This prospective study assessed the value of presenting history, physical examination, and screening laboratory tests in predicting whether diarrhea in a young child is associated with a stool culture positive for a bacterial pathogen. Acutely ill children less than 4 years old were studied in a hospital outpatient setting. Two hundred patients were seen in a 9?-month period, which encompassed the seasons of summer, fall, and winter. One hundred ninety-five patients had cultures completed and twenty-nine (15%) had a bacterial pathogen isolated. The best predictive variable for a stool culture positive for a bacterial pathogen was the presence of polymorphonuclear cells in the stool, with a sensitivity of 85%, a specificity of 88%, and positive and negative predictive values of 59% and 97%, respectively. A cluster of three historical variables—abrupt onset of diarrhea, greater than four stools per day, and no vomiting before the onset of diarrhea—was identified that delineated a subpopulation of patients with an increased probability of having a stool culture positive for a bacterial pathogen (27% v 4% if any of the three variables was absent). It is suggested that these findings can be combined in a stepwise manner using the historical cluster as an initial screening, followed by examination for stool polymorphonuclear cells in the high probability subgroup, to identify those patients with a very high probability of having a bacterial pathogen isolated in their stool.
展开▼
