Reproducibility of the Manchester Triage System: a multicentre vignette study.

Abstract

Background: While several studies have evaluated the performance of the Manchester Triage System (MTS), none have specifically examined its accurate application by triage nurses and its association with clinical outcomes. This study focuses on the agreement between nurse-assigned MTS codes and those assigned by an expert group, as well as their ability to predict clinical outcomes.

Methods: This multicentre simulation study was conducted from January to March 2024 across four EDs in Italy employing MTS in clinical practice. Two emergency physicians developed 30 vignettes derived from real clinical cases to encompass diverse triage scenarios and priority codes. An expert MTS group, composed of three experienced nurses, assigned MTS priority codes following the guidelines outlined in the official MTS textbook. Subsequently, the vignettes were presented to triage nurses, who independently assigned MTS codes. Error rate, agreement between nurse-assigned and expert MTS group codes, and the predictive ability for secondary clinical outcomes (mortality within 72 hours, hospitalisation, life-saving intervention, severe condition in the ED and time-dependent pathology) were compared between the MTS priority assigned by the expert MTS group codes and nurse-assigned triage codes.

Results: 77 nurses from four EDs participated. The triage code assignment error rate was 28.6% (660/2310). The overall agreement between the triage and expert nurses yielded a Cohen's kappa of 0.59 (95% CI 0.58 to 0.59). Expert MTS group applications performed better compared with nurse-assigned codes in predicting clinical outcomes. The mean error rate per nurse was 30% (9/30). Nurses with more ED experience and triage expertise had higher error rates.

Conclusion: The application of MTS using case vignettes was suboptimal in our setting, with more senior nurses having higher error rates. Correct application of MTS better predicted clinical outcomes. It is important to conduct future studies to understand how to best support nursing clinical decision-making in triage.