Does emergency medical dispatch priority predict delphi process-derived levels of prehospital intervention?

Abstract

Objective: The Medical Priority Dispatch System (MPDS) is an emergency medical dispatch system widely used to prioritize 9-1-1 calls and optimize resource allocation. This study evaluates whether the assigned priority predicts a Delphi process-derived level of prehospital intervention in each emergency medical dispatch category.

Methods: All patients given a MPDS priority in a suburban California county from 2004-2006 were included. A Delphi process of emergency medical services (EMS) professionals in another system developed the following categories of prehospital treatment representing increasing acuity, which were adapted for this study: advanced life support (ALS) intervention, ALS-Stat, and ALS-Critical. The sensitivities and specificities of MPDS priority for level of prehospital intervention were determined for each MPDS category. Likelihood ratios of low and high priority dispatch codes for the level of prehospital intervention also were calculated for each MPDS category.

Results: A total of 65,268 patients met inclusion criteria, representing 61% of EMS calls during the study period. The overall sensitivities of high-priority dispatch codes for ALS, ALS-Stat, and ALS-Critical interventions were 83% (95% confidence interval 83-84%), 83% (82-84%), and 94% (92-96%). Overall specificities were: ALS, 32% (31-32%); ALS-Stat, 31% (30-31%); and ALS-Critical 28% (28-29%). Compared to calls assigned to a low priority, calls with high-priority dispatch codes were more likely to receive ALS interventions by 22%, ALS-Stat by 20%, and ALS-Critical by 32%. A low priority dispatch code decreased the likelihood of ALS interventions by 48%, ALS-Stat by 45%, and ALS-Critical by 80%. Among high-priority dispatch codes, the rates of interventions were: ALS 26%, ALS-Stat 22%, and ALS-Critical 1.5%, all of which were significantly greater than low-priority calls (p<0.05) [ALS 13%, ALS-Stat 11%, and ALS-Critical 0.2%]. Major MPDS were categories with high sensitivities (>95%) for ALS interventions included breathing problems, cardiac or respiratory arrest/death, chest pain, stroke, and unconscious/fainting; these categories had an average specificity of 3%. Medical Priority Dispatch System categories such as back pain, unknown problem, and traumatic injury had sensitivities for ALS interventions<15%.

Conclusions: The MPDS is moderately sensitive for the Delphi process derived ALS, ALS-Stat, and ALS-Critical intervention levels, but nonspecific. A low MPDS priority is predictive of no prehospital intervention. A high priority, however, is of little predictive value for ALS, ALS-Stat, or ALS-Critical interventions.