Central Nervous System Manifestations of Childhood Shigellosis: Prevalence, Risk Factors, and Outcome

Alterations in consciousness, including seizures, delirium, and coma, are known to occur during Shigella infection. Previous reports have suggested that febrile convulsions and altered consciousness are more common during shigellosis than with other childhood infections. Those reports, however, have been from locations where S dysenteriae type 1 was not common, thus making it difficult to assess the specific contribution that S dysenteriae type 1 infection, and Shiga toxin, might make to the pathogenesis of altered consciousness in children with shigellosis. In this study we seek to determine the prevalence, risk factors, and outcome of altered consciousness in children with shigellosis in Bangladesh, a country where infection with all four species of Shigella is common. We particularly focus on the importance of metabolic abnormalities, which we have previously shown to be a common feature of shigellosis in this population. This study was conducted at the Diarrhea Treatment Centre of the International Centre for Diarrhoeal Disease Research, Bangladesh in Dhaka, Bangladesh, which provides care free of charge to persons with diarrhea. During 1 year, a study physician identified all inpatients infected with Shigella by checking the logs of the Clinical Microbiology Laboratory daily. Study physicians obtained demographic and historical information by reviewing the patient charts and by interviewing patients, or their parents or guardians, to confirm or complete the history of illness obtained on admission. Patients were categorized as being conscious or unconscious based on a clinical scale; having a seizure documented in the hospital; or having a seizure by history during the current illness that was not witnessed by medical personnel. Patient outcome was classified as discharged improved, discharged against medical advice, transferred to another health facility, or died in the Treatment Centre. Laboratory examinations were ordered at the discretion of the attending physician; all such information was recorded on the study form. Clinical management was by the attending physician. Factors independently predictive of a documented seizure, or of unconsciousness, were determined using a multiple logistic regression analysis. For this analysis variables associated with unconsciousness or a documented seizure in the analysis of variance or chi2 analyses were entered into the regression equation and eliminated in a backward stepwise fashion if the probability assoc