RESUMO
OBJECTIVES: Fever is among the most common reasons for emergency department (ED) visits by children. This study compares temporal artery thermometry to rectal thermometry in febrile children in an ED. METHODS: This was a retrospective evaluation of children younger than 36 months treated consecutively in an urban medical center. Patients underwent triage with temporal artery thermometry, and after transfer to the pediatric ED, they underwent rectal thermometry. Fever was defined as rectal temperature of 100.4 °F (38 °C) or greater, and 147 patients met this definition. Data extraction from electronic charts obtained paired temporal artery and rectal temperatures, and these were compared by Bland-Altman analysis. Temperature points of 100.4 °F (38 °C) and 102.2 °F (39 °C) were evaluated to compare temporal artery thermometry with rectal thermometry sensitivity and specificity. RESULTS: A statistically and clinically significant difference between temporal artery and rectal temperature was found. Temporal artery thermometry was 53% sensitive detecting rectal temperature 100.4 °F (38 °C) or greater, and 27% sensitive detecting rectal temperature of 102.2 °F (39 °C) or greater. Mean rectal temperature was 102.36 °F (39.09 °C) (95% confidence interval [CI], 102.14 °F-102.58 °F); mean temporal artery temperature was 100.36°F (37.98 °C) (95% CI, 100.08 °F-100.65 °F), and mean difference between the two was 1.99 °F (1.11 °C) (95% CI, 1.75 °F-2.23 °F). CONCLUSIONS: Temporal artery thermometry is poorly sensitive detecting fever and does not accurately reflect rectal temperature. Temporal artery thermometry should not be used for clinical management of children younger than 36 months if detection of fever is of importance.