Comparison of temporal artery thermometer and other body temperature measurement methods in children aged 1-36 months

Background: Fever is defined as above normal body temperature and is the most common reason for pediatric emergency departments visits. It is still a source of debate as to which non-invasive core body temperature measurement method is the most useful and accurate. In this study, we consider the rectal temperature measurement as the reference core body temperature and compare the temporal artery thermometers (TAT) accuracy, sensitivity, and specificity with cutaneous, and axillary thermometer measurements, and their correlation. The aim of the study is to evaluate each temperature measurement method and to find the highest sensitivity and specificity, which makes the closest measurement to rectal measurement. Materials and methods: This prospective study includes children aged 1 to 36 months, without indications of chronic diseases and neutropenia, presented to the pediatric emergency department of a large city hospital. Rectal, axillary, cutaneous and temporal artery temperatures are obtained. Body temperature was also evaluated tactilely by the mothers. Sensitivity, specificity, positive and negative predictive values are determined by data analysis utilizing appropriate statistical methods. Results: Rectal body temperature is found to be significantly higher than the measurements made from other body parts. The differences between the averages of the rectal and TAT measurement values ​​are found to be 0.7 ± 0.7 SD °C. TAT measurements are found to provide the closest values to the rectal measurements compared to the other methods, while the differences between the averages of the rectal and axillary measurements are found to provides measurements furthest away from the rectal temperature. All measurement methods are found to be correlated with rectal measurements. Axillary thermometer measurements has the strongest correlation. All methods are found to be compatible with the rectal measurement method with Bland-Altman analysis. ROC analysis is performed for the temporal artery thermometer cut-off value and the 37.7°C cut-off value is found to be the ideal. Among temperature cut-off values of 37.5°C, 37.7°C and 38°C, the highest specificity (60.5%) and sensitivity (95.7%) is observed at 37.7°C cut-off value for the temporal artery measurement method when a rectal temperature of 38°C is considered as reference for fever. In non-contact cutaneous and axillary measurements, optimum sensitivity (34.9%, 20.9%, respectively) and specificity (95.7%, 98.9%, respectively) are achieved at a cut-off value of 37.5°C. The sensitivity and specificity of the tactile assessment of mothers are observed to be 48.8% and 94.6% respectively. The sensitivity of maternal sensation of fever in infants aged 1 to 3 months is observed to be 66.7% and higher than that of thermometers. In the infants aged 1 to 3 months, the ideal cut-off value of temporal artery thermometer is found to be 38°C with a sensitivity of 33.3% and a specificity of 100%. Conclusions: Peripheral thermometer measurements cannot replace core body temperature measurements. In patients aged 1 to 36 months, when 37.7°C is considered as the cut-off temperature, the temporal artery thermometer provides the most reliable body temperature measurement. TAT measurements are more accurate than non-contact cutaneous, and axillary measurements for patients aged 1 to 36 months. In the 1 to 3 months age group, C non-contact thermometers provide the best measurements. Mothers tactile body temperature assessment is highly sensitive and specific. Keywords: Body temperature measurement, Fever, Thermometer, Temporal artery, Axillary