Blood pressure variability and prevalence of hypertension using automated readings from multiple visits to a pharmacy-based community-wide programme.

Blood pressure (BP) measurements taken outside the routine office context may be a useful adjunct strategy to monitor BP. Community-based BP data can also provide estimates of the prevalence of elevated BP. We compared multiple readings taken on different days in pharmacies using an automated BpTRU device during a cardiovascular health programme targeting community-dwelling older adults. Mean systolic (S) and diastolic (D) BP values were compared over time using repeated measures analysis of variance for all participants with at least three separate sets of readings (n=317). BP variability was then examined among four subgroups based on report of antihypertensive medication or no treatment, and normal or elevated SBP at the initial visit (< or >or=140, or 130 if diabetes reported). Prevalence of elevated BP was compared across visits. Overall, mean SBP decreased between visits 1 and 2 (140.4 vs 137.1 mm Hg; P<0.001). Among participants with normal SBP at the initial visit, SBP did not vary significantly, whether or not antihypertensive treatment was reported. Those with initially elevated SBP experienced a significant decrease between visits 1 and 2, also regardless of treatment status. Prevalence of elevated BP decreased from visits 1 to 2 (55.8 vs 48.9%; P=0.026) and from visits 1 to 3 (55.8 vs 42.9%; P<0.001). Analyses of BP data from a community-based programme using an accurate device showed that initial readings may inflate the population estimate of elevated BP. Findings suggest that more than one set of BP readings measured on different occasions are needed, particularly if the first set is elevated.