A comparison of pulmonary artery, rectal, and tympanic membrane temperature measurement in the ICU.

OBJECTIVE

To compare tympanic membrane temperature with pulmonary artery (PA) and rectal temperature (calibrated glass mercury) to determine consistency among measures.

DESIGN

Convenience, within-subject, quasi-experimental.

SETTING

Midwestern university-affiliated tertiary medical center.

PATIENTS

128 adult (18 years or older) patients admitted to the surgical intensive care unit with an age range of 18 to 90 years (mean 57 years).

OUTCOME MEASURES

Tympanic, PA, and rectal temperature.

INTERVENTION

Rectal and tympanic membrane temperatures were measured in 60 patients. PA and tympanic membrane temperatures were measured in 68 patients.

RESULTS

Rectal and tympanic membrane temperatures were moderately correlated (r = 0.525). Mean rectal was slightly higher than mean tympanic membrane temperature (mean difference 0.19 degrees C). PA and tympanic membrane temperatures were highly correlated (r = 0.909). Mean tympanic membrane temperature was slightly higher than mean PA temperature (mean difference 0.42 degrees C). For PA temperatures, 57 (84%) of the differences were between 0 and +1 degree Celsius, whereas for rectal temperature, 23 (37%) were between 0 and +1 degree Celsius and 28 (47%) were between 0 and -1 degree Celsius.

CONCLUSIONS

Tympanic temperature measurements would be an appropriate substitute for PA temperature if the PA catheter has been removed or the PA thermistor is nonfunctional. In addition, this study suggests that to effectively track temperature in a particular patient, the site for temperature measurement must be consistent.