Accuracy of non-invasive body temperature measurement methods in adult patients admitted to the intensive care unit: a systematic review and meta-analysis.

Non-invasive thermometers are widely used in both clinical practice and trials to estimate core temperature. We aimed to investigate their accuracy and precision in patients admitted to the intensive care unit (ICU). Systematic review and meta-analysis. We searched MEDLINE, EMBASE and the Cochrane Central Register of Controlled Trials to identify all relevant studies from 1966 to 2017. We selected published trials that reported the accuracy and precision of non-invasive peripheral thermometers (index test) in ICU patients compared with intravascular temperature measurement (reference test). The extracted data included the study design and setting, authors, study population, devices, and body temperature measurements. Two reviewers performed the initial search, selected studies, and extracted data. Study quality was assessed using the QUADAS-2 tool. Pooled estimates of the mean bias between index and reference tests and the standard deviation of mean bias were synthesised using DerSimonian and Laird random effects meta-analyses. We included 13 cohort studies (632 patients, 105 375 measurements). Axillary, tympanic infrared and zero heat flux thermometers all underestimated intravascular temperature. Only oesophageal measurements showed clinically acceptable accuracy. We found an insufficient number of studies to assess precision for any technique. Study heterogeneity was high (99-100%). Risk of bias for the index test was unclear, mostly because of no device calibration or control for confounders. Compared with the gold standard of intravascular temperature measurement, non-invasive peripheral thermometers have low accuracy. This makes their clinical and trial-related use in ICU patients unreliable and potentially misleading.