Abstract

The use of tympanic thermometers has markedly increased over the past few years. While some authors have shown high correlation between rectal and tympanic readings, others have found low correlation. After noticing a failure of the tympanic thermometer to detect fever in obviously febrile pediatric patients, we prospectively evaluated its use in our emergency department. All patients less than 6 years of age who were brought in for care over a 2-week period were eligible for entry into the study. Nurses were trained in the proper use of the instruments before the study began. We used a calibrated tympanic thermometer (First Temp 2000A) and an electronic thermometer (IVAC 2080A). In all children the tympanic temperature was measured first, immediately followed by measurement of the rectal temperature. The 39 children entered into the study ranged in age from 2 to 71 months (mean of 27 months). Rectal temperature ranged from 97.8° to 105.2°F. The correlation between rectal and tympanic was low (r = .830) and was worse in patients with fever (r = .612). The mean difference was 0.8°F, but was as high as 3.4°F. The tympanic thermometer failed to identify five of the 15 febrile patients (T ≥ 100.5°F) in our study. We therefore conclude that the tympanic thermometer is not sensitive enough to be used to screen for fever in pediatric patients.