Table 4.
Study Period (2009–2014) | Period 1 (2009–2011) | Period 2 (2012–2014) | ||||
---|---|---|---|---|---|---|
Estimate | (95% CI) | Estimate | (95% CI) | Estimate | (95% CI) | |
All-cause mortality | ||||||
Marital status (age) | ||||||
Widowed (65+) | 0.844 | (−1.895, 3.583) | −8.919 | (−13.533, −4.306) * | 2.795 | (−1.296, 6.887) |
Widowed (75+) | −2.671 | (−9.906, 4.563) | −21.640 | (−35.97, −7.310) * | −3.673 | (−15.721, 8.376) |
Education (age) | ||||||
None (19–64) | −0.144 | (−0.227, −0.061) * | −0.065 | (−0.188, 0.057) | −0.401 | (−0.544, −0.258) * |
Elementary (65+) | 4.239 | (2.070, 6.408) | −6.979 | (−10.693, −3.266) * | 3.995 | (0.466, 7.524) |
Elementary (75+) | 1.577 | (−3.359, 6.512) | −18.660 | (−28.563, −8.757) * | −0.004 | (−9.234, 9.226) |
Cardiovascular mortality | ||||||
Marital status (age) | ||||||
Widowed (65+) | −0.944 | (−2.438, 0.550) | −0.880 | (−3.396, 1.636) | −1.409 | (−3.604, 0.787) |
Widowed (75+) | −4.524 | (−8.617, −0.431) * | −2.471 | (−10.24, 5.298) | −9.140 | (−15.776, −2.504) * |
Education (age) | ||||||
None (19–64) | −0.091 | (−0.140, −0.042) * | −0.005 | (−0.053, 0.043) | −0.284 | (−0.382, −0.187) * |
Elementary (65+) | 0.631 | (−0.597, 1.859) | 0.100 | (−1.784, 1.984) | 0.510 | (−1.352, 2.372) |
Elementary (75+) | −1.669 | (−4.888, 1.550) | −1.911 | (−6.938, 3.116) | −1.650 | (−6.956, 3.657) |
Respiratory mortality | ||||||
Marital status (age) | ||||||
Widowed (65+) | −0.944 | (−2.438, 0.550) | 0.714 | (−0.594, 2.021) | 2.158 | (0.894, 3.423) |
Widowed (75+) | −4.524 | (−8.617, −0.431) * | 1.239 | (−2.840, 5.318) | 6.705 | (2.778, 10.632) |
Education (age) | ||||||
None (19–64) | −0.032 | (−0.055, −0.009) * | 0.008 | (−0.013, 0.030) | −0.167 | (−0.221, −0.114) * |
Elementary (65+) | −0.884 | (−1.582, −0.185) * | −2.024 | (−3.084, −0.964) * | 2.084 | (1.025, 3.142) |
Elementary (75+) | −3.167 | (−5.030, −1.305) * | −5.123 | (−8.204, −2.041) * | 5.752 | (2.686, 8.818) |
Notes. The effects are the estimates of difference-in-differences (DD) that were obtained by comparing mortality changes due to heat wave alert announcement for heat wave days versus that for non-heat wave days (i.e., counterfactual) through the difference-in-differences analysis. Negative estimates indicate reduction in mortality. Statistically significant results (significant level = 0.05) are marked with an asterisk (*).