Performance of a cuffless blood pressure monitor in the adult and elderly population - a comparative validation study

Abstract

Abstract Background Cuffless blood pressure (BP) monitoring devices are emerging as powerful tools for BP monitoring and management of hypertension, particularly in the elderly, in whom prevalence of hypertension is high. Some cuffless BP monitors allow the hourly and daily tracking of blood pressure. It remains unclear, however, whether cuffless devices yield equivalent performance across age categories. Purpose To compare the performance of BP measurements conducted by a cuffless BP monitor in different body positions in adult and elderly participants with a wide range of BP levels. Methods 2 cohorts of patients (adult, 21-65 years; elderly, 65-85 years) were recruited for a prospective clinical study (NCT04027777) between August/2019–March/2022. During 4 visits over a month, systolic/diastolic BP (SBP, DBP) measurements obtained by the investigated cuffless wrist bracelet were compared to the reference measurement – double auscultation – in a sitting, standing and lying position after a unique initial calibration. The two ISO 81060-2 criteria were used for the comparison of BP measurements by the cuffless monitor and auscultation (BP-reading-level – criterion 1; subject-level – criterion 2) in the sitting position (there are no ISO criteria for other body positions). The mean differences on SBP and DBP between cohorts both at BP-reading-level and at subject-level were compared (unpaired t-test). Results 96 adult (38±12 years)/35 elderly (73±6 years) were included in the study. Of those, 84 adult/35 elderly patients were investigated in sitting position; 91 adult/35 elderly in all body positions. SBP and DBP were significantly higher in the elderly cohort (SBP 116.4±20.3 vs. 133.9±17.7 mmHg, P≤0.0001; DBP 74.5±12.0 vs. 77.9±11.7 mmHg, P≤0.0001). The results for the analyses at BP-reading-level and subject-level are shown in Table 1. For the analyses on sitting position at BP-reading-level, SBP and DBP for both cohorts were within ISO targets for mean error and SD. At subject-level, the SD of the mean paired differences per subject were within individual ISO targets for SBP and DBP on both cohorts. Statistically significant differences between cohorts were found only at BP-reading-level for all body positions, although the magnitude of differences were not clinically relevant (Table 1, bold and *). Figure 1 highlights the biases and correlations of SBP and DBP readings between the cuffless BP monitor and auscultation for both adult and elderly cohorts. Conclusions The cuffless BP monitor investigated in the present study yielded equivalent BP measurements compared to auscultation, with similar performance in the adult and elderly populations.Table 1Figure 1