Comparison of neonatal skin sensor temperatures with axillary temperature: does skin sensor placement really matter?

Purpose: Appropriate thermoregulation affects both morbidity and mortality in the neonatal setting. Nurses rely on information from temperature sensors and radiant warmers or incubators to appropriately maintain a neonate's body temperature. Skin temperature sensors must be repositioned to prevent skin irritation and breakdown. This study addresses whether there is a significant difference between skin sensor temperature readings from 3 locations on the neonate and whether there is a significant difference between skin sensor temperatures compared with digital axillary temperatures.

Subjects: The study participants included 36 hemodynamically stable neonates, with birth weight of 750 g or more and postnatal age of 15 days or more, in a neonatal intensive care unit. Gestational age ranged from 29.6 to 36.1 weeks at the time of data collection.

Design: A method-comparison design was used to evaluate the level of agreement between skin sensor temperatures and digital axillary thermometer measurements.

Methods: When the neonate's skin sensor was scheduled for routine site change, 3 new skin sensors were placed-1 each on the right upper abdomen, left flank, and right axilla. The neonate was placed in a supine position and redressed or rewrapped if previously dressed or wrapped. Subjects served as their own controls, with temperatures measured at all 3 skin sensor sites and followed by a digital thermometer measurement in the left axilla. The order of skin sensor temperature measurements was randomly assigned by a computer-generated number sequence.

Main Outcome Measures: An analysis of variance for repeated measures was used to test for statistical differences between the skin sensor temperatures. The difference in axillary and skin sensor temperatures was calculated by subtracting the reference standard temperature (digital axillary) from the test temperatures (skin temperatures at 3 different locations), using the Bland-Altman method. The level of significance was set at P < .05.

Principal Results: No statistically significant differences were found between skin temperature readings obtained from the 3 sites (F2,70 = 2.993, P = .57). Differences between skin temperature readings and digital axillary temperature were also not significant when Bland-Altman graphs were plotted.

Conclusions: For hemodynamically stable neonates in a supine position, there were no significant differences between skin sensor temperatures on abdomen, flank, or axilla or between skin sensor temperatures and a digital axillary temperature. This may increase nurses' confidence that various sites will produce accurate temperature readings.