Interobserver concordance in simulated vignettes using the Manchester Triage System in emergency departments in Asturias.

OBJECTIVE

To assess the inter-observer concordance of the Manchester Triage System (MTS) through clinical vignettes among emergency nurses in Asturias, Spain.

METHOD: Prospective multicenter study in seven emergency departments using questionnaires containing 40 simulated triage cases based on real cases. Interobserver agreement was calculated, as well as the level of accuracy (correctness) and the level of inaccuracy in triage (over- and under-triage).

RESULTS

A total of 100 nurses participated in the study (median age: 39.1 years). Substantial inter-observer agreement was observed (κ = 0.681; 95% CI 0.614-0.759; P < .001). The median number of correctly triaged vignettes was 30 (SD = 3.93). A positive correlation was found between accuracy and MTS experience was r = 0.250; P = .012. Among the 4000 vignettes evaluated, 403 (10.1%) were undertriage, while overtriage occurred in 253 vignettes (6.3%). Sensitivity was 83.7% and specificity was 94.8%, with undertriage (10%) was more frequent than overtriage (6%).

CONCLUSION

The study demonstrated substantial inter-observer agreement among emergency nurses in seven hospitals using the MTS, with an overall accuracy of 80%. Triage accuracy has a slight correlation with increased years of experience.