Medical myth: Measuring white blood cells in the stools is useful in the management of acute diarrhea.

Abstract

It is routine clinical practice, endorsed by the American College of Gastroenterology,1 that the fecal white blood cell count should be used to determine if an acute episode of diarrhea is invasive. This test is also used to direct further investigations and therapy. This practice assumes that the fecal white blood cell count predicts that the stool culture will be positive for invasive disease and will, therefore, predict the patient's response to antimicrobial therapy. The results of multiple studies note that this is a poor test on both counts. The use of this test dates to the beginning of this century, when clinicians noted that patients with bacillary dysentery were much more likely to have white blood cells in their stool specimens than patients with amebic dysentery. Only the presence of Shigella species and Entamoeba histolytica were reported in these historic works.2 In a systematic review and meta-analysis of the literature from 1970 to 1994, Huicho et al3 determined the accuracy of stool white blood cell counts for predicting the results of stool cultures. Only studies with enough information to determine sensitivity and specificity of the test were included in the meta-analysis. Using a receiver operating characteristic curve, a peak sensitivity of 70% for the test was noted, but the specificity was just 50%. Using a cutoff to maximize specificity (90%) resulted in a sensitivity of about 40%. The test depends heavily on the time taken for the specimen to be examined after it has been retrieved and also on the experience of the operator. Increasing time results in increasing cell lysis and decreased sensitivity. The use of fecal white blood cell counts to predict the results of the stool culture is at best a secondary consideration. The most important question is, “Does the test result predict response to antimicrobial therapy?” Several articles have noted that neither the results of stool culture nor the fecal white blood cell count predict the response to therapy for either domestic4,5 or imported diarrhea.6 Because the presence of white blood cells in stool specimens is a poor predictor of culture and response to therapy, its routine use should be abandoned in favor of a more cost-effective and rational clinical algorithm.