Persistence, adherence, and risk of discontinuation associated with commonly prescribed antihypertensive drug monotherapies.

OBJECTIVE To assess 1-year persistence and adherence with monotherapy using the most commonly dispensed individual agent in 4 antihypertensive drug classes: hydrochlorothiazide (HCTZ), amlodipine, lisinopril, or valsartan. DESIGN Retrospective, longitudinal analysis of initial prescriptions during 2001 to 2002 from a nationwide administrative claims database representing 11 million covered lives in the United States. MEASUREMENTS Drug utilization following initiation. Cox proportional hazards regression models controlled for demographics, case-mix, and concomitant treatments. RESULTS Records for 60,685 subjects were included: HCTZ (n = 18,713), amlodipine (n = 11,520), lisinopril (n = 21,138), or valsartan (n = 9314). Over 1 year, 31% to 44% of subjects utilized no treatment for at least 60 days. Medication possession ratio (MPR) and adherence measures ranged from 73% to 90%. Valsartan was associated with significantly (P < .001) more favorable measures of persistence, length of therapy, time to discontinuation, MPR, and risk of discontinuation, compared with HCTZ, amlodipine, or lisinopril. The risk of discontinuation was 53%, 32%, and 14% greater for HCTZ, amlodipine, and lisinopril, respectively, versus valsartan (all comparisons P < .001). CONCLUSION Among antihypertensive agents studied, valsartan was associated with the most favorable utilization patterns. Health care providers and systems should evaluate the use of antihypertensive drugs within their populations to identify and manage treatment discontinuation.