UNRELIABLE BLOOD PRESSURE MEASUREMENTS USING VALIDATED AUTOMATED UPPER ARM CUFF DEVICES IN 861 INDIVIDUALS: PREVALENCE AND DETERMINANTS OF THE PHENOMENON

Objective: Validated automated upper-arm cuff devices are recommended for blood pressure (BP) measurement, but even validated ones may be inaccurate in some individuals. This study investigated the prevalence and the determinants of unreliable automated BP (UABP) measurement. Design and method: Data on validation studies in adults conducted at Hypertension Center STRIDE-7 using established protocols [ESH-IP 2002/2010, ANSI/AAMI/ISO 2013, AAMI/ESH/ISO Universal Standard] were analysed. Three BP pairs (Ref-Test-Ref) were recorded per participant. Persistent unreliable test systolic BP (UASBP) was defined as >=2 systolic BP (SBP) pairs showing >10 mmHg difference between test and reference measurements per individual. Similarly, persistent unreliable diastolic BP (UADBP) was defined as >=2 diastolic (DBP) pairs with >10 mmHg difference. UABP was defined as UASBP and/or UADBP. Results: 861 adult participants [45% women, mean±SD age 55.9±14.7 (range 21-92) years, arm circumference 31.6±4.7 cm, SBP/DBP 131±21/82±15] were investigated in 14 validation studies (ESH 2002/2010 N=192/33, ANSI/AAMI/ISO 2013 N=51, AAMI/ESH/ISO N=585) on 13 devices (all upper-arm cuff; 4/6/1/2 for office/home/ambulatory/kiosk use), were analysed. UABP occurred in 12.2% (UASBP/UADBP 7.2/5.5%). 12.1/12.1/10.9% of individuals had a >10 mmHg difference in the 1st/2nd/3rd SBP pair (agreement: 83/84%, kappa=0.20/0.22, 1st vs 2nd/2nd vs 3rd) and 8/7.1/8.5% in the 1st/2nd/3rd DBP pair (agreement: 92/89%, kappa=0.43/0.26, 1st vs 2nd/2nd vs 3rd). The automated devices underestimated SBP at higher reference SBP values, and overestimated DBP at lower reference DBP values (Figure). Linear regression showed that average test-reference difference in 3 SBP pairs were determined by arm circumference (beta=0.1, p<0.01) and reference SBP (beta=-0.1, p<0.001), whereas DBP differences were determined by age (beta= 0.1, p<0.001), female sex (beta=-1.1, P<0.01) and reference DBP (beta=-0.1, p<0.001). Logistic regression analysis demonstrated that persistent UABP was determined by reference SBP (OR=1.02, p<0.05). Conclusions: The data suggest that UABP is a common phenomenon occuring in about 12% of people using validated devices and it is mainly determined by outlier BP levels.