| Title |
Response-time component decomposition and performance measurement of an emergency response system under weather-advisory conditions using nationwide EMS dispatch data |
| DOI |
https://doi.org/10.5573/ieie.2026.63.10.103 |
| Keywords |
Emergency response system; Weather advisory; Response time measurement; Response-time component decomposition; Emergency medical service |
| Abstract |
This study constructs a component-decomposition procedure to measure the time performance of the 119 emergency response system under weather-advisory conditions. We linked 16,842,543 nationwide emergency medical service dispatch records from 2021 to 2025 with heat-wave and cold-wave datasets and weather-advisory bulletins. The analysis applied station-by-year-month fixed effects, 2,000 cell-level bootstrap replications, and 200 permutations. Heat-advisory sensitivity was 95.85% across 123,930 station-days. Call-to-arrival time on advisory days increased by 5.63 s (95% CI 2.51?8.77) under equal cell weighting and by 2.73 s (95% CI 0.22?5.42) under case weighting. The increase occurred primarily in the travel component, whereas the turnout component changed by 0.9?1.7 s. Dispatches coded as hyperthermia increased by 14.33%, and only hyperthermia and general weakness increased in the symptom-specific analysis. The advisory variable explained 0.0067% of response-time variance, one-539th of that explained by municipality (3.6047%). None of the 38 component-level tests across four weather conditions showed a reduction in turnout or travel time. In addition, 70.5% of heat-related dispatches were concentrated in the top quintile of predicted station-days. The proposed procedure distinguishes the magnitude and location of time-performance changes in an emergency response system under exogenous conditions. The measured change was about 1% of the baseline response time. No response-time shortening indicative of pre-positioning was observed even on advisory days, suggesting that allocation design should consider administrative district and day-ahead identifiable high-risk station-days in addition to advisory issuance. |