Background: Monitoring patients’ temperatures is an important aspect of clinical nursing. Oral temperature readings have long been considered as a gold standard. However, oral temperature measurements may be contraindicated in some patients, that is cases with level of consciousness and seizure. Axiliary and tympanic temperature monitoring are often the next choices. We decided to evaluate these methods compared with standard oral temperature readings.

Materials and Methods: Subjects were a convenient sample of 81 adult patients over 17 years old in Shahrood Imam Hussein hospital. The study time course spanned a 1-mounth in December, 2004. Body temperature is measured in the evening using mercury oral and axiliary thermometers and “Omeron” tympanic thermometer.

Results: From 81 patients studied, there were 35(43.2%) males and 46(56.8%) females and their mean age was 46.0 ±19.2 (range 17-85). The mean for oral temperature was 36.9±0.56°C, axiliary: 36.5±0.53 °C, tympanic: 36.5±0.49° C. Oral thermometry identified a fever in 8 of 81 (9.9%) patients. The ROC analysis for determining a cut point with 100% sensitivity is used. For axiliary measurement at 36.95 °C, 82% specificity and for tympanic measurement at 36.45 °C, 55% specificity is recorded. Alpha coronbach for reliability of tympanic thermometer is equal to 0.98. The correlation between right to right and left to right ear temperature were 0.96 and 0.94, respectively. Conclusion: We conclude that axiliary thermometer with carefully use and suitable cut point is an effective instrument. With regard to specificity of tympanic thermometer for detection of fever, we do not recommend electronic tympanic measurement.