Quantification of the effects of climatic conditions on French hospital admissions and deaths induced by SARS-CoV-2

Abstract

An estimation of the impact of climatic conditions-measured with an index that combines temperature and humidity, the IPTCC-on the hospitalizations and deaths attributed to SARS-CoV-2 is proposed. The present paper uses weekly data from 54 French administrative regions between March 23, 2020 and January 10, 2021. Firstly, a Granger causal analysis is developed and reveals that past values of the IPTCC contain information that allow for a better prediction of hospitalizations or deaths than that obtained without the IPTCC. Finally, a vector autoregressive model is estimated to evaluate the dynamic response of hospitalizations and deaths after an increase in the IPTCC. It is estimated that a 10-point increase in the IPTCC causes hospitalizations to rise by 2.9% (90% CI 0.7-5.0) one week after the increase, and by 4.1% (90% CI 2.1-6.4) and 4.4% (90% CI 2.5-6.3) in the two following weeks. Over ten weeks, the cumulative effect is estimated to reach 20.1%. Two weeks after the increase in the IPTCC, deaths are estimated to rise by 3.7% (90% CI 1.6-5.8). The cumulative effect from the second to the tenth weeks reaches 15.8%. The results are robust to the inclusion of air pollution indicators.

Figure 1

Comparison of the time series of total hospitalizations and deaths and the average IPTCC, March 23, 2020 to January 10, 2021.

Figure 2

Dynamic responses of hospitalizations and deaths to a shock on the IPTCC. Note: The solid line gives the estimated impulse responses. The dashed lines give the 90% confidence intervals generated by Monte Carlo with 5000 repetitions. The size of the increase in the IPTCC is set to a 10-point increase. The size of the increase in hospitalizations is set to a one-percent increase. The responses are the percentage change in the number of hospitalizations and deaths.