Computer-based medical system for the computation of blood pressure excess in the diagnosis of hypertension.

A computer-based system for assessing hypertension was designed by combining hardware for automatic, long-term blood pressure (BP) measurement with a set of software modules for computing time-specified tolerance intervals and evaluating measures of BP excess. BP is so variable that the identification and proper definition of hypertension are highly ambiguous when based on single measurements. One first step in dealing with such variability is to replace the constant limits for BP frequently used in the assessment of hypertension by a time-specified reference interval. Once such a threshold is available, a hyperbaric index can be computed by numerical integration as the total area (within one cycle) of any given patient's BP above the threshold. In order to examine the extent of normal physiologic excess, a series of 266 systolic and diastolic BP and heart rate (HR) measurements were automatically monitored every 30 minutes for at least 48 hours from clinically healthy subjects, aged 19 to 25 years. Original data were used to obtain 90% tolerance intervals for each gender separately. The hyperbaric index was then computed for each individual BP profile. The distribution of the maximum hyperbaric index (maximum of the values computed for systolic, mean arterial, and diastolic BP levels shows a highest value of 14.839 mmHg x hr for the men and of 10.229 mmHg x hr for the women. These values represent a testable threshold for assessing hypertension based on the proposed approach. The tolerance intervals obtained from the reference population were also used to compute hyperbaric indices for a series of 175 BP measurements sampled from clinically healthy people, as well as a series of 60 measurements sampled from patients with mild hypertension. Sensitivity and specificity in the diagnosis of hypertension based on the hyperbaric index were both 100%, as opposed to values obtained from computing the BP load, the average of the BP series, or the circadian amplitude, all of which provided a much poorer diagnostic test. The software system developed for automatically establishing time-qualified tolerance limits from a reference population and assessing the extent and timing of BP elevation for a test subject may help to establish a prognosis and diagnosis, with a correspondingly better assessment of health status, to initiate treatment if needed, to time treatment when it is most desirable and least harmful in terms of undesired effects, and to gauge the patient's response to treatment.