Objective: Ambulatory blood pressure monitoring (ABPM) is used for confirmation of diagnosis of hypertension. However, conventional cuff-based blood pressure (BP) measurement involves the obtrusive procedure of cuff inflation, making multiple measurements over extended periods uncomfortable. This study aims to compare BP measurement using the unobtrusive procedure of radial applanation tonometry with that of the conventional brachial cuff sphygmomanometer for ABPM. Design and method: ABPM (24 hour) was conducted in 16 participants (53±14 years, 4 female) using validated brachial cuff devices (Microlife WatchBPO3, n = 6; BPLab, n = 10) with BP measurements every 15 minutes during daytime and every 30 minutes during nighttime. BP measurements were obtained with a wrist-worn tonometric device (BPro, Healthstats) every 15 minutes, using an initial calibration obtained with the ABPM device. Results: Averaged over 24 hours, tonometric BP estimates correlated with brachial cuff systolic BP (SBP, r = 0.89, p<0.001) but not with diastolic BP (DBP, r = 0.48, p = 0.059) with mean differences (bias) of -4±14 mmHg (SBP) and 16±17 mmHg (DBP). Daytime and night-time correlations and mean difference were, day: SBP (r = 0.92, p<0.001), -1±13 mmHg; DBP (r = 0.44, p = 0.089), 18±21 mmHg; night: SBP (r = 0.66, p = 0.007), 6±22 mmHg; DBP (r = 0.64, p = 0.010); 15±13 mmHg. Over 24 hours, the total number of successful readings from the tonometric device was 52% less than that from the brachial cuff devices. The average correlation between SBP and DBP using only cuff readings was 0.29±0.40 (range -0.47 to 0.86). The average correlation between SBP and DBP for the tonometric device was 0.99±0.01 (range 0.97 to 1.00). Conclusions: The cuffless tonometric method employed in the BPro device enables unobtrusive monitoring of BP. In comparison to conventional cuff ABPM devices, the BPro device showed low mean differences for SBP but large differences in DBP in both day and night measurements with consistently low precision. The large differences for DBP may be due to the high correlation between tonometric DBP and SBP, suggesting that waveform features are used for SBP estimation only, from which DBP is directly calculated.
Read full abstract