Accuracy and precision of a novel non-invasive core thermometer

Abstract

Background: Accurate measurement of core temperature is a standard component of perioperative and intensive care patient management. However, core temperature measurements are difficult to obtain in awake patients. A new non-invasive thermometer has been developed, combining two sensors separated by a known thermal resistance ('double-sensor' thermometer). We thus evaluated the accuracy of the double-sensor thermometer compared with a distal oesophageal thermometer to determine if the double-sensor thermometer is a suitable substitute.

Methods: In perioperative and intensive care patient populations (n=68 total), double-sensor measurements were compared with measurements from a distal oesophageal thermometer using Bland-Altman analysis and Lin's concordance correlation coefficient (CCC).

Results: Overall, 1287 measurement pairs were obtained at 5 min intervals. Ninety-eight per cent of all double-sensor values were within +/-0.5 degrees C of oesophageal temperature. The mean bias between the methods was -0.08 degrees C; the limits of agreement were -0.66 degrees C to 0.50 degrees C. Sensitivity and specificity for detection of fever were 0.86 and 0.97, respectively. Sensitivity and specificity for detection of hypothermia were 0.77 and 0.93, respectively. Lin's CCC was 0.93.

Conclusions: The new double-sensor thermometer is sufficiently accurate to be considered an alternative to distal oesophageal core temperature measurement, and may be particularly useful in patients undergoing regional anaesthesia.