Reliability of infrared ear thermometry in the prediction of rectal temperature in older inpatients
Article first published online: 14 JAN 2009
© 2009 The Authors. Journal compilation © 2009 Blackwell Publishing Ltd
Journal of Clinical Nursing
Volume 18, Issue 3, pages 451–456, February 2009
How to Cite
Smitz, S., Van de Winckel, A. and Smitz, M.-F. (2009), Reliability of infrared ear thermometry in the prediction of rectal temperature in older inpatients. Journal of Clinical Nursing, 18: 451–456. doi: 10.1111/j.1365-2702.2008.02565.x
- Issue published online: 14 JAN 2009
- Article first published online: 14 JAN 2009
- Accepted for publication: 11 July 2008
- body temperature;
- infectious diseases;
- older patients;
- tympanic thermometer
Aims and objectives. To assess the capability of infrared ear thermometry accurately to predict rectal temperature in older patients.
Background. Infrared ear thermometry is now commonly used for predicting body temperature in older patients. However, ear thermometry has been insufficiently evaluated in geriatric patients.
Design. Prospective, convenience sample, unblinded study.
Methods. All patients (or their guardians) gave informed consent. Patients hospitalised in a geriatric unit underwent sequential ear and rectal temperatures measurements using two different models of infrared ear thermometers (ThermoScan and Genius) and a rectal probe, respectively. After a brief otoscopic examination, ear temperatures were measured twice at both ears with each thermometer, the highest of four measurements being retained for analysis. The rectal temperature was the reference standard.
Results. Hundred patients (31 males), aged 81 (SD 7) years completed the study. The mean rectal temperature was 37·3 °C (SD 0·7) °C (range 36·3–40·7 °C). Eighteen patients were febrile (rectal temperature ≥ 37·8 °C). The mean bias between rectal and ear temperatures as measured with the ThermoScan was –0·20 °C (SD 0·32) °C and the 95% limits of agreement were –0·83 °C and 0·42 °C (95% CI, –0·88–0·48 °C). Using the Genius, the corresponding figures were –0·56 °C (SD 0·39) °C, –1·32 °C and 0·20 °C (95% CI, –1·39–0·27 °C). After correction for bias, the ThermoScan predicted the level of fever with a maximum error of 0·7 °C (mean error 0·3 °C). Using the Genius, the maximum error and the mean error were 1·6 °C and 0·4 °C, respectively.
Conclusions. Infrared ear thermometry can predict rectal temperature in normothermic and in febrile inpatients with an acceptable level of accuracy. However, the predictive accuracy depends on both operator technique and quality of instrumentation.
Relevance to clinical practice. Proper technique (measuring in both ears) and optimal instrumentation (model of ear thermometer) are essential for accuracy.