Using a novel two-stage strategy to characterize the spatial distribution of associations between temperature and COVID-19: A case study in the continental United States.

BACKGROUND

Numerous studies have studied the association between daily average temperature (DAT) and daily COVID-19 confirmed cases, which show considerable heterogeneity, even opposite results, among different regions. Such heterogeneity suggests that characterizing the association on a large area scale would ignore the local variation, even obtain false results in some local regions. So, characterizing the spatial distribution of heterogeneous DAT-COVID-19 associations and exploring the causes plays an important role on making temperature-related region-specific intervention measures and early-warning systems.

METHODS

Aiming to characterize the spatial distribution of associations between DAT and COVID-19 confirmed cases in the continental United States, we proposed a novel two-stage strategy. In the first stage, we used the common stratified distributed lag nonlinear model to obtain the rough state-specific associations. In the second stage, conditional autoregression was used to spatially smooth the rough estimations. Furtherly, based on the idea, two modified strategies were used to investigate the time-varying associations and the modification effects derived from the vaccination campaign.

RESULTS

Around one-third of states exhibit no significant association between DAT and daily confirmed COVID-19 cases. Most of the remaining states present a low risk at low DAT and a high risk at high DAT, but several states present opposite associations. The average association curve presents a 'S' shape with positive association between -8 - 18 °C and keeping flat out of the range. An increased vaccination coverage rate will increase the risk when DAT < 12 °C, but slightly affect the risk when DAT > 12 °C.

CONCLUSION

A considerable spatial heterogeneity of DAT-COVID-19 associations exists in America and the average association curve presents a 'S' shape. The vaccination campaign significantly modifies the association when DAT is low, but only make a slight modification when DAT is high.

Informations de copyright

Copyright © 2022 The Authors. Published by Elsevier B.V. All rights reserved.

Déclaration de conflit d'intérêts

Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.