Combining repeated blood pressure measurements to obtain prevalences of high blood pressure

Blood pressure (BP) levels may be classified as normal, borderline, or high with respect to the World Health Organization (WHO) criteria, but most studies like the MONICA project require at least two BP measurements and must tackle the problem of combining the results of the different readings into a single value for classification. The Swiss MONICA project measured the blood pressure of 1872 individuals in the areas of Vaud and Fribourg. Second BP readings were, on average, lower than the first by 3.2 mmHg for systolic and 1.1 mmHg for diastolic BP. These differences, while trivial at the individual level, nevertheless generate significant effects on prevalences of high BP. The first reading, the second, the mean, and the lowest yield prevalences of 14%, 10%, 11% and 9% respectively. Therefore, any published prevalence of high blood pressure should specify the method of measurement, the number of readings taken, and the way results were combined.