We developed a dual monitor protocol for testing the accuracy and reliability of 24-hr ambulatory blood pressure monitors (ABPMs) and determined in normotensives (n=15), hypertensives (n=14) and alcohol-dependents (n=11) that a popular auscultatory ABPM was highly variable, misclassifying up to 70% of patients. Purpose: To determine the accuracy and reliability of two widely used ocillometric ABPMs, the Oscar 2 (Suntech Medical, Morrisville, NC) and the Spacelabs 90207 (Spacelabs Healthcare, Snoqualmie, WA) under controlled lab conditions. Hypothesis: Oscillometric proprietary algorithms were developed from auscultatory reference BPs, thus ABPMs would differ from each other and from observers (O1, O2) using a Hg column and Thinklabs digital stethoscope. Methods: BPs were measured in triplicate on both arms in 17 seated subjects (10 ♂, 7 ♀) with simultaneous same arm BPs by O1 & O2 alternating with simultaneous opposite arm BPs by ABPMs. Results: The average (x ± SE) systolic (SBP) and diastolic (DBP) BPs for O1, O2, Oscar & Spacelabs ABPMS were 114.2/67.3 ± 2.0/1.8 mm Hg, 114.1/67.1 ± 2.0/1.8 mm Hg, 124.0/67.7 ± 2.1/1.4 mm Hg, and 119.3/67.9 ± 1.9/1.3 mm Hg, respectively. Compared to O1O2, the Oscar overestimated SBP by 9.8 ± 0.9 mm Hg (P < 0.001), while the Spacelabs overestimated SBP by 5.2 ± 0.8 mm Hg (P < 0.001). SBP also differed significantly between ABPMs (P < 0.001). Though the DBP difference was small between observers and ABPMs, (O1O2 - Oscar = -0.6 mm Hg; O1O2 - Spacelabs = -0.7 mm Hg), there was a high degree of variability with the Oscar ranging from 23 mm Hg above to 20 mm Hg below and the Spacelabs 23 mm Hg above to 26 mm Hg below the observers. Compared to O1O2, Oscar SBP and DBP differed by > 5 mm Hg in 15/17 (88%) and 7/17 (41%), while the Spacelabs SBP and DBP differed by > 5 mm Hg in 11/17 (65%) and 8/17 (47%) of subjects, respectively. Conclusions: Controlled lab testing revealed significant measurement errors in widely used oscillometric ABPMs. Given light exercise during 24-hr ambulatory monitoring, the outlook for accuracy and reliability appears worse. Oscillometric nomogram-like equations are population-specific and indirect 2nd generation estimations and cannot account for individual variations making them highly susceptible to errors, though more testing is required in a greater number of hypertensives.
Collapse
