ABSTRACT
Obtaining a patient’s temperature is an important part of a patient’s physical examination. As human medicine transitions to noninvasive temperature measurements, so does veterinary medicine. Historically, temperature measurement has been obtained from rectal readings; however, alternative methods, such as axillary and auricular temperatures, are increasing in popularity. The purpose of the study was to compare these alternative techniques to the gold standard of rectal temperature. Temperatures were obtained three ways for each patient: rectal, axillary, and auricular. Results indicated a positive linear relationship between rectal and axillary temperatures (bivariate correlation coefficient [r] = 0.65, P < .001) and axillary and auricular temperatures (r = 0.55, P < .001). Agreement was strongest between rectal and auricular temperatures (r = 0.80, P < .001). The average discrepancy between axillary and rectal temperature was 1.2°C [2.1°F] with the highest difference being 4.0°C [7.3°F]. The average discrepancy between auricular and rectal temperature was 0.6°C [1.2°F] with the highest difference being 2.2°C [4.1°F]. Despite auricular temperatures having stronger agreement, Bland-Altman Limits of Agreement testing revealed that it was a poor predictor of rectal temperature. Based on these results, axillary and auricular temperatures should not be substituted for rectal temperature.