Objective: The aim of the study is to validate a tympanic thermometer (TT) use in primary care.
Patients and methods: Validation study phase IV of the tympanic thermometer ThermoScan Pro LT. Measurements are compared to axillary temperature (AT) measured with a glass mercury thermometer. 412 subjects 6 months to 15 years old attended in primary care clinics were studied. Exclusion criteria were middle ear conditions, intense crying or sweating. The highest of the three measurements in each ear was used to calculate sensitivity and specificity of the TT and to analyse the concordance with AT. The impact of ambient temperature, age, gender, cerumen, presence of a febrile condition and the examiner on the results was assessed.
Results: Sensitivity of TT to detect fever (AT > 38 degrees C) was 89.7% (95% IC: 81.8-97.5) and specificity 90.7% (95% IC: 87.7-93.7). As a mean TT measures 0.5 degree C higher than AT but there was a high dispersion of the data (-1.59-0.6 degrees C). Accuracy of TT was influenced by the sex of the patient (p = 0.029), examiner (p < 0.01), presence of a febrile condition (p < 0.01) and positioning of the patient (p = 0.026). Repeatability coefficient was 0.34 degree C for the right ear and 0.42 degree C for the left.
Conclusion: The sensitivity of the TT to detect fever was adequate, though there was a poor concordance with AT and had an important variation among repeated measurements in the same ear.