Maternal temperature monitoring during labor: Concordance and variability among monitoring sites

OBJECTIVE: Elevated maternal temperature in labor is associated with adverse immediate and long-term neonatal outcomes. Conventional methods of temperature measurement may not reflect the intrauterine temperature, which constitutes the fetal environment. ne purpose of this study was to ascertain the most reliable noninvasive method of temperature monitoring in labor that would best reflect changes in intrauterine temperature. METHODS: Women in labor receiving epidural analgesia had temperature readings taken every 10 seconds from the uterine cavity, ear canal, and skin surface of the leg and abdomen and hourly from the mouth. RESULTS: Eighteen patients were studied for a mean of 228 minutes (range 56-464 minutes). The best indicator of intrauterine temperature was oral temperature, with a mean intraclass correlation coefficient of 0.6 (95% confidence interval 0.42, 0.77). On average, oral temperature underestimated intrauterine temperature by 0.8degreesC (95% confidence interval 0.7degreesC, 1degreesC). Allowing for this, oral temperature greater than 37.2degreesC detected an intrauterine temperature greater than 38degreesC with a sensitivity of 81% and a specificity of 96%. The intraclass correlation coefficients of all other sites with intrauterine temperature were poor (0.1 or less). As expected, the temperature at all sites increased as labor progressed. CONCLUSION: Oral temperature, measured carefully, has an acceptable correlation with intrauterine temperature and is recommended for routine detection of maternal pyrexia in labor. Continuous skin and external auditory canal temperature measurements did not correlate well. (C) 2004 by The American College of Obstetricians and Gynecologists.