Wrist measurement of blood pressure: some critical remarks to oscillometry

Oscillometric blood pressure measuring devices worn at the wrist are preferred by a growing part of hypertensive patients for self-measurement because of their compactness and ease of use. However, little is known about the accuracy of such instruments. The accuracy of such devices was evaluated by measuring the blood pressure at the end of a coronary angiography procedure in 27 subjects nearly simultaneously in the aortic arch, using a Statham P23 transducer, and at the wrist, using the blood pressure watch (BPW) by NAIS-Matsushita. Four replicate comparative measurements were performed in each subject. On average, the blood pressure measured by the BPW was found to be higher: systolic +1.2+/-(SD)10.2 mm Hg, diastolic +4.1+/-7.2 mm Hg. The correlation coefficient between the two methods was 0.85 for the systolic and 0.84 for the diastolic pressure, but 39% of the systolic and 22% of the diastolic differences were outside the range of +/-1 SD. There was no correlation between systolic differences and blood pressure or patients' variables, but multiple regression analysis revealed that the diastolic blood pressure differences were negatively correlated with the aortic diastolic pressure (r = 0.89, p<0.001), indicating that the BPW is less accurate to detect lower values.

(1) Though the average systolic and diastolic deviations between both methods and the relative correlation coefficients suggest that the BPW is to be used carefully for monitoring blood pressure, the great variability of differences and the inability of the BPW to detect lower diastolic blood pressure calls for a mandatory validation test by an approved protocol before this new device can be recommended for general use. (2) On the basis of our results for wrist oscillometry with the BPW and data reported in the literature for upper-arm oscillometric devices, we recommend to use oscillometric devices only when they were validated properly. In particular, the oscillometric measurement should be employed with caution for epidemiological studies or for investigation of population groups (e.g., children or pregnant women) that are expected to display lower diastolic pressures.