Table 4 Modification by city characteristics of the two-day cumulative effect of extreme heat on mortality. Comparison of the predicted change in mortality at the 25th and 75th percentile of the effect modifier distribution
Mean of warm months’ temperature (°C), % (95% CI)Variance of warm months’ temperature (°C), % (95% CI)Central air conditioning (%), % (95% CI)Population density (population/km2), % (95% CI)
Change in total mortality at the:
    25th percentile9.25* (6.22 to 12.4)4.33* (1.90 to 6.83)9.68* (6.55 to 12.9)3.36* (1.17 to 5.61)
    75th percentile3.81* (1.42 to 6.25)7.77* (5.03 to 10.6)3.24* (0.44 to 6.13)5.46* (3.51 to 7.44)
Change in MI mortality at the:
    25th percentile7.86** (2.44 to 13.6)3.66 (−1.07 to 8.61)7.99** (2.70 to 13.6)2.87 (−2.09 to 8.08)
    75th percentile2.62** (−1.98 to 7.43)5.20 (−0.12 to 10.8)2.05** (−3.38 to 7.78)3.99 (−0.27 to 8.43)
Change in CA mortality at the:
    25th percentile15.0 (−1.08 to 33.6)0.29* (−10.8 to 12.8)11.8 (−2.19 to 27.9)−0.96* (−12.5 to 12.2)
    75th percentile1.64 (−10.4 to 15.3)15.8* (2.04 to 31.3)3.65 (−10.6 to 20.2)3.79* (−6.67 to 15.4)
  • MI, myocardial infarction; CA, cardiac arrest.

  • *Two-sided p value <0.05 in the meta-regression model; **two-sided p value <0.1 in the meta-regression model.