Assessment of adherence to antihypertensive drugs in patients with resistant hypertension receiving optimal treatment

Abstract

To estimate the proportion of non-adherence to antihypertensive drugs in patients with apparent resistant hypertension despite optimal medical treatment. The study was prospective and observational. We screened all consecutive patients managed in our tertiary healthcare center for resistant hypertension between January 2014 and December 2017. Were included only those who already had an ambulatory blood pressure measurement (ABP) to exclude a white coat effect and a thorough etiological work up to exclude a secondary cause. Hypertension was considered resistant if the daytime ABP was over or equal to 135/85 mmHg and/or the 24-hours ABP was over or equal to 130/80 mmHg, under 4 antihypertensive medications combining a renin-angiotensin system inhibitor, amlodipine, a thiazide (or indapamide) and spironolactone, at optimal doses. Adherence to treatment was assessed by the eight-item Morisky Medication Adherence Scale (MMAS-8). Three hundred eighty-six patients were enrolled, with a mean age of 57.4 ± 11.3 years, 48.3% of men. The mean office blood pressure was 178 ± 20.4/101 ± 15.5 mmHg and the 24-hours ABP was 164 ± 17.6/97 ± 15.2 mmHg. The proportions of high, medium and low adherence were 27.9%, 47.6% and 24.5% respectively. Fully adherent, partially non-adherent and completely non-adherent patients differed significantly in terms of proportions of women (25%, 48% and 72%), number of daily drugs, reflecting comorbidities (5.9; 6.1, and 9.8 respectively) and education level (the proportions of patients who did not achieve a secondary school were 10.1%, 28.3% et 53.2% respectively). In our study, more than two out of three patients with apparent resistant hypertension optimally treated and without white coat effect, were partially or completely non-adherent to treatment. Assessment of treatment adherence in these high cardiovascular risk patients should be systematically implemented and if possible by more objective methods.