Rapid estimation of the accuracy of automatic blood pressure measuring devices (READ)

We propose a rapid method for the evaluation of automatic blood pressure me asurement devices (READ) in response to the claim to simplify the validatio n of those devices. The READ is based on numerous blood pressure (BP) measu rements at rest and during a standardised postural challenge in a small num ber of subjects who exhibit a wide range of BPs. Automatic (AU) and mercury sphygmomanometric (MS) brachial BP were measured simultaneously in a blind ed manner. An average of 30 measurements per patient were done in 10-min wh ile in a supine position, followed by 30 min in head-up tilt and again supi ne for 10 min. Two Collin-8800 automated oscillometric devices were tested with the aid of the READ in 15 adults. The consistency of the MS standard w as demonstrated by duplicate MS measurements showing minimal inter-observer differences consistent with class A of the British Hypertension Society (B HS) grading system. AD-to-MS differences of 447 measurements were mean syst olic Delta BP = 1.1 +/- 11.4 mm Hg and mean diastolic Delta BP -13.36 +/- 8 .9 mm Hg, classifying into category D of the BHS. Both tested instruments, each subject, each phase of the tilt test (supine I, tilt, supine II) and e ach mode of measurement (at 5-min intervals or continuously) equally qualif ied as 'D'. We conclude that the READ permitted us to identify quickly and at low expense a grossly inaccurate automatic BP measuring device. Further studies will show whether the READ can be efficient as a pre-validation tes t, post-validation test, and for assessment of ambulatory BP measuring devi ces.