Skip to main content
. 2021 Nov 25;375:e065653. doi: 10.1136/bmj-2021-065653

Table 2.

Excess relative risk and excess absolute risk (95% confidence interval) of emergency department (ED) visits for all causes associated with extreme heat over lag days 0-5, by age, sex, and low income status

Characteristics No (%) of ED visits Excess relative risk (%) P value Excess absolute risk (No/100 000 people at risk/day) P value
Age (years):
 18-24 2 102 380 (9.6) 9.5 (8.0 to 11.0) <0.001 8.5 (7.3 to 9.7) <0.001
 25-34 2 904 132 (13.2) 9.9 (8.6 to 11.1) 10.2 (9.0 to 11.4)
 35-44 2 906 914 (13.2) 7.4 (6.2 to 8.6) 6.9 (5.8 to 7.9)
 45-54 3 183 433 (14.5) 10.3 (9.1 to 11.5) 9.0 (8.0 to 9.9)
 55-64 3 276 096 (14.9) 8.8 (7.6 to 10.0) 8.0 (7.0 to 9.0)
 65-74 3 228 092 (14.7) 7.6 (6.4 to 8.7) 7.6 (6.5 to 8.8)
 ≥75 4 395 623 (20.0) 3.6 (2.7 to 4.6) 4.5 (3.3 to 5.7)
Sex:
 Men 9 314 254 (42.4) 9.5 (8.8 to 10.2) <0.001 9.6 (8.9 to 10.2) <0.001
 Women 12 678 437 (57.6) 6.5 (5.9 to 7.1) 7.4 (6.8 to 8.0)
Low income status*:
 Yes 3 111 751 (41.0) 5.6 (4.6 to 6.6) 0.53 12.6 (10.2 to 14.9) <0.001
 No 4 476 250 (59.0) 6.1 (4.9 to 7.3) 6.0 (5.0 to 7.0)

Ambient temperature (°F)=(°C×9/5)+32.

Extreme heat was defined based on the 95th centiles of local county specific temperature distribution during the warm season, and excess risks are expressed versus the local first centile. On average across the country, extreme heat was 34.4°C.

*

Only among a subset of members (32% (n=23 803 556/74 188 445) of total unique beneficiaries) who were enrolled in Medicare Advantage programmes that include pharmacy benefits.