Key points are not available for this paper at this time.
ISEE-212 Objective: To compare the extent to which short-term mortality displacement can explain heat deaths in Delhi, São Paulo and London given contrasting demographic and epidemiological profiles of these cities. Design: Time-series analysis. Methods: For each city, daily mortality was examined in relation to daily ambient temperature using auto-regressive Poisson models with adjustment for season, relative humidity, rainfall, particulate air pollution, day of week and public holidays. Unconstrained distributed lag models were used to identify the extent to which heat-related excesses were followed by deficits (displacement). Results: For each city, an increase in mortality was observed with temperatures above a threshold of 20 degrees Celsius at lags zero and one. The excess risk per degree Celsius increase above 20 Celsius was greatest in Delhi and smallest in London. In Delhi an excess was apparent on days up to three weeks after heat, after which a deficit was observed which offset just part of the overall excess. In London the heat excess persisted only two days and was followed after three days by deficits, such that the sum of effects was zero by day eleven. The pattern in Sao Paulo was intermediate between that in London and Delhi. The excess risk summed over 21 days was 4.45 (95% CI 2.40 to 6.55) per degree Celsius above the heat threshold in Delhi, 0.86 (−0.18 to 1.92) in São Paulo and −1.02 (−2.71 to 0.70) in London. Conclusion: Short-term mortality displacement accounts for a proportion of heat-related deaths in each of these cities, but its contribution appears to be relatively minor in Delhi where infectious and acute causes of mortality still predominate.
Hajat et al. (Thu,) studied this question.