Objective: To determine which measures of heatwave have the greatest predictive power for increases in health service utilisation in Perth, Western Australia.
Methods:Three heatwave formulas were compared, using Poisson or zero-inflated Poisson regression, against the number of presentations to emergency departments from all causes, and the number of inpatient admissions from heat-related causes. The period from July 2006 to June 2013 was included. A series of standardised thresholds were calculated to allow comparison between formulas, in the absence of a gold standard definition of heatwaves.Results: Of the three heatwave formulas, Excess Heat Factor (EHF) produced the most clear dose-response relationship with Emergency Department presentations. The EHF generally predicted periods that resulted in a similar or higher rate of health service utilisation, as compared to the two other formulas, for the thresholds examined.
Conclusions:The EHF formula, which considers a period of acclimatisation as well as the maximum and minimum temperature, best predicted periods of greatest health service demand. The strength of the dose-response relationship reinforces the validity of the measure as a predictor of hazardous heatwave intensity.
Implications:The findings suggest that the EHF formula is well suited for use as a means of activating heatwave plans and identifies the required level of response to extreme heatwave events as well as moderate heatwave events that produce excess health service demand.