Association between ambulance dispatch priority and patient condition

Objective To compare chief complaints of the Medical Priority Dispatch System in terms of the match between dispatch priority and patient condition. Methods This was a retrospective whole‐of‐population study of emergency ambulance dispatch in Perth, Western Australia, 1 January 2014 to 30 June 2015. Dispatch priority was categorised as either Priority 1 (high priority), or Priority 2 or 3. Patient condition was categorised as time‐critical for patient(s) transported as Priority 1 to hospital or who died (and resuscitation was attempted by paramedics); else, patient condition was categorised as less time‐critical. The χ 2 statistic was used to compare chief complaints by false omission rate (percentage of Priority 2 or 3 dispatches that were time‐critical) and positive predictive value (percentage of Priority 1 dispatches that were time‐critical). We also reported sensitivity and specificity. Results There were 211 473 cases of dispatch. Of 99 988 cases with Priority 2 or 3 dispatch, 467 (0.5%) were time‐critical. Convulsions/seizures and breathing problems were highlighted as having more false negatives (time‐critical despite Priority 2 or 3 dispatch) than expected from the overall false omission rate. Of 111 485 cases with Priority 1 dispatch, 6520 (5.8%) were time‐critical. Our analysis highlighted chest pain, heart problems/automatic implanted cardiac defibrillator, unknown problem/collapse, and headache as having fewer true positives (time‐critical and Priority 1 dispatch) than expected from the overall positive predictive value. Conclusion Scope for reducing under‐triage and over‐triage of ambulance dispatch varies between chief complaints of the Medical Priority Dispatch System. The highlighted chief complaints should be considered for future research into improving ambulance dispatch system performance.