Objective: To compare the measurements of body temperature with the tympanic infrared thermometer and the digital rectal thermometer.
Design: Prospective, comparative.
Setting: Beatrix Hospital, Gorinchem, the Netherlands.
Patients and methods: A total of 2057 almost simultaneous measurements of rectal and tympanic temperature were performed in 164 patients in 9 different wards.
Results: The mean difference between the two methods was 0.45 degree C with a standard deviation of 0.57 degree C. The tympanic temperature was lower than the rectal temperature. The differences ranged from -1.5 to 3.6 degrees C. The correlation coefficient was 0.69. If a rectal temperature > 37.8 degrees C was applied as the criterion of fever, the diagnosis was not made in 175/291 measuring moments (60%) with the tympanic thermometer. If a tympanic temperature > 37.8 degrees C was applied as the criterion of fever, the rectal thermometer failed to show fever in 16/132 measuring moments (12%).
Conclusion: The low sensitivity of the tympanic measurement to establish fever renders the tympanic infrared thermometer unsuitable for use as a fever thermometer.