Validity of four indirect methods to measure adherence in primary care hypertensives.

High blood pressure (HBP) is one of the most important risk factors for morbidity and mortality in the world. Non-adherence to medication is associated with the lack of control of HBP. The objective of this study was to assess the validity of four indirect methods for measuring adherence to medication to control HBP in primary health care. A random sample of 120 hypertensive patients who were undergoing treatment for at least 2 months in a primary care unit in Florianópolis (Brazil) were included in the study. The independent variables were four indirect methods to measure adherence to medication: knowledge regarding the HBP medication, the blood pressure level, attitude regarding the medication intake (Morisky-Green test) and self-reported adherence. The classification of HBP was based on criteria established by the Brazilian Ministry of Health. The gold standard used for measuring adherence was the pill count. Logistic regression was used to estimate sensitivity (highest value of 88.2% for self-report), specificity (highest value of 70.7% for HBP control), positive predictive value (highest value of 46.4% for HBP control) and negative predictive value (highest value of 79.1% for Morisky-Green test) for each of the indirect methods. No indirect method of measuring adherence had a good positive predictive value for adherence, which was best predicted by patients' age and whether they managed to control HBP. The results also revealed low treatment adherence (31.2%) and low control of HBP (37.6%). Non-adherence was mainly associated with side effects of the treatment.