Temporal Artery (TA) thermometry has become increasingly popular in the past decade, but no study to date directly compares its accuracy to that of axillary and oral thermometry in pediatric immunocompromised patients.

  • Due to the high risk of immunosuppressed patients for sepsis, accurate temperature readings are imperative.
  • This study examined a sample of immunocompromised pediatric patients in order to describe decisive thermometer performance on a specialized population.

The aim of this study was to evaluate whether TA temperature readings have comparable accuracy to oral and/or axillary measurements in the immunocompromised patient population.


  • This was a descriptive correlational research study.
  • Patients 6 months to 22 years with any immunosuppressive condition were enrolled.
  • Staff were instructed on study criteria, including verbal consent at time of temperature measurements and proper use of the Exergen TA thermometer.
  • Clinical decision algorithms in this subset of patients were based on whether the patient had a fever of at least 38 degrees and checked using TA along with a traditional oral or axillary route of measurement.
  • 154 children participated in the study, contributing to 313 measurements. Each sex had equal representation in the sample (77 male and female patients). Most patients were between the ages of 2-11.


  • TA measurements, compared to axillary and oral trend slightly higher (expected > 0.03°C more) when body temperature was near 36°C, essentially the same when body temperatures were between 37°C and 38°C, and lower (up to an expected difference of nearly 0.5°C) at higher temperatures.
  • The Bland-Altman analyses estimated a maximum expected difference of less than 1°C between TA and other measurements, even at extremely high levels of fever.
  • There was higher sensitivity with TA and axillary and higher specificity with TA and oral.
  • There were wider limits of agreement (LOA) between TA and axillary vs TA and oral measurements.
  • There was a higher bias when compared oral to TA (0.38) versus axillary to TA (-0.503).
  • Two temperatures (38.0 C and 38.5 C) were used to define fever. Temperatures correlated better with temperatures > 38.5 C. There was 5 % better overall agreement from oral and TA versus axillary and TA.


  • TA thermometry has shown variability in different physiologic conditions. This can also be true for both axillary and oral routes.
  • For low grade fevers, TA is at a different defined range as shown in literature which would need more quantification.
  • TA thermometry has been referenced as core temperature, however, secondary to the sensitivity and vast movement/fluctuations reported during fevers, it has shown wide variations. These wide variations show TA thermometer’s ability to quickly detect a change in body temperature.
  • Reynolds et al. (2014) compared TA thermometry to axillary which showed bias and precision values favoring the TA thermometer. In 52 pediatric ED children < 4 years old, “The percentage of temporal artery and axillary temperatures that were >± 1.0°C and/or >± 1.5°C above or below the clinical reference temperature 15% and 6%, respectively, for the temporal artery thermometer and 39% and 14%, respectively, for the axillary thermometer.”
  • In children, the oral route is not always appropriate. Kiekkas et al. (2019) mentioned TA thermometry accuracy and precision in children were much higher compared to adults with a mean difference of -0.17 degrees C (32. F) or 95% LOA (-1.14 to 0.79 degrees C (34.0-33.4 F).
  • Barnason (2011) concluded oral and TA thermometers were highly recommended for use in children three to 18 years of age febrile or not, whereas axillary thermometers were not recommended in febrile patients and only moderately recommended in non-febrile patients.


  • TA thermometer should be considered as a first choice over oral or axillary for temperature measurement in immunocompromised pediatric patients.
  • When checking temperatures with a TA, oral or axillary thermometer, consider rechecking another temperature site if there are concerns for fever.
  • Displaying and analyzing higher temperature comparisons would be valuable to clearly document when overestimations or quick decline in fevers occur.
  • Future studies may be needed to further validate its use in immunocompromised pediatric patients.
  • Limitations to this study include wide age group categories. There were also relatively few temperatures in the 38.0-degree range recorded in this study, yet they were most influential describing the TA measurement discrepancy TA thermometry has shown variability in different physiologic conditions. This can also be true for both axillary and oral routes.

References and Acknowledgements

  • References available upon request at
  • Acknowledgements: Jennifer Anders and Kevin