Objective: To assess the agreement between noncontact infrared thermometer (noncontact) with infrared tympanic thermometer (tympanic) and electronic axillary thermometer (axillary) in an adult emergency department population.
Materials and methods: This is a single-center, cross-sectional, prospective trial carried out in a Joint Commission accredited private hospital in Turkiye. All consecutive patients above 16 years were included in the study. The agreements between three methods were analyzed by Bland-Altman analysis with MedCalc 11.0.4 statistical software.
Results: Body temperatures were measured on 400 patients (48% were men, mean 35.9±17.3°C). Mean noncontact, tympanic, and axillary measurements (±SD) were 37.22±1.03, 36.72±0.95, and 36.91±0.96°C, respectively, whereas Intraclass Correlation Coefficient of all measurements was 0.892 (95% confidence interval 0.821-0.929). Binary comparisons between body temperature measurements produced mean differences Δ axillary-tympanic, Δ axillary-noncontact, and Δ tympanic-noncontact as 0.5±0.63, 0.2±0.71, and 0.31±0.61°C, respectively. However, the agreement limits for axillary and noncontact was between -1.2 and 1.6°C; -1.74 and 0.74°C for tympanic and noncontact, and -1.52 and 0.9°C for tympanic and noncontact.
Conclusion: There is a lack of agreement between body temperature measurements by noncontact, tympanic, and axillary in the adult emergency department population. The easy application may lead noncontact to be the preferable method for healthcare providers but large agreement limits should be considered.