Hypertension control and medication increase in primary care.

Over half of treated patients with hypertension are not well controlled. However, little is known about physicians' prescribing behaviour for these patients. Our objective was to clarify whether physicians increase antihypertensive medication in patients with poorly controlled hypertension and what characteristics are predictors of medication increase. This was a retrospective cohort study by surveying medical records in primary care clinics in Tochigi, Japan. Twenty-nine of 79 randomly selected physicians agreed to select 20 consecutive hypertensive patients. This resulted in 547 patients (women 60%; mean (s.d.) age, 68 (12) years) who had blood pressure measurements taken in 1998 and prescription of antihypertensive medication in 1998 and 1999. Mean (s.d.) systolic/diastolic blood pressure was 142 (12)/81 (9) mm Hg and the percentage of patients in good control (<140/90 mm Hg), fair (140-159/90-94) and poor (> or =160/95) were 42%, 47%, and 11%, respectively. Physicians increased medication in 28% of poorly controlled patients (95% confidence interval (CI), 17-41%), which was more than those in fair (12%, 95%CI 8-16%) or good control (7%, 95%CI 4-12%). Multivariate logistic regression analysis showed that systolic and diastolic blood pressures were positively, and the number of kinds of antihypertensive medications and the age of the physician were negatively, associated with an increase in medication. In conclusion, primary care physicians did not increase antihypertensive medication adequately for patients with uncontrolled hypertension. Attempts to understand and to change physicians' prescription behaviour could reduce the burden of uncontrolled hypertension among treated hypertensive patients.