Non-pharmacological interventions of travel restrictions and cancelation of public events had a major reductive mortality affect during pre-vaccination coronavirus disease 2019 period

Haoyu Wen; Fang Shi; Yan Liu; Cong Xie; Guiyu Qin; Fang Wang; Xiaoxue Liu; Jianjun Bai; Qiumian Hong; Runxue Ma; Chuanhua Yu

doi:10.3389/fmed.2022.914732

Non-pharmacological interventions of travel restrictions and cancelation of public events had a major reductive mortality affect during pre-vaccination coronavirus disease 2019 period

Front Med (Lausanne). 2022 Aug 22:9:914732. doi: 10.3389/fmed.2022.914732. eCollection 2022.

Authors

Haoyu Wen¹, Fang Shi¹, Yan Liu¹, Cong Xie², Guiyu Qin¹, Fang Wang³, Xiaoxue Liu¹, Jianjun Bai¹, Qiumian Hong¹, Runxue Ma¹, Chuanhua Yu^{1

4}

Affiliations

¹ Department of Epidemiology and Biostatistics, School of Public Health, Wuhan University, Wuhan, China.
² Hubei Provincial Center for Disease Control and Prevention, Institute of Preventive Medicine Information, Wuhan, China.
³ Department of Biostatistics, School of Public Health, Xuzhou Medical University, Xuzhou, China.
⁴ China Global Health Institute, Wuhan University, Wuhan, China.

Abstract

Background: The coronavirus disease 2019 (COVID-19) is a severe acute respiratory disease that poses a continuous threat to global public health. Many non-pharmacological interventions (NPIs) have been implemented to control the COVID-19 pandemic since the beginning. The aim of this study was to assess the impact of various NPIs on COVID-19 mortality during pre-vaccination and vaccination periods.

Methods: The COVID-19 data used in this study comes from Our World in Data, we used the Oxford Strict Index (OSI) and its five combination interventions as independent variables. The COVID-19 mortality date (MRT) was defined as a date when daily rate of 0.02 COVID-19 deaths per 100,000 population in a country was reached, and the COVID-19 vaccination date (VRT) was defined as people vaccinated reaching 70%. Linear regression and random forest models were used to estimate the impact of various NPI implementation interventions during pre-vaccination and vaccination periods. The performance of models was assessed among others with Shapley Additive Explanations (SHAP) explaining the prediction capability of the model.

Results: During the pre-vaccination period, the various NPIs had strong protective effect. When the COVID-19 MRT was reached, for every unit increase in OSI, the cumulative mortality as of June 30, 2020 decreased by 0.71 deaths per 100,000 people. Restrictions in travel (SHAP 1.68) and cancelation of public events and gatherings (1.37) had major reducing effect on COVID-19 mortality, while staying at home (0.26) and school and workplace closure (0.26) had less effect. Post vaccination period, the effects of NPI reduced significantly: cancelation of public events and gatherings (0.25), staying at home (0.22), restrictions in travel (0.14), and school and workplace closure (0.06).

Conclusion: Continued efforts are still needed to promote vaccination to build sufficient immunity to COVID-19 in the population. Until herd immunity is achieved, NPI is still important for COVID-19 prevention and control. At the beginning of the COVID-19 pandemic, the stringency of NPI implementation had a significant negative association with COVID-19 mortality; however, this association was no longer significant after the vaccination rate reached 70%. As vaccination progresses, "cancelation of public events and gatherings" become more important for COVID-19 mortality.

Keywords: COVID-19; mortality; public health interventions; random forest; vaccines.