Validation of e-Celsius gastrointestinal telemetry system as measure of core temperature

Abstract

The main objective of this study was to validate gastrointestinal measurement with the e-Celsius® system composed of an ingestible electronic capsule and a monitor. Twenty-three healthy volunteers aged 18–59 years stayed at the hospital for 24 h under fasting conditions. They were only allowed for quiet activity and were asked to keep their sleeping habits. Subjects ingested a Jonah capsule and an e-Celsius® capsule, and a rectal probe and an esophageal probe were inserted. Mean temperature measured by the e-Celsius® device was lower than that measured by Vitalsense® (−0.12 ± 0.22°C; p < 0.001) and the rectal probe (−0.11 ± 0.03°C; p = 0.003) and higher than that measured by the esophageal probe (0.17 ± 0.05; p = 0.006). Mean difference (bias) and 95% confidence intervals between temperature of e-Celsius capsule, Vitalsense Jonah capsule, esophageal probe, and rectal probe were computed using Bland and Altman procedure. The magnitude of the measurement bias is significantly greater when comparing the e-Celsius® and the Vitalsense® device pair with any other device pairs containing the esophageal probe. Amplitude of confidence interval between the e-Celsius® system and the Vitalsense® system was 0.67°C. This amplitude was significantly lower than those of the esophageal probe–e-Celsius® pairing (0.83°C; p = 0.027), of the esophageal probe–Vitalsense (0.78°C; p = 0.046) and of the esophageal probe–rectal probe (0.83°C; p = 0.002). The statistical analysis did not reveal any effect of time on the amplitude of bias, whatever the device concerned. When comparing missing data rate of the e-Celsius® system (0.23 ± 0.15%) and the Vitalsense® devices (0.70 ± 0.11%) during the whole experiment, no differences was observed (p = 0.09). The e-Celsius® system could be used when a continuous following of internal temperature is needed.