Background Triage is an important method to manage patient flows safely at the Emergency Department (ED), by prioritising patients according to their clinical urgency. To date, no studies have assessed the safety of the Manchester Triage System (MTS) to identify critically ill children.

Objective To determine how well the MTS identifies critically ill children as reflected by ICU admission, and to identify areas for improvement.

Methods In this observational study, we collected routinely available electronic data of all children <16 years visiting the ED in a Dutch university hospital between 2006 and 2012. Under triage was defined as patients triaged to low MTS urgencies (level 3 to 5, indicating 60, 120 and 240 min maximum waiting time) who were admitted to ICU. Patients with low MTS urgencies and ICU-admission were compared with low urgency patients who were not admitted to ICU, to identify risk factors for under triage.

Results During the 7-year study period, data on 44 476 ED-visits were collected of which 837 (1,9%) were from patients admitted to ICU. Of all ICU admissions, 251 (28.3%) were assigned to low MTS triage categories. Risk factors for under triage were young age, non-surgical presenting problem and general illness, gastrointestinal or respiratory complaints (all p < 0.001). Moreover, under triaged children were more often referred by a medical specialist, possibly indicating that co morbidity is a risk factor too.

Conclusion The MTS under triages an impressive amount of paediatric patients. Adaptations of the MTS based on risk factors for under triage are needed to improve the safety of the MTS.