No abstract

Electronic oral thermometry is performed routinely in most medical centres. From the studies available on this subject it seems difficult to find any documentation for this practice. We have conducted clinically controlled studies in which the accuracy of electronic oral thermometry (CRAFTEMP and TERUMO WCT) and that of electronic rectal thermometry (TERUMO WCT) were tested. Rectal glass mercury thermometry was used as a reference method. Two studies were designed. In study 1, 184 patients (72 women, 112 men), median age 70 (18-95) years were investigated. In study 2, 91 patients (41 women, 50 men), median age 59 (18-96) years were investigated. Electronic oral thermometry was found unacceptably inaccurate under daily routine conditions. Electronic rectal thermometry was found to be accurate. Calculations of mean temperature difference between reference measurements and test measurements (mean +/- SD) were found to be: routine oral CRAFTEMP at 0.70 +/- 0.50 degrees C, optimum oral TERUMO WCT at 0.75 +/- 0.74 degrees C, routine rectal TERUMO WCT at 0.08 +/- 0.26 degrees C and optimum rectal TERUMO WCT at 0.02 +/- 0.17. In the screening procedure for fever oral thermometry showed low sensitivity (routine CTAFTEMP 0.47 and optimum TERUMO WCT 0.59) whereas rectal thermometry showed high sensitivity (routine TERUMO WCT 0.74 and optimum TERUMO WCT 0.91). It was concluded that rectal thermometry must be preferred to oral thermometry for daily routine measurements.