Postoperative febrile morbidity in the benign gynecologic patient. Identification and management

To identify a useful temperature threshold and occurrence frequency that might represent genuine morbidity and to identify elements of the fever evaluation that were most useful in discriminating ill patients from those with transient and inconsequential temperature elevations.

A retrospective chart review of postoperative gynecologic patients was undertaken for a one-year period. Data regarding temperatures, frequencies and workup were collected, as were descriptive data regarding the patient and case.

Some of the previously reported associations, such as menopausal status and insurance coverage, failed to materialize as associated with temperature increases in this study. Other associations, such as blood loss and length of surgery, were confirmed. The commonly used threshold for "fever" of 38 degrees C was not as discriminatory as a threshold of 38.3 degrees C. Frequency, as well as absolute degree of elevation, did appear to relate to level of morbidity. The most useful studies in assessing temperature elevations were the white blood cell count and chest radiography.

Although prospective evaluation is difficult and much work remains to be done, our data are sufficiently compelling to suggest a minimum temperature and frequency at which the otherwise-well-appearing postoperative patient should be assessed for infection. We also found the yield sufficient to make routine ordering useful only in the case of the white blood cell count and possibly the chest radiograph. Other tests should be reserved for use if directed by the physical examination or patient history.