An index of severity for acute pediatric illness.

Publisher Website
Google Scholar
Abstract
A study population of 1,106 patients presenting to a pediatric emergency room for care of acute non-traumatic illness were prospectively assessed to determine what clinical manifestations and/or functional status indicators predicted the severity of their illnesses. From these data, a "Severity Index" was developed using a 0-1-2 point score for the five most significant predictors to give a maximum score of 10 points. The predictor variables include respiratory effort, skin color, activity, temperature, and ability to play. The index has a predictive accuracy for non-severe illness of 98.7 per cent with only 1.3 per cent false negative predictions. In predicting major illness, a false positive prediction rate of 15.8 per cent was obtained. The index appears valid both across and within diagnostic categories.