We read with great interest Hay et al's article on infrared thermometry.1 They used axilla temperature as the gold standard against which the infrared tympanic membrane temperature (TMT) was compared. The fundamental fault of this study lies with the fact that axilla temperature was not a gold standard for body temperature measurement. It was found to be poorly correlated with core temperature by a previous meta- analysis.2 The best reference site would be rectal temperature — the generally accepted gold standard for temperature measurement in children.3 As exercise increases body temperature, the study was further hampered by a lack of stipulated rest before measurement.
In a previous study done in this department, we found infrared thermometers to be reliable in neonates;4 previous meta-analysis showed such thermometers to be unreliable.5 However, the main problem of the meta-analysis was grouping different infrared ear thermometers as one single entity when each brand should be tested on its own.6 The accuracy of TMT could be enhanced by taking two readings and recording the higher one as representative of core temperature.4
- © British Journal of General Practice, 2004.
