Triage of the pediatric patient in the emergency department: are we all in agreement?

OBJECTIVE: To compare triage categorization as a measure of perceived patient acuity on presentation to the emergency department by pediatric emergency medicine (PEM) attending physicians, nurses, and pediatric residents with their general emergency medicine (GEM) counterparts.

METHODS: A questionnaire that contained 12 pediatric triage scenarios was sent to all PEM attending physicians, triage-trained nurses, and pediatric residents and their GEM counterparts at a large urban hospital with separate pediatric and general emergency departments. Participants were asked to use a 3-tier triage system (emergent, urgent, nonurgent) to assign a triage level for each patient scenario.

RESULTS: The response rate was 99%. The kappa level of agreement was highest (.39) among the PEM physicians. Significantly more GEM attending physicians triaged the following scenarios at a higher acuity level as compared with PEM attending physicians with a trend toward emergent triage: simple febrile seizure, 50% (95% confidence interval [CI]: 30%-70%) versus 7.7% (95% CI: 1%-34%); 18-month-old with fever and bumps on lips, 21% (95% CI: 9%-43%) versus 0% (95% CI: 0%-23%); and 15-month-old well-appearing child with high fever, 50% (95% CI: 30%-70%) versus 7.7% (95% CI: 1%-34%). Significant differences were found between GEM and PEM triage-trained nurses only in the 15-month-old high fever scenario and between GEM and pediatric residents in the 15-month-old high fever scenario, the 18-month-old with fever and bumps on lips scenario, and a fever/limp scenario.

CONCLUSIONS: The level of agreement of triage assignment within each group was only fair. GEM participants and PEM participants agreed on most scenarios. However, GEM participants were more likely to triage children with certain febrile illnesses at higher acuity levels as compared with their PEM counterparts.