Abstract
OBJECTIVE: To assess the validity and reliability of a triage system for pediatric emergency care (CLARIPED) developed in Brazil. METHODS: Validity phase: prospective observational study with children aged 0 to 15 years who consecutively visited the pediatric emergency department (ED) of a tertiary hospital from July 2 to 18, 2013. We evaluated the association of urgency levels with clinical outcomes (resource utilization, ED admission rate, hospitalization rate, and ED length of stay); and compared the CLARIPED performance to a reference standard. Inter-rater reliability phase: a convenience sample of patients who visited the pediatric ED between April and July 2013 was consecutively and independently double triaged by two nurses, and the quadratic weighted kappa was estimated. RESULTS: In the validity phase, the distribution of urgency levels in 1,416 visits was the following: 0.0% red (emergency); 5.9% orange (high urgency); 40.5% yellow (urgency); 50.6% green (low urgency); and 3.0% blue (no urgency). The percentage of patients who used two or more resources decreased from the orange level to the yellow, green, and blue levels (81%, 49%, 22%, and 2%, respectively, p<0.0001), as did the ED admission rate, ED length of stay, and hospitalization rate. The sensitivity to identify patients with high urgency level was 0.89 (confidence interval of 95% [95%CI] 0.78-0.95), and the undertriage rate was 7.4%. The inter-rater reliability in 191patients classified by two nurses was substantial (kw2=0.75; 95%CI 0.74-0.79). CONCLUSIONS: The CLARIPED system showed good validity and substantial reliability for triage in a pediatric emergency department.