1678. Assessing Performance of Multiple Methods for Measurement of Body Temperature, Bangladesh

Abstract Background CDC and icddr,b are conducting an acute febrile illness (AFI) research project in four hospitals in Bangladesh. Enrolled subjects have measured fever of ≥100.4°F. To determine the most-sensitive temperature measurement method, we collected multiple measurements on patients with fever history between March and April 2019. Methods Patients were screened in outpatient departments of four hospitals in Bangladesh between March 7 and April 15, 2019. Screening used at least two of three methods: tympanic, oral, or axillary. Records were consolidated using Microsoft Excel and analyzed in R3.4.3. We examined the correlation between temperatures measured by different methods for each patient. For records with all three measures, we calculated the likelihood of meeting AFI inclusion criterion of ≥100.4°F (38°C) by measurement method. Results 3,060 subjects were enrolled. The highest correlation among measurements was between axillary and oral (r = 0.882, 95% CI 0.868–0.895). The lowest correlation was between tympanic and oral (r = 0.71, 95% CI 0.69–0.73). Axillary and oral had the highest correlation in both children and adults (peds: 0.88, 95% CI 0.86–0.90; adult: 0.89, 95% CI 0.86–0.90). By site, the highest correlation was axillary to oral among children in Hospital 1 (r = 0.98, 95% CI 0.92–1.00), while the lowest was axillary to tympanic for adults in Hospital 3 (r = 0.71, 95% CI 0.65–0.77). 882 subjects (334 pediatric, 548 adult) were assessed using all three measurement methods. 313 (159 pediatric, 154 adult) met AFI inclusion criterion by at least one method. From 49% to 63% of subjects at any site met the criterion by two or three methods (table). Results in hospitals 1, 2 and 4 were similar and grouped for analysis. In every site, subjects were detected by oral who would not have been detected using axillary or tympanic. Only in Hospital 3, subjects were detected by tympanic alone. No subjects in any site met the criterion by axillary measurement alone. Conclusion Accurate measurement of body temperature is essential for AFI surveillance, but literature on methodology is limited. We demonstrate that multiple modes of measurement increased detection of febrile patients. The most sensitive combination was oral and tympanic. Axillary measurement did not improve detection. Disclosures All authors: No reported disclosures.