To determine the inter-rater reliability and validity of the Netherlands Triage Standard (NTS) for paediatric triage.
A cross-sectional study using fictional cases for telephone and physical triage.
An expert panel established in advance the urgency of 40 cases concerning emergency help requests from non-referred children (the reference standard). These requests were presented in an online survey to triagists from three general practitioner (GP) out-of-hours practices, three ambulance dispatching centres and three hospital emergency departments. Triagists assessed all cases, using the NTS. We determined the agreement on degrees of urgency between different triagists and compared them with the reference standard. The outcome measure for inter-rater reliability was the intraclass correlation coefficient (ICC). The outcome measures for validity were the degree of agreement with the reference standard, under-triage and over-triage, and sensitivity and specificity in identifying high-urgency (U0-U2) versus low-urgency cases (U30U5).
In total, 116 triagists participated in the study (response: 86%). The ICC was 0.73 among all triagists, and was highest in the out-of-hours GP cooperatives. There was 62.3% agreement with the reference standard, 17.4% under-triage and 20.2% over-triage. Of the divergent urgencies, 77% differed by only one urgency category. The sensitivity was 85.2% and the specificity 89.7%. The sensitivity and specificity of triage by the GP out-of-hours practices (82.7% and 92.7%, respectively) were almost the same as that by the hospital emergency departments (79.6% and 92.5%, respectively). Triage by the ambulance dispatching centres had relatively high sensitivity (93%), but relatively low specificity (82.4%).
The results of the study contribute to the evidence that the NTS is a reliable and valid triage standard for paediatric patients. The urgency assessments by triagists in the GP out-of-hours practices, ambulance dispatching centres and hospital emergency departments were broadly in agreement. Results were limited by cases being on paper and triage only on anamnestic characteristics.
Conflict of interest and financial support: potential conflicts of interest have been reported for this article. ICMJE forms provided by the authors are available online along with the full text of this article.