Objective: To use weather data to predict the daily number of ambulance transports required for heat stroke cases.
Methods: We analyzed weather data and the number of ambulance transports for heat stroke cases in Tokyo, Kanagawa, and Osaka between July 1 and September 30, 2013. Weather data included daily temperature (average, maximum, and minimum), humidity, hours of daylight, wind velocity, and precipitation. The correlation between each parameter and the number of ambulance transports was analyzed, and we developed a prediction formula using these data.
Results: The daily average and maximum temperatures were strongly correlated with the number of transports for heat stroke (both, r = 0.73). Therefore, we created a formula using the daily average temperatures in Tokyo, and verified this formula in Tokyo, Osaka, and Kanagawa:
yi ~
f (
Tav,i) ≡ a exp (b
Tav,i) + c = 0.3800 exp (0.00007
Tav,i)
This formula provided accurate predictions for August and September, although it underestimated the number of transports that occurred in July. Therefore, we created an adjusted formula that provided accurate predictions for July:
yi ~
f (
Tav,i) + Δ
f (
xi) ≡ (a exp (b
Tav,i) + c) + (
α T*,i +
β)
={0.3800 exp (0.00007
Tav,i) + 1.209
Thigh,i − 33.47 (
Thigh)
={0.3800 exp (0.00007
Tav,i) + 1.416
Tlow,i − 28.59 (
Tlow)
Conclusion: Accurately predicting the number of ambulance transports for heat stroke can facilitate better allocation of ambulance resources. Our revised formula provided accurate predictions in all 3 months and regions that we examined.
View full abstract