School of Public Health, University of Queensland, Brisbane, QLD, Australia.
Queensland Alliance for Environmental Health Sciences, University of Queensland, 266 Herston Rd, 4006, Herston, QLD, Australia.
Int J Biometeorol. 2024 Aug;68(8):1637-1647. doi: 10.1007/s00484-024-02691-9. Epub 2024 May 6.
Extreme heat alerts are the most common form of weather forecasting services used in Australia, yet very limited studies have documented their effectiveness in improving health outcomes. This study aimed to examine the temporal changes in temperature-related mortality in relation to the activation of the heat-health alert and response system (HARS) in the State of Victoria, Australia. We examined the relationship between temperatures and mortality using quasi-Poisson regression and the distributed lag non-linear model (dlnm) and compared the temperature-mortality association between the two periods: period 1- prior-HARS (1992-2009) and period 2- post-HARS (2010-2019). Since the HARS heavily weights heatwave effects, we also compared the main effects of heatwave events between the two periods. The heatwaves were defined for three levels, including 3 consecutive days at 97th, 98th, and 99th percentiles. We also controlled the potential confounding effect of seasonality by including a natural cubic B-spline of the day of the year with equally spaced knots and 8 degrees of freedom per year. The exposure-response curve reveals the temperature mortality was reduced in period 2 in comparison with period 1. The relative risk ratios (RRR) of Period 2 over Period 1 were all less than one and gradually decreased from 0.86 (95% CI, 0.72-1.03) to 0.64 (95% CI, 0.33-1.22), and the differences in attributable risk percent increased from 13.2 to 25.3%. The reduction in the risk of heatwave-related deaths decreased by 3.4% (RR 1.068, 95% CI, 1.024-1.112 versus RR 1.034, 95% CI, 0.986-1.082) and 10% (RR 1.16, 95% CI, 1.10-1.22 versus RR 1.06, 95% CI, 1.002-1.119) for all groups of people. The study indicated a decrease in heat-related mortality following the operation of HARS in Victoria under extreme heat and high-intensity heatwaves conditions. Further studies could investigate the extent of changes in mortality among populations of differing socio-economic groups during the operation of the heat-health alert system.
极端高温警报是澳大利亚最常用的天气预报服务形式，但很少有研究记录其在改善健康结果方面的有效性。本研究旨在检查与维多利亚州（澳大利亚）热健康警报和响应系统（HARS）激活相关的与温度相关的死亡率的时间变化。我们使用拟泊松回归和分布式滞后非线性模型（dlnm）检查温度与死亡率之间的关系，并比较了两个时期之间的温度-死亡率关系：时期 1- HARS 之前（1992-2009 年）和时期 2- HARS 之后（2010-2019 年）。由于 HARS 高度重视热浪的影响，我们还比较了两个时期之间热浪事件的主要影响。热浪定义为三个水平，包括连续三天处于第 97、98 和 99 百分位。我们还通过包含具有均等间隔结和每年 8 自由度的年的自然立方 B 样条来控制季节性的潜在混杂影响。暴露-反应曲线显示，与时期 1 相比，时期 2 的温度死亡率降低。时期 2 相对于时期 1 的相对风险比（RRR）均小于 1，并且逐渐从 0.86（95%CI，0.72-1.03）降至 0.64（95%CI，0.33-1.22），归因风险百分比的差异从 13.2%增加到 25.3%。与热浪相关的死亡风险降低了 3.4%（RR 1.068，95%CI，1.024-1.112 与 RR 1.034，95%CI，0.986-1.082）和 10%（RR 1.16，95%CI，1.10-1.22 与 RR 1.06，95%CI，1.002-1.119）对于所有人群。该研究表明，在极端高温和高强度热浪条件下，维多利亚州 HARS 运行后，与热相关的死亡率有所下降。进一步的研究可以调查在热健康警报系统运行期间，不同社会经济群体的人群死亡率变化的程度。
