Assessing the prevalence of hypertension in populations: are we doing it right?

Background Although it is well recognized that the diagnosis of hypertension should be based on blood pressure (BP) measurements taken on several occasions, notably to account for a transient elevation of BP on the first readings, the prevalence of hypertension in populations has often relied on measurements at a single visit. Objective To identify an efficient strategy for assessing reliably the prevalence of hypertension in the population with regards to the number of BP readings required. Design Population-based survey of BP and follow-up information. Setting and participants All residents aged 25-64 years in an area of Dar es Salaam (Tanzania). Main outcome measures Three BP readings at four successive visits in all participants with high BP (n = 653) and in 662 participants without high BP, measured with an automated BP device. Results BP decreased substantially from the first to third readings at each of the four visits. BP decreased substantially between the first two visits but only a little between the next visits. Consequently, the prevalence of high BP based on the third reading - or the average of the second and third readings - at the second visit was not largely different compared to estimates based on readings at the fourth visit. BP decreased similarly when the first three visits were separated by 3-day or 14-day intervals. Conclusions Taking triplicate readings on two visits, possibly separated by just a few days, could be a minimal strategy for assessing adequately the mean BP and the prevalence of hypertension at the population level. A sound strategy is important for assessing reliably the burden of hypertension in populations. (C) 2003 Lippincott Williams Wilkins.