Hipertensão arterial sistêmica no Brasil : estimação da prevalência real obtida por medida autorreferida

Abstract

The prevalence of systemic arterial hypertension (SAH) is often measured in health surveys, which commonly use the self-reported measure as a measure. However, this measurement process can generate a biased measure. The objective was to correct the prevalence of SAH in Brazil, using the sensitivity and specificity of the self-reported question. The self-referenced prevalence correction methods were used, as well as the presentation and comparison of new techniques (empirical method and integral transformation method), which allow the correction of large samples. In addition, a systematic review, including meta-analysis of Brazilian studies that validated selfreported hypertension, was performed for use in the correction. Due to the high heterogeneity of the studies, the meta-analysis used the random effects model and biases were evaluated by the QUADAS-2 scale. The prevalence of corrected SAH for Brazil was 14.5%, while self-reported SAH was 22.1%. In the meta-analysis, sensitivity and specificity combined were 77% and 88%, respectively. The empirical and integral transformation methods presented close estimates, however, due to poor conditioning in the correction formula, reflected in large differences in corrected prevalence. The integral transformation method should be preferred in any sample size. When selecting Brazilian studies, it was possible to identify combined sensitivity and specificity values higher than those found with studies of greater heterogeneity. In summary, the study presented corrected estimates of hypertension, closer to true, without the need to directly measure all individuals.