Clostridium difficile infection in patients with unexplained leukocytosis

Purpose To determine whether unrecognized Clostridium difficile infection is responsible for a substantial proportion of cases of unexplained leukocytosis in a tertiary care hospital setting. Methods We prospectively identified 60 patients who had unexplained leukocytosis (white blood cell count ≥15,000/mm3). Fecal specimens were tested for C. difficile toxin using an enzyme immunosorbent assay. We compared the clinical features of patients who had positive or negative assay results, as well as of 26 hospitalized control patients who did not have unexplained leukocytosis. Results Thirty-five (58%) of the patients with unexplained leukocytosis had C. difficile toxin in at least one fecal specimen as compared with 3 (12%) of the controls (P<0.001). Symptoms of colitis were often mild or absent at the time the white blood cell count was first elevated or, if present, had not been recognized by the attending physicians. Leukocytosis resolved promptly in most patients who were treated with metronidazole. In the 25 patients (42%) who had a negative test for C. difficile toxin, leukocytosis also tended to resolve during empiric therapy with metronidazole; some of these patients may have had C. difficile infection. Conclusion The majority of patients in our hospital who had unexplained leukocytosis had C. difficile infection. Unexplained leukocytosis in hospitalized patients should prompt a search for symptoms and signs consistent with C. difficile infection and a study to detect C. difficile. Empiric therapy with metronidazole may be effective in the appropriate epidemiologic setting.