RESUMO
BACKGROUND: Patients' adherence to antihypertensive drug therapy-especially at the beginning of treatment-is essential for preventing serious cardiovascular complications over the long term. OBJECTIVES: This study was conducted to assess adherence among hypertensive patients who initiated antihypertensive pharmacotherapy and to identify whether it was related to the medical provider, dispensing patterns, or comorbidities. METHODS: We reviewed the computerized claim records submitted to Korea's Health Insurance Review Agency (which maintains data for all medication prescriptions for Korean residents) between July 2004 and December 2006. We processed the claims of adult hypertensive patients who initiated therapy with an antihypertensive medication in 2005. Medication adherence was assessed by the cumulative medication adherence (CMA), calculated by dividing the sum of a day's supply (obtained over a series of intervals) by the total number of days in the time period. Good adherence was defined as CMA > or =80%. RESULTS: The records of 725,220 antihypertensive patients aged > or =20 years were included in the analysis. The mean CMA value of the study group was 59.6% (median, 67.6%), and 39.2% of the patients had good adherence (CMA > or =80%). In multiple logistic regression analysis, the likelihood of a good adherence rate was greater when the medical provider was from a public health center (adjusted odds ratio [AOR], 2.71) or private clinic (AOR, 1.99) than a general hospital (ie, a hospital with >100 beds and > or =9 major departments) (AOR, 1.00). The likelihood of good adherence was greater when the medical provider's specialty was internal medicine (AOR, 1.00) versus family medicine (AOR, 0.96) or another specialty (AOR, 0.85). The odds of good adherence were greater among patients prescribed an angiotensin-converting enzyme inhibitor or angiotensin II receptor blocker (AOR, 1.00) or combined drugs without a diuretic (AOR, 1.01) as the first-line drug rather than other drugs (AOR, < or =0.92). The likelihood of good adherence was also better when the mean daily number of antihypertensive pills was >1 to 2 (AOR, 1.22) or >2 to 3 (AOR, 1.34) than when it was < or =1 (AOR, 1.00). The likelihood of good adherence was lower among patients without target organ disease or metabolic syndrome (AOR, 1.00) and highest among those with > or =4 relevant comorbidities (AOR, 1.85). CONCLUSIONS: The overall CMA of these hypertensive Korean patients who started antihypertensive therapy for the first time was <60%, and the rate of good adherence (CMA > or =80%) was <40%. Good medication adherence appeared to be related to the type of medical provider who prescribed the medication, the type of antihypertensive agent prescribed, the number of agents used, and the number of related comorbidities that a patient had.