Failure of the community-based Vita-Stat automated blood pressure device to accurately measure blood pressure

Abstract

Objective: To evaluate the Vita-Stat automated blood pressure computer (a patient-operated blood pressure measuring device available in the community) to determine its value as an instrument to monitor blood pressure in the ambulatory patient.

Design: Comparative study using the Vita-Stat vs a gold standard, the mercury sphygmomanometer.

Setting: Three local grocery stores.

Participants: Sixty-three passersby who agreed to answer questions and to sit for several measurements of blood pressure.

Interventions: Simultaneous measurement of blood pressure with each subject wearing a Vita-Stat cuff on the left arm and a mercury sphygmomanometer cuff on the right arm. Two pressures were measured sequentially in the same manner.

Main outcome measures: The reproducibility, accuracy, sensitivity, and specificity of the Vita-Stat computer compared with the gold standard.

Results: In sequential measurements, the Vita-Stat readings of both systolic and diastolic blood pressure correlated less well with each other than did the mercury readings (intramachine differences). The Vita-Stat readings also correlated poorly with the mercury readings of systolic and diastolic blood pressure (intermachine differences). The variability in readings recorded by the Vita-Stat were striking, with differences of up to 60 mm Hg from the mercury readings. More than half (63.2%) of the subjects had Vita-Stat readings that were more than 5 mm Hg different from the mercury readings. Vita-Stat systolic readings were usually lower than mercury readings and also varied by as much as 60 mm Hg below in one patient to 58 mm Hg above the mercury reading in another. The sensitivity of the Vita-Stat in correctly diagnosing hypertension was 0.26; the negative predictive value was 0.45.

Conclusions: Our data suggest that the Vita-Stat is not only inconsistent but inaccurate in measuring blood pressure in the ambulatory patient and is, therefore, not appropriate to use as a monitoring device.