Continuous remote monitoring of postoperative vital signs with the Biobeat patch: A multicenter prospective observational study

Abstract

Objective This study, conducted under standard clinical conditions, evaluated the Biobeat wearable sensor in postoperative patients monitored over multiple days without human intervention. The study aimed to assess technical performance (vital sign data collection and artifact frequency), reliability (agreement with nurse-measured values), the sensor’s ability to detect vital sign abnormalities, patient satisfaction, and incidence of adverse skin reactions. Design This multicenter prospective observational study included patients who underwent abdominal surgery with a planned postoperative ward stay between December 2020 and December 2022. Methods Alongside routine care–comprising (2–3 times daily nurse assessments of blood pressure (BP), pulse rate (PR), respiratory rate (RR), peripheral oxygen saturation (SpO2), and temperature)–participants wore a precordial Biobeat sensor. The sensor transmitted data wirelessly to a cloud-based repository via Wi-Fi without human intervention. Results Of 109 enrolled patients, 90 were included in the analysis. The median recording duration was 63.8 hours (interquartile range: 42.6–72.3 hours). Data loss occurred in 45% of the recording time, and pulse oximetry data were absent in 16% of the available measurements. Artifacts were infrequent, comprising 277 of 35,998 measurement points. Compared with nurse measurements, Biobeat showed small mean differences for most variables, except for RR and temperature, yet exhibited wide limits of agreement across all variables. Clarke Error Grid analysis revealed excellent concordance between Biobeat and nurse measurements for SpO2 and PR, with lower agreement for mean BP, RR, and temperature. The sensor’s continuous monitoring detected more vital sign abnormalities, including severe hypotension (mean BP < 60 mmHg), in 10.0% of patients compared to 3.3% with nurse measurements (P = 0.04). The patients reported high satisfaction and no adverse skin reactions were observed. Conclusions Significant data loss presents a substantial challenge for analysis, underscoring the critical need for improved data transmission and flow control measures in study protocols and clinical deployment. Registry ClinicalTrials.gov (NCT04585178, October 14, 2020).