Full dataset for fitting the linear mixed models.

In the absence of definitive treatments or vaccines, the primary strategy to mitigate the COVID-19 pandemic relied on non-pharmaceutical interventions. By the end of 2020, COVID-19 vaccines had been developed and initiated for preventive purposes. To better understand the association between various mitigation measures and their impact on the pandemic, we analyzed the effect of vaccination coverage, international travel, traveler positivity rates, and the stringency of public health measures on the incidence of COVID-19 cases and hospitalizations at the provincial level in Belgium. We identified several important interactions among the covariates that influence the incidence of COVID-19 confirmed cases. Specifically, the best-fitting model (AIC = 965.658) revealed significant interactions between lagged vaccination coverage and the stringency index, as well as between incoming travel rates and positivity rates. Additionally, when modeling COVID-19 hospitalizations, a significant interaction was observed between the incoming travel rate and the stringency index. Model performance improved substantially when incorporating the incidence of confirmed cases as a covariate (AIC = 1061.516 vs. AIC = 432.708), while highlighting key interactions between confirmed cases and traveler positivity rates, as well as between lagged vaccination coverage and incoming travel rates. These findings underscore the intricate interplay between public health interventions, population immunity, and mobility patterns in shaping the course of the COVID-19 pandemic.

在尚无特效治疗药物与疫苗的阶段，遏制新冠病毒病（COVID-19）大流行的主要策略依赖于非药物干预手段。截至2020年末，新冠疫苗已完成研发并投入预防性使用。为更清晰地明晰各类防控措施及其对大流行的影响之间的关联，我们针对比利时省级层面的新冠确诊病例与住院病例数据，分析了疫苗接种覆盖率、国际旅行、入境旅客阳性率以及公共卫生措施严格程度等因素对当地疫情的影响。我们识别出了影响新冠确诊病例发病数的协变量间存在多组重要交互效应。具体而言，最优拟合模型（赤池信息准则AIC=965.658）显示，滞后疫苗接种覆盖率与公共卫生严格指数之间、入境旅行率与阳性率之间均存在显著交互效应。此外，在构建新冠住院病例的预测模型时，我们观察到入境旅行率与公共卫生严格指数之间存在显著交互效应。将确诊病例发病数作为协变量纳入模型后，模型性能得到显著提升（赤池信息准则从1061.516优化至432.708），同时还凸显出确诊病例与旅客阳性率之间、滞后疫苗接种覆盖率与入境旅行率之间的关键交互效应。本研究结果凸显了公共卫生干预措施、人群免疫水平与流动模式之间的复杂交互作用，而这些因素共同塑造了新冠大流行的发展态势。