Resumen de Comparison of rectal and axillary temperatures in dogs and cats
Objective—To compare rectal versus axillary temperatures in dogs and cats.
Design—Prospective observational study.
Animals—94 dogs and 31 cats.
Procedures—Paired axillary and rectal temperatures were measured in random order with a standardized method. Animal signalment, initial complaint, blood pressure, blood lactate concentration, and variables associated with vascular perfusion and coat were evaluated for associations with axillary and rectal temperatures.
Results—Axillary temperature was positively correlated with rectal temperature (ρ = 0.75 in both species). Median axillary temperature (38.4°C [101.1°F] in dogs, and 38.4°C [101.2°F] in cats) was significantly different from median rectal temperature in dogs (38.9°C [102.0°F]) but not in cats (38.6°C [101.5°F]). Median rectal-axillary gradient (difference) was 0.4°C (0.7°F; range, −1.3° to 2.3°C [−2.4° to 4.1°F]) in dogs and 0.17°C (0.3°F; range −1.1° to 1.6°C [−1.9° to 3°F]) in cats. Sensitivity and specificity for detection of hyperthermia with axillary temperature were 57% and 100%, respectively, in dogs and 33% and 100%, respectively, in cats; sensitivity and specificity for detection of hypothermia were 86% and 87%, respectively, in dogs and 80% and 96%, respectively, in cats. Body weight (ρ = 0.514) and body condition score (ρ = 0.431) were correlated with rectal-axillary gradient in cats.
Conclusions and Clinical Relevance—Although axillary and rectal temperatures were correlated in dogs and cats, a large gradient was present between rectal temperature and axillary temperature, suggesting that axillary temperature should not be used as a substitute for rectal temperature.
Accurate assessment of body temperature is an important aspect of the physical examination in veterinary patients. Rectal temperature is the standard measurement used to assess body temperature in animals, but the use of axillary temperature has been described1,2 and is often used as a substitute. This is true especially when measurement of rectal temperature is complicated by animal temperament or rectal lesions or when frequent temperature monitoring is required. Although multiple studies3–25 evaluating axillary temperature in human populations have been conducted, only 1 study2 has critically evaluated axillary temperature in dogs, and to the author's knowledge, axillary temperature has not been evaluated in cats.
A large number of human studies have compared axillary temperature with internal body temperature. Although some data are conflicting, most studies have detected a wide variation between axillary and rectal temperatures in adults as well as in pediatric and neonatal populations.4,5,7–12,22,24 Sensitivity of axillary temperature for detecting fever ranges from 25% to 87.5%,8,11–13 with increased sensitivity in neonatal populations.11,17,19,21 Patient age,7,11,12 body condition,16 time of day,4 axilla of measurement,23 and presence of hemiplegia18 affect the reliability of axillary temperature, whereas IV fluid infusion,3 environmental temperature,9,12 and amount of clothing worn9 do not. Weight, body condition, breed, and coat length are associated with differences between axillary and rectal temperature in dogs.2 In humans, the gradient between core temperature and the temperature of the great toe correlates with the severity of circulatory disease,26,27 and increasing temperature of the toe in response to treatment is associated with an improved prognosis.26 Although measurement of axillary temperature is convenient and may reduce patient stress, compared with measurement of rectal temperature, it is important to establish the accuracy of axillary temperature measurements in dogs and cats before use in clinical practice. Inaccurate body temperature measurements can lead to inappropriate diagnostic and treatment decisions that may affect overall patient care. Therefore, the purpose of the study reported here was to compare axillary with rectal temperature in dogs and cats and evaluate multiple clinical variables to identify factors that may alter the accuracy of axillary temperature measurements. The hypothesis was that, as in humans, axillary temperature would not be strongly correlated with rectal temperature.
