AN EVALUATION OF TYMPANIC THERMOMETRY IN AN OBSTETRIC POPULATION

Objective: In an obstetric population, we set out to establish normal temperature ranges and reproducibility for core temperatures measured by the tympanic thermometer, compare these core temperatures with oral electronic temperatures, acid determine the accuracy of the tympanic thermometer in detecting fever. Methods: Simultaneous tympanic core te mperatures and oral electronic temperatures were measured in obstetric inpatients and outpatients(n = 73) between the gestational ages of 24 weeks and 4 days postpartum. Repeat measurements were taken 10 min la ter. For the detection of fever, linear regression analysis was used t o calculate the tympanic temperature that correlated with 38.0 degrees C oral. Results: Tympanic core temperatures were significantly higher than oral temperatures (P < 0.00001), with means for the two groups o f 37.76 and 35.95 degrees C, respectively. Both thermometers showed eq uivalent reproducibility between the initial and repeat measurements, with Pearson's correlation coefficients of 0.98 for tympanic temperatu res and 0.97 for oral temperatures. Linear regression analysis reveale d that 38.0 degrees C oral correlated with 38.8 degrees C tympanic. Th us, for the detection of fever with the tympanic thermometer, we found a sensitivity of 91.7%, a specificity of 98.4%, a positive predictive value of 91.7%, and a negative predictive value of 98.4%. Conclusions : We conclude that: (1) temperature ranges, accuracy, and reproducibil ity for the tympanic thermometer in a population of obstetric patients are demonstrated. (2) The mean tympanic temperature was significantly higher, by an average of 0.81 degrees C, than the mean oral temperatu re. (3) Acceptable sensitivity, specificity, and positive and negative predictive values for the detection of fever were obtained with the t ympanic thermometer. (4) The level of diagnosing fever for tympanic te mperatures in obstetric patients is shown to be 38.8 degrees C. Confir mation of this difference from the commonly accepted standard of 38.6 degrees C tympanic for nonpregnant patients awaits larger studies.