Performance of the Manchester Triage System in patients with dyspnoea: A retrospective observational study

Background: Between 1% and 7.5% of patients admitted to the emergency department (ED) suffer from dyspnoea. The Manchester Triage System is one of Europe's most used triage systems considering five levels to prioritize patients in the ED: level 1 (red), immediate; level 2 (orange), very urgent; level 3 (yellow), urgent; level 4 (green), standard; level 5 (blue), non-urgent. With this study we aimed to evaluate the performance of the Manchester Triage System in patients with dyspnoea. Methods: With this retrospective, observational study we analysed data from 4076 patients with dyspnoea accessing the ED in one Italian hospital between January 1, 2017 and June 30, 2019. To determine the sensitivity and specificity we computed the area under the curve (AUC) of the receiver operating characteristics (ROC) comparing the sensitivity and specificity for each of the five priority codes and seven-day mortality. Results: A total of 3.6% (n = 145) of the patients died within seven days after access to the ED. From the patients dying within seven days, 6.2% (9/145) had a blue or green priority code, 17.9% (26/145) had a yellow priority code and 75.9% had an orange or red priority code. A high priority code (orange or red) had a sensitivity of 75.9%, a specificity of 66.1%, a PPV of 7.6% and a NPV of 98.7%. The AUC was 0.734 (CI 95% 0.695-0.773, p < 0.001). Conclusions: The Manchester Triage System showed acceptable sensitivity and negative predictive value in patients with dyspnoea, yet a low specificity related to the risk of death. The system is safe, resulting in a low under-triage rate for death. Improving the clinical assessment during nurses' triage (e.g. considering clinical history and auscultation of lungs) and prospective validation studies with surrogate endpoints is needed in this population.