Analysis of fecal leukocytes and erythrocytes in Shigella infections in urban Bangladesh.

We evaluated the usefulness of enumeration of fecal leukocytes and erythrocytes in making an early diagnosis of Shigella infection, where Shigella is a leading cause of invasive diarrhea. Stool specimens from 561 invasive diarrhea patients were submitted for microscopic examination. A presumptive diagnosis of shigellosis based on microscopic examination was made in 389 of them; 227 had stool cultures positive for Shigella spp (Shigella patients). One hundred sixty-two patients with no detectable Shigella infection (non-Shigella invasive diarrhea cases) served as a comparison group. Two hundred twenty-seven randomly selected Shigella patients and 227 non-Shigella infectious diarrhea cases from the surveillance system database of the hospital constituted another group for comparative evaluation. The stool specimens of the patients were examined under the microscope, and isolation, biochemical characterization and serotyping of Shigella were performed. In comparison with non-Shigella invasive diarrhea cases, the presence of >50 WBC/hpf in association with any number of RBC in the fecal sample had a modest sensitivity of 67%, specificity of 59%, positive predictive value of 70%, negative predictive value of 56%, accuracy of 64%, and positive likelihood ratio of 1.6 in predicting shigellosis. Comparison between Shigella and non-Shigella infectious diarrhea patients revealed the presence of >20 WBC/hpf was a less accurate predictor of shigellosis (sensitivity 51%, specificity 88%, positive predictive value 81%, negative predictive value 64%, accuracy 69%, and positive likelihood ratio 4.1). Direct microscopical examination of stool specimens for the presence of WBC and RBC may facilitate the early diagnosis of shigellosis, and may be a cheap alternative to stool culture in this setting.