Accuracy of temporal artery thermometry in neonatal intensive care infants.

PURPOSE To determine the accuracy of temporal artery and axillary temperatures and the discomfort level of stable neonates during temperature measurement. SUBJECTS Convenience sample of neonates between the ages of 32 and 40 weeks' gestation cared for in an isolette or crib. DESIGN A method-comparison design was used to compare different methods for noninvasive temperature monitoring (infra-red temporal artery; axillary electronic) to core body temperatures (indwelling rectal probe). MAIN OUTCOME MEASURE Bias and precision of test temperature devices (temporal artery; axillary). RESULTS Bias and precision for the temporal artery and axillary devices were 0.30 ± 0.44 and 0.28 ± 0.33, respectively. Analysis of variance found significant differences between both temporal and axillary temperatures compared to rectal temperatures (P < .01). Statistical differences were small and did not represent a clinically important difference. No statistical difference was found between temporal artery and axillary temperatures (P = .81). Increases in neonate discomfort after temperature measurement were significantly greater with axillary than increases after temporal artery temperature measurement (P = .03). CONCLUSIONS This study found that body temperature measured with the temporal artery thermometer was similar to temperatures obtained with an axillary thermometer in stable, afebrile neonates. The use of temporal artery thermometry appears to be an acceptable approach for noninvasive temperature measurement in neonates, which causes less discomfort in neonates.