Abstract
Diarrhoeal stools from 11358 patients were microscopically examined for the presence of white and red blood cells (WBC and RBC)·and cultured for enteric pathogens. Of the 3895 patients who had Shigella as sole pathogens, 2681 (72·3%) showed the presence of both WBC and RBC and the remainder had only WBC ( P < 0·001). The presence of both WBC and RBC was as good a predictor of shigellosis as the presence of >25 WBC per high power field (hpf) of the microscope with or without RBC. However, the best predictor of shigellosis (positive predictive value 85%, negative predictive value 83%) was the presence of >25 WBC/hpf and the presence of RBC regardless of their number. More patients infected with S. dysenteriae 1, S. flexneri and S. boydii shed both WBC and RBC than those infected with S. sonnei , most of whom shed WBC only ( P < 0·001). A greater number of patients infected with S. dysenteriae 1 shed more WBC and RBC than those infected with S. flexneri ( P < 0·001). The same trend was found when patients infected with S. flexneri were compared with those infected with S. boydii (though the difference was not statistically significant) and when patients infected with S. boydii were compared with patients infected with S. sonnei ( P < 0·001). There was a progressive decline in the isolation rate of Shigella as the duration of dysentery, before stool culture, increased.