Correlation between the accuracy of the emergency response centre’s urgency assessment and emergency medical services non-conveyance: a retrospective register-based study in Finland

Abstract

Background: In modern emergency medical services (EMS), ambulances increasingly focus on examining and treating the patient at the scene. This has led to increased levels of non-conveyance. In Finland, for instance, approximately 40% of EMS dispatches end up in non-conveyance. As EMS systems evolve, the proportion of non-conveyance could serve as a cost-effective measure to assess the quality of the dispatch criteria, if a link to the performance of urgency assessment would be established. The purpose of this study was to investigate whether the proportion of non-conveyance is associated with the test performance levels of the urgency assessment. This investigation was done separately within each dispatch category. Methods: A retrospective evaluation of the data was conducted on all EMS dispatches in the Pirkanmaa Hospital District from 1 August 2021 through 31 August 2021. There were a total of 7,245 EMS dispatches during the study period of which 829 were excluded. This study was conducted by comparing the existing test performance levels (sensitivity, specificity and under- or overestimation) of the emergency response centre’s urgency assessment with the non-conveyance rate (%) of each dispatch category. The relationships between the variables were measured using Spearman’s rank correlation coefficient. Results: The proportion of over-triage was the only urgency assessment’s test performance variable that had a statistically significant correlation with the proportion of non-conveyance (r = 0.568; p = 0.003). Other test performance variables of the urgency assessment had no or little correlation to the proportion of non-conveyance. Of the 6,416 EMS dispatches in the study period, 42% (2,672) resulted in non-conveyance of the patient. In nine dispatch categories, at least half (51–69%) of the dispatches ended in non-conveyance. Conclusions: Based on this study, it seems that the percentage of non-conveyance in the dispatch category could be used, with certain limitations, to assess the proportion of over-triage in the dispatch category. The method is particularly applicable in scenarios where the dispatch criteria have undergone modifications and there is a need to monitor the effect of the changes on the level of over-triage.