[Self-reported hypertension: validation in a representative cross-sectional survey].

To assess the validity of self-reported hypertension, a population-based cross-sectional study was carried out in Pelotas, Southern Brazil. Individuals aged 20 years were selected through probabilistic two-stage sampling. All 2,949 participants had their blood pressure measured twice (5 minutes apart) during a household visit. Those presenting systolic pressure 140 mmHg and/or diastolic pressure 90 mmHg were revisited. The gold standard for hypertension was the mean of the two measures taken at the second visit and/or use of anti-hypertensive drugs. Self-reported hypertension was obtained through the question: 'Has any doctor ever said that you have high blood pressure?'. Prevalence rates for self-reported and measured hypertension were 33.6% and 29.5%, respectively. Sensitivity was 84.3% (81.7-86.7%), specificity 87.5% (86.0-88.9%), positive predictive value 73.9% (71.0-76.6%), and negative predictive value 93.0% (91.8-94.1%). Self-reporting is a reliable method for monitoring prevalence of hypertension, one of the most important risk factors for chronic non-communicable diseases.