Accuracy and reliability of wrist-cuff devices for self-measurement of blood pressure

Object Self-measurement of blood pressure (1313) might offer some advantages in diagnosis and therapeutic evaluation and in patient management of hypertension. Recently, wrist-cuff devices for self-measurement of BID have gained more than one-third of the world market share. In the present study, we validated wrist-cuff devices and compared the results between wrist- and arm-cuff devices. The factors affecting the accuracy of wrist-cuff devices were also studied. Method The research group to assess the validity of automated blood pressure measuring device consisted of 13 institutes in Japan, which validated two wrist-cuff devices (WC-1 and WC-2) and two arm-cuff devices (AC-1 and AC-2). They used a crossover method, where the comparison was done between auscultation, by two observers by means of a double stethoscope on one arm and the device on the opposite arm or wrist. Results There was good inter-observer agreement for the auscultation method in each institute (systolic blood pressure (SBP), -0.1 +/- 2.8 mmHg; diastolic blood pressure (DBP), -0.1 +/- 2.6 mmHg, n = 498). The mean difference between auscultation and the device was minimal both in arm-cuff devices (mean difference for AC-1, 2.2/1.9 mmHg, n = 97 and for AC-2, 5.1/2.9 mmHg, n = 136, SBP/DBP) and wrist-cuff devices (mean difference for WC-1, -2.1/1.2 mmHg, n = 173 mmHg and for WC-2, -2.3/-5.6 mmHg, n = 92). The standard deviation of the difference (SDD) in wrist-cuff devices, however (SDD for WC-1, 9.7/7.3 mmHg and for WC-2, 10.2/ 8.6 mmHg), was larger than that of the arm-cuff devices (SDD for AC-1, 5.6/6.6 mmHg and for AC-2, 6.3/ 5.1 mmHg). Grading of AC-1 and AC-2 based on criteria of British Hypertension Society was A/A and B/A, respectively, while that of WC-1 and WC-2 was C/B and D/ B, respectively. Using the same validation protocol, the results of validation for one device were divergent in each institute. In wrist-cuff devices, the BP value obtained in palmar flexion was significantly higher and that obtained in palmar dorsiflexion was significantly lower than that in palmar extension. In some cases, finger plethysmogram did not disappear during maximum inflation of the wrist-cuff (congruent to 250 mmHg), even in palmar extension and especially in palmar flexion, suggesting incomplete obstruction of radial and/or ulnar arteries during inflation. Conclusion The results suggest that wrist-cuff devices in the present form are inadequate for self-measurement of blood pressure and, thus, are inadequate for general use or clinical and practical use. However, there is much possibility in wrist-cuff device and the accuracy and reliability of wrist-cuff device are warranted by an improvement of technology. J Hypertens 20:629-638 (C) 2002 Lippincott Williams Wilkins.