Objective: To determine the simplest combination of symptoms and signs for use as a triaging procedure in young infants presenting to hospital, and to compare this with the 'Baby Check' system.
Methodology: Six hundred and eighty-two babies were assessed when presenting with acute illness to a children's hospital emergency department. Results of analysis were compared to the results of the Baby Check system, which was derived from the data on the same babies.
Results: The presence of at least one of the following markers: drowsiness on examination, significant chest wall recession, generalized pallor, a history of feeding less than 50% or decreased activity, had a sensitivity of 91%, a specificity of 72%, a positive predictive value of 29% and a negative predictive value of 98% when predicting the need for intervention in hospital. The 'Baby Check' system with a score of 13 or more had a sensitivity of 95%, a specificity of 70%, a positive predictive value of 28% and a negative predictive value of 99%. Less common but predictive markers such as bile-stained vomiting, respiratory grunt, apnoea and convulsions add further to the sensitivity. Based on this data set, if historical variables alone (the presence of either drowsiness, difficult breathing, being more pale than usual, feeding less than 50% or decreased activity) were used (as in telephone screening) 92% of babies needing treatment in hospital would be identified.
Conclusion: A simple five-marker system is almost as powerful as the 19-marker Baby Check system and provides a useful basis for a triaging system and educational guidelines when assessing sick babies.