THE ETIOLOGY AND MANAGEMENT OF DIARRHEA IN THE GYNECOLOGIC ONCOLOGY PATIENT

Purpose: We determined the etiology and evaluated the usefulness of diagnostic tests in 42 gynecologic oncology patients with diarrhea. Methods: All inpatients who developed diarrhea during treatment for gynecologic cancer were prospectively evaluated during the 12-month study period ending June 1992. Diarrhea was defined as five or more loose stools per 24 hr lasting for 48 hr or more. Diagnostic tests were complete blood cell count; electrolytes; stool for occult blood, white blood cells, and Clostridia difficile toxin; and stool culture for Shigella, Salmonella, Yersinia, and Campylobacter. Results: During the study period 351 women with gynecologic cancer received inpatient care. Forty-two (12%) patients developed diarrhea. One patient had three episodes and another had two episodes of diarrhea. There were 39 evaluable patients; 3 patients were excluded due to incomplete tests. The only cause of bacterial-associated diarrhea was C. difficile, which occurred in 4 of 39 (10.3%) patients. No patient had community-acquired diarrhea, and 1 patient had diarrhea due to Crohn's Disease. Surgery was performed on 265 patients; 29 of 265 (10.9%) developed postoperative diarrhea. Eighty-six patients had medical treatment and 13 (15.1%) developed diarrhea. The distribution by admission status of patients with diarrhea did not differ statistically from the patients without diarrhea. The four patients with C. difficile-associated diarrhea received appropriate therapy at the time of diagnosis. Nevertheless, the mean (SD) duration of C. difficile-associated diarrhea (342 (103) hr) was significantly longer than the mean (SD) duration of diarrhea not associated with this pathogen (79 (32) hr). The difference between the sample means is 263 hr with a 95% confidence interval for the difference between the means of 218 and 308 hr. Electrolyte abnormalities occurred in 20 patients (51%). Fecal leukocytes were present in 11 (28%) patients and occult blood in 5 (12.8%) of the 39 patients. Neither study was predictive for the presence of C. difficile. Conclusion: Hospital-acquired diarrhea in gynecologic oncology patients is common, although specific bacterial pathogens requiring antibiotic therapy account for only 10% of cases. Diagnostic tests for infectious causes besides C. difficile are infrequently useful and may be omitted unless the condition persists more than 72 hr. © 1993 Academic Press, Inc.