Evaluation of the non-invasive Temple Touch Pro temperature monitoring system compared with oesophageal temperature in paediatric anaesthesia (PETER PAN): A prospective observational study.

BACKGROUND: Monitoring peri-operative body temperature in children is currently mainly achieved through invasive devices. The Temple Touch Pro Temperature Monitoring System estimates core temperature noninvasively based on heat flux thermometry. OBJECTIVE: To investigate the agreement of this noninvasive sensor against standard oesophageal core temperature. DESIGN: A prospective observational study. SETTING: University hospital recruiting between April and July 2021. PATIENTS: One hundred children (32 girls) aged 6 years or younger scheduled for noncardiac surgery, resulting in 6766 data pairs. Exclusion criteria were contraindication for the insertion of an oesophageal temperature probe, and procedures in which one of the measurement methods would interfere with the surgical field. MAIN OUTCOME MEASURES: Primary outcome was the agreement analysis by a Bland-Altman comparison with multiple measurements. Posthoc, we performed another agreement analysis after exclusion of a statistically determined equilibration time. Secondary outcomes were the temperature differences over time and subgroup analysis of hypothermic, normothermic and hyperthermic temperature ranges, age, sex and sensor's side by type III analysis of variance. Further, we correlated the sonographically determined depth of the artery with trueness. RESULTS: The mean difference was -0.07°C (95% CI -0.15 to +0.05) with limits of agreement of -1.00 and +0.85°C. After adjusting for an equilibration time of 13 min, the mean difference improved to -0.04°C (95% CI -0.08 to +0.01) with limits of agreement of -0.68 and +0.60°C. Concordance correlation coefficient was 0.83 (95% CI 0.82 to 0.84). Differences between the skin sensor and oesophageal reference increased over time by -0.05°C per hour. Subgroup analysis showed no clinically relevant differences. Depth of artery negatively correlated with trueness by 0.03°C per millimetre. CONCLUSIONS: Although the Temple Touch Pro sensor showed acceptable accuracy after allowing for an equilibration time, it still needs further investigation for routine use in children. This particularly affects accuracy in hypothermic ranges, imprecise positioning and applicability in children with immature or vulnerable skin. TRIAL REGISTRATION: German Clinical Trials Register, identifier: DRKS00024703.