Ear measurement of temperature is only useful for screening for fever in an adult emergency department

Abstract

Background: A new generation of ear thermometers with preheated tips and several measurements points should allow a more precise temperature measurement. The aim of the study was to evaluate if the ear temperature measured by this ear thermometer can be used to screen for fever and whether the thermometer is in agreement with the rectal temperature and if age, use of hearing devices or time after admission influences the temperature measurements.

Methods: Open cross-sectional clinical single site study patients, > 18 years old, who were acutely admitted to the short stay unit at the ED. A sample size of 99 patient per subgroup was recruited as random convenience series. As ear thermometer Braun Thermoscan Pro 4000® and as rectal thermometer Omron Flex Temp Smart ® was used. For different cut off of temperature the AUC was calculated and Bland-Altman analysis for calculation of 95% limits of agreement with rectal temperature, with subgroup analysis concerning age, time span from admission time and use of hearing aid.

Results: Among 599 patients the sensitivity to detect fever with an ear thermometer varied between 68 and 70% with AUC from 0.88-0.97. If the ear temperature was ≥37.5 oC, the sensitivity to detect patients with ≥38.0 oC rectally was 95% which raised to 100% for a rectal temperature of ≥38.3 oC. For the ear thermometer's ability to determine the exact temperature the 95% limits of agreement were +/- 0.8 oC. with no influence from age, duration of hospital stay or hearing aids.

Conclusion: The examined ear thermometer is able to detect fever, defined as ≥38 oC rectally in an adult ED population by using an ear cut-point of 37.5 oC, but not to measure the exact temperature. Used in this way around a fifth of the patients will still be in need of a rectal temperature measurement, but less than 5% with fever ≥38.0 oC will remain undetected and none with fever ≥38.3 oC. Age, admission time and use of hearing aid did not influence the temperature measurements.

Trial registration: Clinical Trials: ID NCT02977481 , date 11/18/2016.

Keywords: Fever; Rectal temperature; Temperature measurement; Tympanic.

Conflict of interest statement

Ethics approval and consent to participate

Approval from the Regional Committees on Health Research Ethics for Southern Denmark was waived after request (S-20160154). The study was registered with the Danish Data Protection Agency (16/39984).

Eligible patients were informed about the study and written information was provided. If the patient required time to consider participation, up to one hour was accepted. If the patient requested to discuss participation with a lay representative before a decision to participate was made, this was secured by a telephone call to the lay representative appointed by the patient if the representative was not present. When a written consent to participate was obtained the patient was included in the study.

Consent for publication

Not applicable.

Competing interests

Prof. Christian Backer Mogensen is associate editor for BMC Emergency Medicine.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Figures

Fig. 1

Recruitment and inclusion of patients

Fig. 2

Scatter diagram of rectal and ear temperature measurements

Fig. 3

Bland Altman plot for TM and rectal temperatures