Quality assessment of blood pressure measurements in epidemiological surveys. The impact of last digit preference and the proportions of identical duplicate measurements. WHO Monica Project [corrected]

Abstract

In the WHO MONICA Project, cardiovascular risk factor surveys including measurements of arterial blood pressure (BP) were conducted in more than 50 different populations. In the course of a retrospective BP measurement quality assessment effort, two indicators of "prejudiced" blood pressure reading, last digit preference and high proportions of identical results in duplicate measurements, are used in addition to other items to evaluate blood pressure measurement quality. We used fictitious blood pressure distributions and applied to them Last Digit Preference scores and Proportions of Identical Duplicate Measurements actually found in the MONICA surveys. The analysis showed that Last Digit Preference affects predominantly the shape of the BP distribution curve, whereas high Proportions of Identical Duplicate Measurements may cause a shift of the entire BP distribution curve. Although the two items are partly interrelated, a clear distinction between them and their effects is advocated.