Clinical validation of a novel accelerometer-based respiratory rate monitor

Abstract

Respiratory rate (RR) is a vital sign often unused in clinical practice due to labour-intensive manual counting or capnography limitations in spontaneously breathing patients. This study validated RespX, a novel accelerometer-based wireless sensor, for continuous respiratory rate measurement in hospitalised patients. This prospective, single-centre observational study included hospitalised adults (> 18 years) expected to remain undisturbed for one hour while measuring RR. The RespX sensor, placed on the left chest wall, was compared with continuous capnography and manual counting every 20 min for one hour. The primary endpoint was respiratory rate measured by RespX within ± 3 breaths per minute compared to capnography. Linear mixed-effects models analyzed mean differences between methods. Twenty-nine patients completed the study. The proportion of RespX measurements within ± 3 breaths per minute of capnography was 77% (95% CI: 74.8–79.0%) for raw data and 82.8% (95% CI: 80.7–84.7%) after removing outliers. The proportion of RespX measurements within ± 3 breaths per minute of manual counting was 87,5% (95% CI: 81,1%-93,8%). For the capnography the proportions for the raw data was 87,8% (95% CI: 81,8–93,8%) and 94,5% (95% CI: 90,3%-98,7%) after removing outliers. Linear mixed-effects analysis showed mean differences of -0.40 (p = 0.43) for raw data and 0.39 (p = 0.15) for filtered data between RespX and capnography, supporting their equivalence. RespX demonstrated lower measurement variability (SD: 0.75–3.05) compared to capnography (SD: 0.99–5.19). No adverse events were observed during the study. The wireless RespX accelerometer-based sensor provides a novel and accurate method for assessing the respiratory rate in awake hospitalised patients compared with capnography. Trial registration: NCT06911541.