ between patients living in rural areas vs. urban areas. Further analyses by stratifying the outcomes according to each covariate also resulted in similar findings. When an additional group on the duration of initial prescription (≥60 days) was included, or when a different categorization of prescription for duration of prescriptions was used, similar association between prescription duration and PDC was also found. There were no variable interactions or multicollinearity in all the regression analyses, implying the robustness of the regression models. 4. Discussion 4.1. Statement of the principal findings This cohort study was conducted among a large population of Chinese patients who received their first ever antihypertensive medications. It was found that shorter duration of initial antihypertensive prescription was positively associated with poorer drug adherence in the subsequent 12 months. The strength of this finding was consistent in various sensitivity and stratification analyses. 4.2. Relationship with existing literature and explanation of the findings We have used the duration of initial prescriptions 1–6 days, 7– 14 days, 15–28 days, and ≥29 days as there were few initial prescriptions which lasted for more than 60 days in database. There were three smaller scale cohort studies which have explored on such an association. Batal and colleagues [32] assessed the relationship between the quantities of statin dispensed by the pharmacy with medication adherence by the Denver Health Medical Center in the state of Colorado. From an integrated health care system, they found that patients who received a larger quantity of statins (60 day supply) were significantly more adherent than those who received lower quantities (30 day supply). The
Fig. 2. The levels of medication adherence measured by Proportion of Days Covered (PDC) according to the duration of initial antihypertensive prescriptions.
506
M.C.S. Wong et al. / International Journal of Cardiology 182 (2015) 503–508
Table 3 Factors associated with optimal medication adherence (N = 203,259). Adjusted odds ratio (95% C.I.)
p value
Duration of first prescription
58.65 (17.32)
1.02 (1.01–1.02)
b0.001
89725 (44.1%) 113533 (55.9%)
1.00 (referent) 0.87 (0.85–0.89)
– b0.001
28891 (14.2%) 172523 (84.9%)
1.12 (1.09–1.15) 1.00 (referent)
b0.001
53230 (26.2%) 63455 (31.2%) 16692 (8.2%) 64187 (31.6%) 5687 (2.8%)
1.00 (referent) 0.98 (0.94–1.02) 0.25 (0.23–0.27) 1.03 (0.99–1.07) 1.10 (1.03–1.18)
– 0.342 b0.001 0.156 0.003
20691 (10.2%) 12554 (6.2%) 95014 (46.7%) 58530 (28.8%) 16470 (8.1%)
1.75 (1.67–1.83) 0.54 (0.51–0.57) 1.02 (0.98–1.06) 1.18 (1.13–1.23) 1.00 (referent)
b0.001 b0.001 0.338 b0.001 –
2 years 1–6 days 7–14 days 15–28 days 29 days or more 3 years 1–6 days 7–14 days 15–28 days 29 days or more 4 years 1–6 days 7–14 days 15–28 days 29 days or more 5 years 1–6 days 7–14 days 15–28 days 29 days or more
35576 (17.5%) 67919 (33.4%) 99764 (49.1%)
1.00 (0.97–1.03) 0.94 (0.92–0.96) 1.00 (referent)
0.950 b0.001 –
48114 (23.7%) 49329 (24.3%) 30683 (15.1%) 75133 (37.0%)
1.00 (referent) 1.17 (1.12–1.22) 1.90 (1.82–1.99) 4.13 (3.96–4.31)
– b0.001 b0.001 b0.001
n% Age (mean, SD) Sex Male Female Public assistance Yes No Service types In-day-patient Special-out-patient Accident & emergency General-out-patient Others Types of first prescription ACEIs Alpha-blocker Beta-blocker CCB Thiazide diuretics District of residence Hong Kong Island Kowloon New Territories Duration of first prescription 1–6 days 7–14 days 15–28 days 29 days or more
Table 5 Factors associated with optimal medication adherence at different time points.
ACEI: angiotensin converting enzyme inhibitors; CCB: calcium channel blockers.
added benefit of supplying more medications in each clinic visit also translated into better clinical outcomes, including lower levels of lipid profile. Steiner and colleagues [31] studied the association among patients in the Veterans Affairs Medical Centres in the Rocky Mountain region. They reported that patients receiving longer prescriptions (≥ 90 days vs. b90 days) were more likely to obtain at least 100% of their prescribed amounts of digoxin. It was concluded that long prescriptions may facilitate the acquisition of maintenance medications. The third study was performed by Ivers et al. [33], who conducted a population-level cohort study to examine adherence with cardiac medications for 18 months after coronary angiography in elderly patients with coronary artery disease. It was found that initial prescriptions of ACEIs/ARBs, β-blockers and statins for at least 60 days were more likely to result in high long-term adherence for all three types of medications. Moreover, in another cross-sectional study conducted among hypertensive patients in the community clinic setting, it was reported that shorter duration of antihypertensive agents prescribed were negatively associated with drug adherence measured using a Morisky Medication Adherence Scale [37]. Therefore, the findings of the present study are Table 4 Sensitivity analyses where patients who were initial prescribed lasted for N180 days and N90 days were excluded.
Duration of first prescription among those prescribed for ≤180 days (n = 202,829) 1–6 days 7–14 days 15–28 days 29 days or more Duration of first prescription among those prescribed for ≤90 days (n = 189,873) 1–6 days 7–14 days 15–28 days 29 days or more
Adjusted odds ratio (95% C.I.)
p values
1.00 (referent) 1.17 (1.12–1.22) 1.91 (1.82–1.99) 4.13 (3.95–4.31)
– b0.001 b0.001 b0.001
1.00 (referent) 1.19 (1.13–1.24) 1.95 (1.86–2.04) 4.10 (3.92–4.28)
– b0.001 b0.001 b0.001
n (%)
Adjusted odds ratio (95% C.I.)
p values
44752 (22.9%) 47899 (24.5%) 29787 (15.2%) 73351 (37.5%)
1.00 (referent) 0.94 (0.89, 0.99) 1.63 (1.55, 1.71) 3.67 (3.50, 3.85)
– 0.015 b0.001 b0.001
42739 (22.3%) 47175 (24.7%) 29309 (15.3%) 72147 (37.7%)
1.00 (referent) 1.20 (1.13, 1.27) 1.56 (1.47, 1.65) 2.33 (2.21, 2.46)
– b0.001 b0.001 b0.001
41000 (21.9%) 46501 (24.8%) 28825 (15.4%) 70933 (37.9%)
1.00 (referent) 1.21 (1.13, 1.30) 1.60 (1.50, 1.72) 2.56 (2.40, 2.74)
– b0.001 b0.001 b0.001
38741 (21.6%) 44434 (24.8%) 27614 (15.4%) 68406 (38.2%)
1.00 (referent) 1.11 (1.01, 1.22) 1.64 (1.50, 1.80) 2.64 (2.42, 2.88)
– 0.024 b0.001 b0.001
consistent with the existing literature, and due to its larger sample size these findings can be regarded as confirmatory evaluations of this positive association between shorter duration of initial antihypertensive prescription and poorer drug adherence. There is one obvious reason which could explain the better adherence profiles among patients with their initial antihypertensive agents prescribed for longer periods [31–33]. Longer prescriptions could partly address one of the major contributors to non-adherence, namely forgetfulness [38]. This refers to the observation that medications prescribed for longer periods may facilitate the patients to follow a regular medication regimen — by reducing the likelihood of their medications running out of stock. Shorter prescriptions might create an avoidable inconvenience especially for elderly patients [33], and a higher frequency of pharmacy refill — the costs of which are paid by patients. If the level of continuity of information exchange among physicians is low, the transparency in the transfer of responsibility for refill of medications could be hampered [39]. Yet one more possibility is that shorter prescriptions may be falsely perceived by patients and their family members, that long-term adherence is not required [33]. 4.3. Study strengths and limitations To the best of our knowledge, this study is the first-ever study to show that longer initial prescriptions were associated with better medication adherence among the Chinese population. It involved a large number of hypertensive patients who were antihypertensive drugnaive. The inclusion of all eligible patients in the public healthcare sector of the entire population in the region further enhanced its representativeness and generalizability to patient care. The high quality prescribing and dispensing practices in the public clinics in Hong Kong increased the reliability of the findings. However, some limitations should be addressed here. Firstly, the PDC ratio is a measure of intent to treatment rather than actual medication bottle audits; albeit the latter cannot assure that a dose that was removed was actually consumed or administered correctly. Thus there is one inherent assumption of database research — namely patients were taking the drugs as exactly recorded in the study dataset. Patients with longer duration of initial prescriptions would automatically have higher PDC, when compared to those whose initial prescriptions were of shorter duration. Also, this study evaluated PDC in a 12 month interval, which was chosen because the first year of initiating antihypertensive drugs represents the most crucial period predicting future adherence. Thirdly, it is worthy of note that the percentage of patients who were first treated with alpha- or
M.C.S. Wong et al. / International Journal of Cardiology 182 (2015) 503–508
beta-blocker in this study was relatively high, and these drugs are known to cause unpleasant side effects particularly in the first few days of treatment. Thus the high degree of failure to adhere after a short-term prescription might be related to the active choice by the patients who discontinued their therapy due to the side effects. In addition, this observational study is not a randomized trial and there may exist differences in baseline characteristics of the study participants or residual confounding, despite our efforts to control confounder variables using regression models and stratified analyses. Some variables like patients' educational level, blood pressure control and medical history were not available, and they might influence medication adherence. Lastly, the study findings might not be generalizable to Chinese patients living in other parts of China, due to differences in health literacy and other clinical characteristics. 4.4. Conclusions: implications to clinical practice and future research In summary, this study shows that the duration of initial antihypertensive prescriptions for newly diagnosed hypertensive patients should not be too short to minimize subsequent medication non-adherence. This finding bears a substantial implication for the prescribing decisions among clinicians and policy-making for pharmacy refill systems [31]. Longer duration of antihypertensive prescriptions for medicationnaive patients may represent a new strategy to modify long-term adherence [33]. This conclusion should, however, be cautiously weighed against some potential harm caused by very long prescription periods — which could lead to medication wastage and minimize co-payment collection [32]. Too long a prescription is also a concern as some patients might warrant an earlier follow-up for further assessment and review of medication side-effects. The major message of this study is, therefore, to call for this balance when physicians consider to prescribe antihypertensive agents for newly diagnosed hypertensive patients. Policy and system changes required to make dispensing larger amounts costeffective are needed [32]. Future studies should evaluate the difference in longer-term adherence and other important clinical outcomes between short and long initial prescriptions, preferably in trial settings. Grant support None. Conflict of interest The authors report no relationships that could be construed as a conflict of interest. Acknowledgments We express our gratitude for all the healthcare professionals who entered the data into the clinical database. We thank the Hospital Authority of the Hong Kong Government for allowing our research team to use the database. References [1] S. Yusuf, S. Reddy, S. Ounpuu, S. Anand, Global burden of cardiovascular diseases: part I: general considerations, the epidemiologic transition, risk factors, and impact of urbanization, Circulation 104 (2001) 2746–2753. [2] C.M.M. Lawes, S. Vander Hoorn, A. Rodgers, I.S. Hypertens, Global burden of bloodpressure-related disease, 2001, Lancet 371 (2008) 1513–1518. [3] H.H.X. Wang, J.J. Wang, S.Y.S. Wong, M.C.S. Wong, F.J. Li, P.X. Wang, et al., Epidemiology of multimorbidity in China and implications for the healthcare system: crosssectional survey among 162,464 community household residents in southern China, BMC Medicine 12 (2014) 188. [4] M.C.S. Wong, H.H.X. Wang, M.C.M. Leung, C.S.H. Tsang, S.V. Lo, S.M. Griffiths, The rising prevalence of self-reported hypertension among Chinese subjects: a population-based study from 121 895 household interviews, QJM-An Int J Med 108 (2014) 9–17.
507
[5] D. Lloyd-Jones, R. Adams, M. Carnethon, G. De Simone, T.B. Ferguson, K. Flegal, et al., Heart disease and stroke statistics—2009 update: a report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee, Circulation 119 (2009) 480–486. [6] K. Wolf-Maier, R.S. Cooper, H. Kramer, J.R. Banegas, S. Giampaoli, M.R. Joffres, et al., Hypertension treatment and control in five European Countries, Canada, and the United States, Hypertension 43 (2004) 10–17. [7] T. Dorjgochoo, X.O. Shu, X. Zhang, H. Li, G. Yang, L. Gao, et al., Relation of blood pressure components and categories and all-cause, stroke and coronary heart disease mortality in urban Chinese women: a population-based prospective study, J. Hypertens. 27 (2009) 468–475. [8] Blood Pressure Lowering Treatment Trialists CF. Turnbull, B. Neal, T. Ninomiya, C. Algert, H. Arima, et al., Effects of different regimens to lower blood pressure on major cardiovascular events in older and younger adults: meta-analysis of randomised trials, BMJ 336 (2008) 1121–1123. [9] P.M. Ho, C.L. Bryson, J.S. Rumsfeld, Medication adherence its importance in cardiovascular outcomes, Circulation 119 (2009) 3028–3035. [10] G. Mazzaglia, E. Ambrosioni, M. Alacqua, A. Filippi, E. Sessa, V. Immordino, et al., Adherence to antihypertensive medications and cardiovascular morbidity among newly diagnosed hypertensive patients, Circulation 120 (2009) 1598–1605. [11] E. Sabate, World Health Organization, WHO Adherence to Long Term Therapies Project., Global Adherence Interdisciplinary Network. Adherence to Long-Term Therapies: Evidence for Action, World Health Organization, Geneva, 2003. [12] B. Vrijens, G. Vincze, P. Kristanto, J. Urquhart, M. Burnier, Adherence to prescribed antihypertensive drug treatments: longitudinal study of electronically compiled dosing histories, BMJ 336 (2008) 1114–1117. [13] B.L. van Wijk, W.H. Shrank, O.H. Klungel, S. Schneeweiss, M.A. Brookhart, J. Avorn, A cross-national study of the persistence of antihypertensive medication use in the elderly, J. Hypertens. 26 (2008) 145–153. [14] J.J. Caro, M. Salas, J.L. Speckman, G. Raggio, J.D. Jackson, Persistence with treatment for hypertension in actual practice, CMAJ 160 (1999) 31–37. [15] C. Bourgault, M. Senecal, M. Brisson, M.A. Marentette, J.P. Gregoire, Persistence and discontinuation patterns of antihypertensive therapy among newly treated patients: a population-based study, J. Hum. Hypertens. 19 (2005) 607–613. [16] J. Mar, F. Rodriguez-Artalejo, Which is more important for the efficiency of hypertension treatment: hypertension stage, type of drug or therapeutic compliance? J. Hypertens. 19 (2001) 149–155. [17] D.A. Hughes, A. Bagust, A. Haycox, T. Walley, The impact of non-compliance on the cost-effectiveness of pharmaceuticals: a review of the literature, Health Econ. 10 (2001) 601–615. [18] G. Mancia, R. Fagard, K. Narkiewicz, J. Redon, A. Zanchetti, M. Bohm, et al., 2013 ESH/ ESC guidelines for the management of arterial hypertension: the Task Force for the Management of Arterial Hypertension of the European Society of Hypertension (ESH) and of the European Society of Cardiology (ESC), Eur. Heart J. 34 (2013) 2159–2219. [19] M.C. Sokol, K.A. McGuigan, R.R. Verbrugge, R.S. Epstein, Impact of medication adherence on hospitalization risk and healthcare cost, Med. Care 43 (2005) 521–530. [20] S. Baroletti, H. Dell'Orfano, Medication adherence in cardiovascular disease, Circulation 121 (2010) 1455–1458. [21] M.C.S. Wong, J.Y. Jiang, J.L. Tang, A. Lam, H. Fung, S.W. Mercer, Health services research in the public healthcare system in Hong Kong: an analysis of over 1 million antihypertensive prescriptions between 2004–2007 as an example of the potential and pitfalls of using routinely collected electronic patient data, BMC Health Serv. Res. 8 (2008) 138. [22] M.C.S. Wong, J.Y. Jiang, S.M. Griffiths, Factors associated with compliance, discontinuation and switching of calcium channel blockers in 20,156 Chinese patients, Am. J. Hypertens. 22 (2009) 904–910. [23] M.C.S. Wong, J.Y. Jiang, S.M. Griffiths, Antihypertensive drug adherence among 6408 Chinese patients on angiotensin-converting enzyme inhibitors in Hong Kong: a cohort study, J. Clin. Pharmacol. 50 (2010) 598–605. [24] M.C.S. Wong, J.Y. Jiang, S.M. Griffiths, Adherence to combination therapy among ethnic Chinese patients: a cohort study, Hypertens. Res. 33 (2010) 416–421. [25] M.C.S. Wong, J.Y. Jiang, S.M. Griffiths, Factors associated with compliance to thiazide diuretics among 8551 Chinese patients, J. Clin. Pharm. Ther. 36 (2011) 179–186. [26] M.C.S. Wong, W.W.S. Tam, C.S.K. Cheung, E.L.H. Tong, A.C.H. Sek, N.T. Cheung, et al., Medication adherence to first-line antihypertensive drug class in a large Chinese population, Int. J. Cardiol. 167 (2013) 1438–1442. [27] M.C.S. Wong, W.W.S. Tam, C.S.K. Cheung, E.L.H. Tong, A.C.H. Sek, N.T. Cheung, et al., Antihypertensive prescriptions over a 10-year period in a large Chinese population, Am. J. Hypertens. 26 (2013) 931–938. [28] M.C.S. Wong, W.W.S. Tam, C.S.K. Cheung, H.H.X. Wang, E.L.H. Tong, A.C.H. Sek, et al., Drug adherence and the incidence of coronary heart disease- and stroke-specific mortality among 218,047 patients newly prescribed an antihypertensive medication: a five-year cohort study, Int. J. Cardiol. 168 (2013) 928–933. [29] M.C.S. Wong, W.W.S. Tam, H.H.X. Wang, C.S.K. Cheung, E.L.H. Tong, A.C.H. Sek, et al., Predictors of the incidence of all-cause mortality and deaths due to diabetes and renal diseases among patients newly prescribed antihypertensive agents: a cohort study, Int. J. Cardiol. 168 (2013) 4705–4710. [30] M.C.S. Wong, J.Y. Jiang, S.M. Griffiths, Short-term adherence to beta-blocker therapy among ethnic Chinese patients with hypertension: a cohort study, Clin. Ther. 31 (2009) 2170–2177. [31] M.C.S. Wong, C.H.M. Wu, H.H.X. Wang, H.W. Li, E.M.T. Hui, A.T. Lam, et al., Association between the 8-item Morisky Medication Adherence Scale (MMAS-8) score and glycaemic control among Chinese diabetes patients, J. Clin. Pharmacol. 55 (2015) 279–287.
508
M.C.S. Wong et al. / International Journal of Cardiology 182 (2015) 503–508
[32] H.A. Batal, M.J. Krantz, R.A. Dale, P.S. Mehler, J.F. Steiner, Impact of prescription size on statin adherence and cholesterol levels, BMC Health Serv. Res. 7 (2007) 175. [33] N.M. Ivers, J.D. Schwalm, C.A. Jackevicius, H. Guo, J.V. Tu, M. Natarajan, Length of initial prescription at hospital discharge and long-term medication adherence for elderly patients with coronary artery disease: a population-level study, Can. J. Cardiol. 29 (2013) 1408–1414. [34] N.K. Choudhry, W.H. Shrank, R.L. Levin, J.L. Lee, S.A. Jan, M.A. Brookhart, et al., Measuring concurrent adherence to multiple related medications, Am. J. Manage. Care 15 (2009) 457–464. [35] B.C. Martin, E.K. Wiley-Exley, S. Richards, M.E. Domino, T.S. Carey, B.L. Sleath, Contrasting measures of adherence with simple drug use, medication switching, and therapeutic duplication, Ann. Pharmacother. 43 (2009) 36–44.
[36] S.E. Andrade, K.H. Kahler, F. Frech, K.A. Chan, Methods for evaluation of medication adherence and persistence using automated databases, Pharmacoepidemiol. Drug Saf. 15 (2006) 565–577. [37] G.K.Y. Lee, H.H.X. Wang, K.Q.L. Liu, Y. Cheung, D.E. Morisky, M.C.S. Wong, Determinants of medication adherence to antihypertensive medications among a Chinese population using Morisky Medication Adherence Scale, PLoS ONE 8 (2013) e62775. [38] L. Osterberg, T. Blaschke, Adherence to medication, N. Engl. J. Med. 353 (2005) 487–497. [39] S. Kripalani, F. LeFevre, C.O. Phillips, M.V. Williams, P. Basaviah, D.W. Baker, Deficits in communication and information transfer between hospital-based and primary care physicians: implications for patient safety and continuity of care, JAMA 297 (2007) 831–841.
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Duration of initial antihypertensive prescription and medication adherence: A cohort study among 203,259 newly diagnosed hypertensive patients☆ Martin C.S. Wong a, Wilson W.S. Tam a,b, Harry H.X. Wang a, Clement S.K. Cheung c, Ellen L.H. Tong c, N.T. Cheung c, Stephen R. Leeder d, Sian M. Griffiths a,⁎ a
JC School of Public Health and Primary Care, Faculty of Medicine, Chinese University of Hong Kong, Hong Kong, China Alice Lee Centre for Nursing Studies, National University of Singapore, Singapore c Hospital Authority Information Technology Services — Health Informatics Section, Hong Kong, China d Menzies Centre for Health Policy, University of Sydney, Australia b
a r t i c l e
i n f o
Article history: Received 25 August 2014 Received in revised form 25 November 2014 Accepted 21 December 2014 Available online 23 December 2014 Keywords: Duration Initial antihypertensive prescription Medication adherence Cohort study Chinese population
a b s t r a c t Background: Optimal adherence with antihypertensive medications is crucial to prevent hypertension-related complications. This study evaluated whether the duration of initial antihypertensive prescription is associated with better medication adherence in a large sample of Chinese hypertensive patients. Methods and results: From a validated clinical database which consists of all patients in the public healthcare sector in Hong Kong, all patients on their first-ever antihypertensive agent from 2001 to 2005 (N = 203,259) were included and followed-up for 12 months (and up to 5 years in separate analyses). The average age was 58.7 years (SD 17.3), and the overall rate of optimal adherence (as measured by having the Proportion of Days Covered ≥ 0.80) was 32.4%. The proportion of patients whose initial prescriptions lasted for ≤6 days; 7–14 days; 15–28 days and ≥29 days was 23.7%, 24.3%, 15.1% and 37.0%, respectively. The corresponding proportion of optimal adherence was 18.1%, 20.1%, 31.0% and 50.3%. The binary logistic regression analysis showed that after controlling for age, sex, socioeconomic status, service type, drug class, and district of residence, those whose initial prescription was 7–14 days (adjusted odds ratio [AOR] = 1.17, 95% C.I. 1.12–1.22); 15–28 days (AOR = 1.90, 95% C.I. 1.82–1.99) and ≥29 days (AOR = 4.13, 95% C.I. 3.96–4.31) were significantly more likely to be adherent than those who were prescribed for ≤6 days (all p b 0.001). These findings remained significant in separate analyses where the period of follow-up was extended to 5 years. Conclusions: Shorter duration of first antihypertensive prescriptions was associated with poorer medication adherence, and this practice should be avoided if possible. © 2014 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Hypertension is currently the most common chronic disease in clinical practice, and carries a substantial global burden to patients, physicians and the healthcare system [1]. Worldwide, around 54% of stroke and 47% of ischemic heart disease were due to hypertension alone [2]. Hypertensive patients are more prone to develop concomitant complications, including cardiovascular diseases, diabetes, retinopathy and renal failure [2,3]. Its prevalence is rising in different parts of the world, while its control rate remains low [4–7].
☆ Statement of authorship: All authors take responsibility for all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation. ⁎ Corresponding author at: 2/F, School of Public Health and Primary Care, Prince of Wales Hospital, Chinese University of Hong Kong, Shatin, NT, Hong Kong, China. E-mail address: siangriffi[email protected] (S.M. Griffiths).
http://dx.doi.org/10.1016/j.ijcard.2014.12.058 0167-5273/© 2014 Elsevier Ireland Ltd. All rights reserved.
The effectiveness of antihypertensive therapies to protect against stroke and heart attack has been extensively documented [8,9]. Mazzaglia and colleagues [10] showed that higher antihypertensive adherence was associated with a 38% lower risk of acute cardiovascular events, from a cohort of more than 18,800 newly diagnosed hypertensive patients. It demonstrated that even among those who had not suffered from any previous events, optimal antihypertensive adherence confers primary prevention to a substantial extent. However, it is widely recognized that a large number of hypertensive patients do not take antihypertensive therapies as prescribed [11], and adherence with prescriptions declined with time — especially in the first year of drug initiation [12–15]. Poor medication adherence may also lead to increased morbidity, mortality, hospitalization admissions and escalated healthcare costs [16–20]. Previous studies on the determinants of medication adherence mostly focused on patients' socio-demographic factors and the service settings [21–30]. The World Health Organization (WHO) has identified
504
M.C.S. Wong et al. / International Journal of Cardiology 182 (2015) 503–508
evidence for early action to enhance adherence with long-term therapies, and antihypertensive prescription is among one of them [11]. Three recent small studies investigated the association between the duration of physician prescription and long-term adherence with cardiovascular medications [31–33], and this raised a knowledge gap of whether the duration of initial antihypertensive drugs prescribed may have an effect on subsequent adherence. In light of these studies, we tested the a priori hypothesis that shorter duration of the initial antihypertensive prescription among newly diagnosed hypertensive patients was associated with poorer medication adherence in the subsequent 12 months of their followup, based on the findings of the existing literature which implied that such an association might exist [31–33]. 2. Methods 2.1. Ethics statement This study was approved by the Clinical Ethics Research Committee of the Hospital Authority, and the Survey and Behavioral Research Ethics Committee of The Chinese University of Hong Kong. The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki. Informed consent was not required as all subjects were anonymized with unique identifiers. 2.2. Data source This study used an electronic Clinical Management System (CMS) in Hong Kong, which has been previously described in details [21–30]. The electronic healthcare records capture information on patients' demographics, prescription details, and clinical diagnoses, which were coded by the International Classification of Diseases (ICD-9 and ICD-10) or the International Classification of Primary Care (ICPC-2). It is a computerized system adopted in all clinics in the public healthcare sector which serves as the major source of health care in Hong Kong. The CMS has been the only source of data entry, allowing the linkage of physician-entered information in each patient visit for all public clinical settings situated in all districts of Hong Kong. All prescriptions are entered by the consulting physicians, and are duly checked by the dispensers or pharmacists according to standardized procedures. Any amendments to the medications prescribed after the clinical consultations are also documented in the system. This database was found to be a validated research platform, and it has a high level of data completeness in terms of sociodemographic information (100%) and prescription details (99.8%) [21]. The present study extracted information from the entire Hong Kong population, which is more than 7 million in 2014. Hong Kong is divided into three distinct regions, namely Hong Kong Island (most urbanized); Kowloon; and the New Territories (most rural). 2.3. Patients We included all patients who had a clinical consultation at any public practices and who were prescribed their first-ever antihypertensive agents between the calendar years 2001 and 2005. These consisted of α-blockers, β-blockers, calcium channel blockers (CCBs), thiazide diuretics, and angiotensin converting enzyme inhibitors (ACEIs). Patients who received angiotensin receptor blockers (ARBs), polytherapy and fixed dose combination therapies were not analyzed due to the small number of such prescriptions. Patients who were prescribed any antihypertensive medications before index date or died 12 months after the index date were also excluded. All patients were followed up for 12 months and their medication adherence in the study period was evaluated. 2.4. Outcome variables and covariates The primary outcome variable was the adherence level 12 months after the initial prescription. We used the Proportion of Days Covered (PDC) as the proxy measure for medication adherence [34]. PDC is defined as the number of days where medication is supplied in a specified period divided by the total number of days within the period. Interval-based PDC is widely recognized as an internationally accepted metric for assessment of medication adherence in large database studies [34–36]. PDC was used as the measure for medication adherence since it was adopted extensively in published studies, and tends to be operationally defined more consistently than other adherence measures [34]. Its use also facilitates comparison of our study findings with existing literature. PDC ≥ 80% indicated that the patient is medication adherent [34,35], similar to an approach adopted by Mazzaglia and colleagues [10] to classify drug adherence. The variable tested for association with the outcome was the duration of the initial prescription, defined as the number of days supplied in the initial antihypertensive medication as recorded in the computerized system. We made reference to existing literature which compared the length of prescriptions [33]. Since less than 20% of the initial prescriptions among all patients lasted for more than 60 days, we divided the prescription durations into four distinct groups: ≤6 days; 7–14 days; 15–28 days; and ≥29 days. The covariates included patients' age, sex, payment status (recipients of comprehensive social security assistance did not need to pay; and for non-recipients each one-off consultation
costs approximately US$5.77, including both investigation and prescription fees regardless of the prescription duration), the service settings (in- or day-patient clinics; specialist outpatient clinics [SOPCs], accident and emergency departments (AEDs); general outpatient clinic [GOPC], and other clinic type), the district of residence (Hong Kong Island vs. Kowloon vs. the New Territories) and the first antihypertensive drug class received. 2.5. Statistical analyses The Statistical Package for Social Sciences version 16.0 (SPSS, Inc., Chicago, Illinois) was used for all data analyses. Student's t tests and chi-square tests of heterogeneity were used to compare continuous and categorical variables, respectively. The sociodemographic characteristics were compared among patients according to the duration of the initial prescription. A binary logistic regression model was constructed with the optimal medication adherence (PDC ≥ 80%) as the outcome variable. The explanatory variable (prescription length) and all the covariates listed above were entered into the regression analysis. We evaluated for multicollinearity and interactions among variables to ensure the robustness of the regression model. As sensitivity analyses, we (1) excluded patients whose initial prescriptions were N90 days and N180 days, respectively; (2) included the frequency of follow-up visits as an additional variable; (3) stratified our patients according to the district of residence (rural vs. urban regions);(4) included one more category of the prescription duration (≥60 days); and (5) used a different categorization for duration of prescriptions (1–6 days (less than 1 week), 7–13 days (1 week to less than 2 weeks), 14–27 days (2 weeks to less than 4 weeks), 28 days or more (4 weeks or more)), and similar analyses were performed to detect for any heterogeneity. In addition, medication adherence using interval PDC for 2, 3, 4 and 5-years was used, respectively, as the outcome variable while all the analyses were repeated. Since there may exist a possibility of residual confounding, stratified analyses according to each covariate were conducted in separate regression models to explore if the findings were different. All p values b 0.05 were regarded as statistically significant.
3. Results 3.1. Participant characteristics A total of 203,259 eligible patients were included in the study (Table 1). The average age was 58.7 years (SD 17.3), with 55.9% being female subjects. The majority were not recipients of public assistance (84.9%). The most common service settings where the antihypertensive medications were prescribed were the GOPCs (31.6%), followed by the SOPCs (31.2%) and in- or day-patient clinics (26.2%). Among them, βblockers (46.7%) and CCB (28.8%) were the most frequently prescribed, followed by ACEIs (10.2%), thiazide diuretics (8.1%) and the α-blockers Table 1 Demographic characteristics of patients (N = 203,259). n (%) Age (SD) Sex Male Female Public assistance Yes No Service types In-day-patient Special-out-patient Accident & emergency General-out-patient Others Types of first prescription ACEIs Alpha-blocker Beta-blocker CCB Thiazide diuretics District of residence Hong Kong Island Kowloon New Territories Duration of first prescription 1–6 days 7–14 days 15–28 days 29 days or more
58.65 (17.32) 89,725 (44.1%) 113,533 (55.9%) 28,891 (14.2%) 172,523 (84.9%) 53,230 (26.2%) 63,455 (31.2%) 16,692 (8.2%) 64,187 (31.6%) 5,687 (2.8%) 20,691 (10.2%) 12,554 (6.2%) 95,014 (46.7%) 58,530 (28.8%) 16,470 (8.1%) 35,576 (17.5%) 67,919 (33.4%) 99,764 (49.1%) 48,114 (23.7%) 49,329 (24.3%) 30,683 (15.1%) 75,133 (37.0%)
ACEI: angiotensin converting enzyme inhibitors; CCB: calcium channel blockers.
M.C.S. Wong et al. / International Journal of Cardiology 182 (2015) 503–508
(6.2%). Significantly more patients were residents in the most rural region (The New Territories; 49.1%). The proportion of patients who had follow-up visits after the initial prescription in the 4 patient subgroups was 74.3%, 67.4%, 75.6% and 79.0%, respectively. 3.2. The interval of initial antihypertensive prescription The proportion of patients whose initial prescriptions lasted for b6 days, 7–14 days, 15–28 days, and ≥ 29 days was 23.7%, 24.3%, 15.1%, and 37.0%, respectively (Table 1; Fig. 1). Table 2 shows the distribution of the period of initial prescriptions. The median duration of the initial prescription was 21 days, and 90% of the prescriptions lasted for 84 days or less (Table 2). 3.3. The association between the duration of initial prescription and medication adherence The overall 12-month adherence rate, as defined by PDC ≥ 80%, was 32.4%. The proportion of optimal adherence was 18.1%, 20.1%, 31.0% and 50.3% among patients prescribed for b6 days, 7–14 days, 15–28 days, and ≥29 days, respectively. The mean PDC in these patients was 0.34 (SD 0.44), 0.35 (SD 0.39), 0.52 (SD 0.42) and 0.73 (SD 0.41), respectively, showing a gradual increase in PDC with longer duration of antihypertensive prescriptions (Fig. 2). When a binary logistic regression analysis was modeled with optimal adherence as the outcome variable, patients prescribed for 7–14 days (adjusted odds ratio [AOR] = 1.17, 95% C.I. 1.12–1.22); 15–28 days (AOR = 1.90, 95% C.I. 1.82–1.99) and ≥29 days (AOR = 4.13, 95% C.I. 3.96–4.31) were significantly more likely to be adherent than those who were prescribed for ≤6 days (all p b 0.001) (Table 3). The other factors associated with optimal adherence included older age (AOR = 1.02, 95% C.I. 1.01–1.02, p b 0.001); male gender (AOR for female subjects = 0.87, 95% C.I. 0.85–0.89, p b 0.001); receiving public assistance (AOR = 1.12, 95% C.I. 1.09– 1.15); and drug class being ACEIs (AOR = 1.75, 95% C.I. 1.67–1.83) or CCBs (AOR = 1.18, 95% C.I. 1.13–1.23). Service setting at the Accident and Emergency Department (AOR = 0.25, 95% C.I. 0.23–0.27, p b 0.001) and prescription of α-blockers (AOR = 0.54, 95% C.I. 0.51–0.57, p b 0.001) were significantly associated with poorer medication adherence (Table 3). 3.4. Sensitivity analyses Additional analyses excluding patients whose antihypertensive prescription lasted for N 180 days and 90 days, respectively, generated similar conclusions with respect to the positive association between the duration of initial prescription and optimal adherence (Table 4). Also, this association was found to be significant even when the duration of observation was extended for up to 5 years (Table 5). Controlling for the frequency of follow-up visits as an additional variable in sensitivity
Fig. 1. The proportion of participants according to the interval of the first prescription.
505
Table 2 Percentile table on the duration of first prescription among patients (N = 203,259). Percentiles (%)
Duration of the first prescription (days)
10 20 30 40 50 60 70 80 90
2 4 7 14 21 28 42 56 84
analysis showed that longer duration of initial prescription was associated with higher levels of optimal adherence with antihypertensive agents. The key factors affecting medication adherence did not differ between patients living in rural areas vs. urban areas. Further analyses by stratifying the outcomes according to each covariate also resulted in similar findings. When an additional group on the duration of initial prescription (≥60 days) was included, or when a different categorization of prescription for duration of prescriptions was used, similar association between prescription duration and PDC was also found. There were no variable interactions or multicollinearity in all the regression analyses, implying the robustness of the regression models. 4. Discussion 4.1. Statement of the principal findings This cohort study was conducted among a large population of Chinese patients who received their first ever antihypertensive medications. It was found that shorter duration of initial antihypertensive prescription was positively associated with poorer drug adherence in the subsequent 12 months. The strength of this finding was consistent in various sensitivity and stratification analyses. 4.2. Relationship with existing literature and explanation of the findings We have used the duration of initial prescriptions 1–6 days, 7– 14 days, 15–28 days, and ≥29 days as there were few initial prescriptions which lasted for more than 60 days in database. There were three smaller scale cohort studies which have explored on such an association. Batal and colleagues [32] assessed the relationship between the quantities of statin dispensed by the pharmacy with medication adherence by the Denver Health Medical Center in the state of Colorado. From an integrated health care system, they found that patients who received a larger quantity of statins (60 day supply) were significantly more adherent than those who received lower quantities (30 day supply). The
Fig. 2. The levels of medication adherence measured by Proportion of Days Covered (PDC) according to the duration of initial antihypertensive prescriptions.
506
M.C.S. Wong et al. / International Journal of Cardiology 182 (2015) 503–508
Table 3 Factors associated with optimal medication adherence (N = 203,259). Adjusted odds ratio (95% C.I.)
p value
Duration of first prescription
58.65 (17.32)
1.02 (1.01–1.02)
b0.001
89725 (44.1%) 113533 (55.9%)
1.00 (referent) 0.87 (0.85–0.89)
– b0.001
28891 (14.2%) 172523 (84.9%)
1.12 (1.09–1.15) 1.00 (referent)
b0.001
53230 (26.2%) 63455 (31.2%) 16692 (8.2%) 64187 (31.6%) 5687 (2.8%)
1.00 (referent) 0.98 (0.94–1.02) 0.25 (0.23–0.27) 1.03 (0.99–1.07) 1.10 (1.03–1.18)
– 0.342 b0.001 0.156 0.003
20691 (10.2%) 12554 (6.2%) 95014 (46.7%) 58530 (28.8%) 16470 (8.1%)
1.75 (1.67–1.83) 0.54 (0.51–0.57) 1.02 (0.98–1.06) 1.18 (1.13–1.23) 1.00 (referent)
b0.001 b0.001 0.338 b0.001 –
2 years 1–6 days 7–14 days 15–28 days 29 days or more 3 years 1–6 days 7–14 days 15–28 days 29 days or more 4 years 1–6 days 7–14 days 15–28 days 29 days or more 5 years 1–6 days 7–14 days 15–28 days 29 days or more
35576 (17.5%) 67919 (33.4%) 99764 (49.1%)
1.00 (0.97–1.03) 0.94 (0.92–0.96) 1.00 (referent)
0.950 b0.001 –
48114 (23.7%) 49329 (24.3%) 30683 (15.1%) 75133 (37.0%)
1.00 (referent) 1.17 (1.12–1.22) 1.90 (1.82–1.99) 4.13 (3.96–4.31)
– b0.001 b0.001 b0.001
n% Age (mean, SD) Sex Male Female Public assistance Yes No Service types In-day-patient Special-out-patient Accident & emergency General-out-patient Others Types of first prescription ACEIs Alpha-blocker Beta-blocker CCB Thiazide diuretics District of residence Hong Kong Island Kowloon New Territories Duration of first prescription 1–6 days 7–14 days 15–28 days 29 days or more
Table 5 Factors associated with optimal medication adherence at different time points.
ACEI: angiotensin converting enzyme inhibitors; CCB: calcium channel blockers.
added benefit of supplying more medications in each clinic visit also translated into better clinical outcomes, including lower levels of lipid profile. Steiner and colleagues [31] studied the association among patients in the Veterans Affairs Medical Centres in the Rocky Mountain region. They reported that patients receiving longer prescriptions (≥ 90 days vs. b90 days) were more likely to obtain at least 100% of their prescribed amounts of digoxin. It was concluded that long prescriptions may facilitate the acquisition of maintenance medications. The third study was performed by Ivers et al. [33], who conducted a population-level cohort study to examine adherence with cardiac medications for 18 months after coronary angiography in elderly patients with coronary artery disease. It was found that initial prescriptions of ACEIs/ARBs, β-blockers and statins for at least 60 days were more likely to result in high long-term adherence for all three types of medications. Moreover, in another cross-sectional study conducted among hypertensive patients in the community clinic setting, it was reported that shorter duration of antihypertensive agents prescribed were negatively associated with drug adherence measured using a Morisky Medication Adherence Scale [37]. Therefore, the findings of the present study are Table 4 Sensitivity analyses where patients who were initial prescribed lasted for N180 days and N90 days were excluded.
Duration of first prescription among those prescribed for ≤180 days (n = 202,829) 1–6 days 7–14 days 15–28 days 29 days or more Duration of first prescription among those prescribed for ≤90 days (n = 189,873) 1–6 days 7–14 days 15–28 days 29 days or more
Adjusted odds ratio (95% C.I.)
p values
1.00 (referent) 1.17 (1.12–1.22) 1.91 (1.82–1.99) 4.13 (3.95–4.31)
– b0.001 b0.001 b0.001
1.00 (referent) 1.19 (1.13–1.24) 1.95 (1.86–2.04) 4.10 (3.92–4.28)
– b0.001 b0.001 b0.001
n (%)
Adjusted odds ratio (95% C.I.)
p values
44752 (22.9%) 47899 (24.5%) 29787 (15.2%) 73351 (37.5%)
1.00 (referent) 0.94 (0.89, 0.99) 1.63 (1.55, 1.71) 3.67 (3.50, 3.85)
– 0.015 b0.001 b0.001
42739 (22.3%) 47175 (24.7%) 29309 (15.3%) 72147 (37.7%)
1.00 (referent) 1.20 (1.13, 1.27) 1.56 (1.47, 1.65) 2.33 (2.21, 2.46)
– b0.001 b0.001 b0.001
41000 (21.9%) 46501 (24.8%) 28825 (15.4%) 70933 (37.9%)
1.00 (referent) 1.21 (1.13, 1.30) 1.60 (1.50, 1.72) 2.56 (2.40, 2.74)
– b0.001 b0.001 b0.001
38741 (21.6%) 44434 (24.8%) 27614 (15.4%) 68406 (38.2%)
1.00 (referent) 1.11 (1.01, 1.22) 1.64 (1.50, 1.80) 2.64 (2.42, 2.88)
– 0.024 b0.001 b0.001
consistent with the existing literature, and due to its larger sample size these findings can be regarded as confirmatory evaluations of this positive association between shorter duration of initial antihypertensive prescription and poorer drug adherence. There is one obvious reason which could explain the better adherence profiles among patients with their initial antihypertensive agents prescribed for longer periods [31–33]. Longer prescriptions could partly address one of the major contributors to non-adherence, namely forgetfulness [38]. This refers to the observation that medications prescribed for longer periods may facilitate the patients to follow a regular medication regimen — by reducing the likelihood of their medications running out of stock. Shorter prescriptions might create an avoidable inconvenience especially for elderly patients [33], and a higher frequency of pharmacy refill — the costs of which are paid by patients. If the level of continuity of information exchange among physicians is low, the transparency in the transfer of responsibility for refill of medications could be hampered [39]. Yet one more possibility is that shorter prescriptions may be falsely perceived by patients and their family members, that long-term adherence is not required [33]. 4.3. Study strengths and limitations To the best of our knowledge, this study is the first-ever study to show that longer initial prescriptions were associated with better medication adherence among the Chinese population. It involved a large number of hypertensive patients who were antihypertensive drugnaive. The inclusion of all eligible patients in the public healthcare sector of the entire population in the region further enhanced its representativeness and generalizability to patient care. The high quality prescribing and dispensing practices in the public clinics in Hong Kong increased the reliability of the findings. However, some limitations should be addressed here. Firstly, the PDC ratio is a measure of intent to treatment rather than actual medication bottle audits; albeit the latter cannot assure that a dose that was removed was actually consumed or administered correctly. Thus there is one inherent assumption of database research — namely patients were taking the drugs as exactly recorded in the study dataset. Patients with longer duration of initial prescriptions would automatically have higher PDC, when compared to those whose initial prescriptions were of shorter duration. Also, this study evaluated PDC in a 12 month interval, which was chosen because the first year of initiating antihypertensive drugs represents the most crucial period predicting future adherence. Thirdly, it is worthy of note that the percentage of patients who were first treated with alpha- or
M.C.S. Wong et al. / International Journal of Cardiology 182 (2015) 503–508
beta-blocker in this study was relatively high, and these drugs are known to cause unpleasant side effects particularly in the first few days of treatment. Thus the high degree of failure to adhere after a short-term prescription might be related to the active choice by the patients who discontinued their therapy due to the side effects. In addition, this observational study is not a randomized trial and there may exist differences in baseline characteristics of the study participants or residual confounding, despite our efforts to control confounder variables using regression models and stratified analyses. Some variables like patients' educational level, blood pressure control and medical history were not available, and they might influence medication adherence. Lastly, the study findings might not be generalizable to Chinese patients living in other parts of China, due to differences in health literacy and other clinical characteristics. 4.4. Conclusions: implications to clinical practice and future research In summary, this study shows that the duration of initial antihypertensive prescriptions for newly diagnosed hypertensive patients should not be too short to minimize subsequent medication non-adherence. This finding bears a substantial implication for the prescribing decisions among clinicians and policy-making for pharmacy refill systems [31]. Longer duration of antihypertensive prescriptions for medicationnaive patients may represent a new strategy to modify long-term adherence [33]. This conclusion should, however, be cautiously weighed against some potential harm caused by very long prescription periods — which could lead to medication wastage and minimize co-payment collection [32]. Too long a prescription is also a concern as some patients might warrant an earlier follow-up for further assessment and review of medication side-effects. The major message of this study is, therefore, to call for this balance when physicians consider to prescribe antihypertensive agents for newly diagnosed hypertensive patients. Policy and system changes required to make dispensing larger amounts costeffective are needed [32]. Future studies should evaluate the difference in longer-term adherence and other important clinical outcomes between short and long initial prescriptions, preferably in trial settings. Grant support None. Conflict of interest The authors report no relationships that could be construed as a conflict of interest. Acknowledgments We express our gratitude for all the healthcare professionals who entered the data into the clinical database. We thank the Hospital Authority of the Hong Kong Government for allowing our research team to use the database. References [1] S. Yusuf, S. Reddy, S. Ounpuu, S. Anand, Global burden of cardiovascular diseases: part I: general considerations, the epidemiologic transition, risk factors, and impact of urbanization, Circulation 104 (2001) 2746–2753. [2] C.M.M. Lawes, S. Vander Hoorn, A. Rodgers, I.S. Hypertens, Global burden of bloodpressure-related disease, 2001, Lancet 371 (2008) 1513–1518. [3] H.H.X. Wang, J.J. Wang, S.Y.S. Wong, M.C.S. Wong, F.J. Li, P.X. Wang, et al., Epidemiology of multimorbidity in China and implications for the healthcare system: crosssectional survey among 162,464 community household residents in southern China, BMC Medicine 12 (2014) 188. [4] M.C.S. Wong, H.H.X. Wang, M.C.M. Leung, C.S.H. Tsang, S.V. Lo, S.M. Griffiths, The rising prevalence of self-reported hypertension among Chinese subjects: a population-based study from 121 895 household interviews, QJM-An Int J Med 108 (2014) 9–17.
507
[5] D. Lloyd-Jones, R. Adams, M. Carnethon, G. De Simone, T.B. Ferguson, K. Flegal, et al., Heart disease and stroke statistics—2009 update: a report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee, Circulation 119 (2009) 480–486. [6] K. Wolf-Maier, R.S. Cooper, H. Kramer, J.R. Banegas, S. Giampaoli, M.R. Joffres, et al., Hypertension treatment and control in five European Countries, Canada, and the United States, Hypertension 43 (2004) 10–17. [7] T. Dorjgochoo, X.O. Shu, X. Zhang, H. Li, G. Yang, L. Gao, et al., Relation of blood pressure components and categories and all-cause, stroke and coronary heart disease mortality in urban Chinese women: a population-based prospective study, J. Hypertens. 27 (2009) 468–475. [8] Blood Pressure Lowering Treatment Trialists CF. Turnbull, B. Neal, T. Ninomiya, C. Algert, H. Arima, et al., Effects of different regimens to lower blood pressure on major cardiovascular events in older and younger adults: meta-analysis of randomised trials, BMJ 336 (2008) 1121–1123. [9] P.M. Ho, C.L. Bryson, J.S. Rumsfeld, Medication adherence its importance in cardiovascular outcomes, Circulation 119 (2009) 3028–3035. [10] G. Mazzaglia, E. Ambrosioni, M. Alacqua, A. Filippi, E. Sessa, V. Immordino, et al., Adherence to antihypertensive medications and cardiovascular morbidity among newly diagnosed hypertensive patients, Circulation 120 (2009) 1598–1605. [11] E. Sabate, World Health Organization, WHO Adherence to Long Term Therapies Project., Global Adherence Interdisciplinary Network. Adherence to Long-Term Therapies: Evidence for Action, World Health Organization, Geneva, 2003. [12] B. Vrijens, G. Vincze, P. Kristanto, J. Urquhart, M. Burnier, Adherence to prescribed antihypertensive drug treatments: longitudinal study of electronically compiled dosing histories, BMJ 336 (2008) 1114–1117. [13] B.L. van Wijk, W.H. Shrank, O.H. Klungel, S. Schneeweiss, M.A. Brookhart, J. Avorn, A cross-national study of the persistence of antihypertensive medication use in the elderly, J. Hypertens. 26 (2008) 145–153. [14] J.J. Caro, M. Salas, J.L. Speckman, G. Raggio, J.D. Jackson, Persistence with treatment for hypertension in actual practice, CMAJ 160 (1999) 31–37. [15] C. Bourgault, M. Senecal, M. Brisson, M.A. Marentette, J.P. Gregoire, Persistence and discontinuation patterns of antihypertensive therapy among newly treated patients: a population-based study, J. Hum. Hypertens. 19 (2005) 607–613. [16] J. Mar, F. Rodriguez-Artalejo, Which is more important for the efficiency of hypertension treatment: hypertension stage, type of drug or therapeutic compliance? J. Hypertens. 19 (2001) 149–155. [17] D.A. Hughes, A. Bagust, A. Haycox, T. Walley, The impact of non-compliance on the cost-effectiveness of pharmaceuticals: a review of the literature, Health Econ. 10 (2001) 601–615. [18] G. Mancia, R. Fagard, K. Narkiewicz, J. Redon, A. Zanchetti, M. Bohm, et al., 2013 ESH/ ESC guidelines for the management of arterial hypertension: the Task Force for the Management of Arterial Hypertension of the European Society of Hypertension (ESH) and of the European Society of Cardiology (ESC), Eur. Heart J. 34 (2013) 2159–2219. [19] M.C. Sokol, K.A. McGuigan, R.R. Verbrugge, R.S. Epstein, Impact of medication adherence on hospitalization risk and healthcare cost, Med. Care 43 (2005) 521–530. [20] S. Baroletti, H. Dell'Orfano, Medication adherence in cardiovascular disease, Circulation 121 (2010) 1455–1458. [21] M.C.S. Wong, J.Y. Jiang, J.L. Tang, A. Lam, H. Fung, S.W. Mercer, Health services research in the public healthcare system in Hong Kong: an analysis of over 1 million antihypertensive prescriptions between 2004–2007 as an example of the potential and pitfalls of using routinely collected electronic patient data, BMC Health Serv. Res. 8 (2008) 138. [22] M.C.S. Wong, J.Y. Jiang, S.M. Griffiths, Factors associated with compliance, discontinuation and switching of calcium channel blockers in 20,156 Chinese patients, Am. J. Hypertens. 22 (2009) 904–910. [23] M.C.S. Wong, J.Y. Jiang, S.M. Griffiths, Antihypertensive drug adherence among 6408 Chinese patients on angiotensin-converting enzyme inhibitors in Hong Kong: a cohort study, J. Clin. Pharmacol. 50 (2010) 598–605. [24] M.C.S. Wong, J.Y. Jiang, S.M. Griffiths, Adherence to combination therapy among ethnic Chinese patients: a cohort study, Hypertens. Res. 33 (2010) 416–421. [25] M.C.S. Wong, J.Y. Jiang, S.M. Griffiths, Factors associated with compliance to thiazide diuretics among 8551 Chinese patients, J. Clin. Pharm. Ther. 36 (2011) 179–186. [26] M.C.S. Wong, W.W.S. Tam, C.S.K. Cheung, E.L.H. Tong, A.C.H. Sek, N.T. Cheung, et al., Medication adherence to first-line antihypertensive drug class in a large Chinese population, Int. J. Cardiol. 167 (2013) 1438–1442. [27] M.C.S. Wong, W.W.S. Tam, C.S.K. Cheung, E.L.H. Tong, A.C.H. Sek, N.T. Cheung, et al., Antihypertensive prescriptions over a 10-year period in a large Chinese population, Am. J. Hypertens. 26 (2013) 931–938. [28] M.C.S. Wong, W.W.S. Tam, C.S.K. Cheung, H.H.X. Wang, E.L.H. Tong, A.C.H. Sek, et al., Drug adherence and the incidence of coronary heart disease- and stroke-specific mortality among 218,047 patients newly prescribed an antihypertensive medication: a five-year cohort study, Int. J. Cardiol. 168 (2013) 928–933. [29] M.C.S. Wong, W.W.S. Tam, H.H.X. Wang, C.S.K. Cheung, E.L.H. Tong, A.C.H. Sek, et al., Predictors of the incidence of all-cause mortality and deaths due to diabetes and renal diseases among patients newly prescribed antihypertensive agents: a cohort study, Int. J. Cardiol. 168 (2013) 4705–4710. [30] M.C.S. Wong, J.Y. Jiang, S.M. Griffiths, Short-term adherence to beta-blocker therapy among ethnic Chinese patients with hypertension: a cohort study, Clin. Ther. 31 (2009) 2170–2177. [31] M.C.S. Wong, C.H.M. Wu, H.H.X. Wang, H.W. Li, E.M.T. Hui, A.T. Lam, et al., Association between the 8-item Morisky Medication Adherence Scale (MMAS-8) score and glycaemic control among Chinese diabetes patients, J. Clin. Pharmacol. 55 (2015) 279–287.
508
M.C.S. Wong et al. / International Journal of Cardiology 182 (2015) 503–508
[32] H.A. Batal, M.J. Krantz, R.A. Dale, P.S. Mehler, J.F. Steiner, Impact of prescription size on statin adherence and cholesterol levels, BMC Health Serv. Res. 7 (2007) 175. [33] N.M. Ivers, J.D. Schwalm, C.A. Jackevicius, H. Guo, J.V. Tu, M. Natarajan, Length of initial prescription at hospital discharge and long-term medication adherence for elderly patients with coronary artery disease: a population-level study, Can. J. Cardiol. 29 (2013) 1408–1414. [34] N.K. Choudhry, W.H. Shrank, R.L. Levin, J.L. Lee, S.A. Jan, M.A. Brookhart, et al., Measuring concurrent adherence to multiple related medications, Am. J. Manage. Care 15 (2009) 457–464. [35] B.C. Martin, E.K. Wiley-Exley, S. Richards, M.E. Domino, T.S. Carey, B.L. Sleath, Contrasting measures of adherence with simple drug use, medication switching, and therapeutic duplication, Ann. Pharmacother. 43 (2009) 36–44.
[36] S.E. Andrade, K.H. Kahler, F. Frech, K.A. Chan, Methods for evaluation of medication adherence and persistence using automated databases, Pharmacoepidemiol. Drug Saf. 15 (2006) 565–577. [37] G.K.Y. Lee, H.H.X. Wang, K.Q.L. Liu, Y. Cheung, D.E. Morisky, M.C.S. Wong, Determinants of medication adherence to antihypertensive medications among a Chinese population using Morisky Medication Adherence Scale, PLoS ONE 8 (2013) e62775. [38] L. Osterberg, T. Blaschke, Adherence to medication, N. Engl. J. Med. 353 (2005) 487–497. [39] S. Kripalani, F. LeFevre, C.O. Phillips, M.V. Williams, P. Basaviah, D.W. Baker, Deficits in communication and information transfer between hospital-based and primary care physicians: implications for patient safety and continuity of care, JAMA 297 (2007) 831–841.
