Performance of the Dutch Triage standard in managing fever in children in out-of-hours primary care: a secondary analysis of the chili study.

Abstract

The Dutch Triage Standard (NTS) supports nurses in assessing case urgency in out-of-hours general practice (GP) care; however, its suitability for fever-related pediatric cases remains debated. This study examines the agreement between pre-consultation NTS urgency classifications (U-scores) and post-consultation management decisions in febrile children, and whether the NTS reduces between-center variation in decision-making. We performed a secondary analysis of the CHILI study, including 22 089 consultations for children under 12 with fever, across 18 out-of-hours GP centers (2015- 2016). Adjusted logistic regression models, with natural cubic splines and random intercepts, explored the associations between U-scores and three outcomes: referral to secondary care, antibiotic prescription, and other medication use. Intraclass correlation coefficients assessed between-center variability. Unadjusted and adjusted percentages of the three outcomes showed similar trends. 24.5% of consultations were classified at low urgency levels (U4-5) before the consultation. Adjusted referral probabilities decreased with lower urgency: 29% (U1) to 0.8% (U5). Strikingly, antibiotic prescription rates were highest for U3 (26.2%) and U4 (25.1%). The probability of other medication prescriptions decreased steadily across urgency levels. ICCs revealed low between-center variability, which did not decrease after accounting for U-scores. Higher NTS urgency classifications are associated with increased referral rates during fever-related out-of-hours GP consultations. However, overall referral rates remain low, while a substantial number of consultations are pre-classified by the NTS as low urgency, contributing to a high workload of unnecessary consultations that highlights the need for more efficient triage and resource allocation.