Objectives: To pool data from individual subjects in an attempt to determine an operational threshold for making clinical decisions based on the self-recorded blood pressure (SRBP) and to examine how many hypertensive subjects, diagnosed by conventional blood pressure (CBP) measurement, would have a self-recorded blood pressure within the normotensive range if the proposed thresholds were applied.
Data sources: Thirteen research groups studied 4668 untreated subjects.
Results: In total 2401 subjects were normotensive, 494 were borderline hypertensive and 1773 were definitely hypertensive. Hypertension had been diagnosed from the mean of 1-6 (median 3) CBP measurements obtained during 1-3 (median 1) visits. The reference values for SRBP measurements determined from the 95th percentiles of the distributions for normotensive subjects were 137 mmHg systolic and 85 mmHg diastolic. Of the subjects with systolic hypertension, 16% had systolic SRBP </= 137 mmHg. Similarly, 25% of those with diastolic hypertension had diastolic SRBP </= 85 mmHg. The probability that hypertensive subjects had SRBP below these thresholds was 34% (diastolic) to 62% (systolic) greater for women than it was for men, was 2-3 times greater if fewer than three CBP measurements had been averaged for establishing the diagnosis of hypertension, and increased by 50% (diastolic) to 126% (systolic) when SRBP had been measured on more than 3 days. In contrast, for each 10 mmHg increment in systolic CBP, this probability decreased by 35% for systolic SRBP and by 11% for diastolic SRBP; for each 5 mmHg increment in diastolic CBP it decreased by 36% for diastolic SRBP. In addition, for systolic blood pressure, the probability decreased by 31% for each 10-year increment in age.
Conclusions: Until the relationship between SRBP and the incidence of cardiovascular morbidity and mortality has been clarified further by prospective studies, a mean self-recorded blood pressure above 137 mmHg systolic or 85 mmHg diastolic may be considered hypertensive. Among the hypertensive patients 16-31% had SRBP below these thresholds, but this proportion decreased if the hypertensive subjects had higher CBP.