The development of automatic instrumentation for ambulatory blood pressure monitoring makes it possible to follow the time-course of blood pressure variation over 24 h or more in large groups of individuals. Whenever samples from a reference group of individuals are available, one may construct a prediction interval that is expected to include any single future observation from the reference population, with a specified confidence. Alternatively, the reference interval may consist of a tolerance interval that will include at least a specified proportion of the population with a stated confidence.
To examine prospectively whether a new, combined tolerance–hyperbaric test approach of establishing tolerance intervals for the circadian variability of blood pressure as a function of gestational age and then computing the hyperbaric index as a measure of blood pressure excess provides high sensitivity and specificity in the early identification of pregnant women who subsequently will develop gestational hypertension or preeclampsia.
We used data sampled for 48 h from 148 normotensive men and women to compute and compare time-specified tolerance and prediction intervals for blood pressure. Once the threshold, given by the upper limit of the tolerance interval, was available, the hyperbaric index, as a measure of blood pressure excess, could be calculated by numerical integration as the total area of any given patient's blood pressure above the threshold. The hyperbaric index, in addition to the duration of excess, could then be used as nonparametric endpoints for assessing hypertension. This combined approach, the tolerance–hyperbaric test, was examined prospectively in the early identification of pregnant women who subsequently will develop gestational hypertension or preeclampsia. We analysed 1494 blood pressure series sampled for 48 h in 124 women with uncomplicated pregnancies and 78 women who developed gestational hypertension or preeclampsia.
Sensitivity of the tolerance–hyperbaric test was 93% for women sampled during the first trimester of gestation, and increased up to 99% in the third trimester. Results further indicated lower sensitivity and specificity from other parameters also computed from data sampled by ambulatory monitoring, including the blood pressure load and average values of blood pressure.
The results suggested the use of tolerance intervals as the appropriate threshold for the circadian variability of blood pressure, primarily because the practical use of prediction limits was restricted by the need for assumptions of normality and symmetry; also, the prediction intervals could not be considered generally applicable when constructed on the basis of a small sample.
Despite the limitations of ambulatory blood pressure monitoring, the tolerance-hyperbaric test represents a reproducible, noninvasive, and highly sensitive test for diagnosing hypertension that has also been validated prospectively for the very early identification of subsequent gestational hypertension and preeclampsia.