PIN68 COVID-19 Simulator: An Interactive Tool to Inform COVID-19 Intervention Policy Decisions in the United States
PIN68 COVID-19 Simulator: An Interactive Tool to Inform COVID-19 Intervention Policy Decisions in the United States
复制标题
DOI：
10.1016/j.jval.2020.08.909
复制
发表时间：
2020-12-11
期刊：
影响因子：
4.5
通讯作者：
Ayer T
中科院分区：
文献类型：
--
作者：
Chhatwal J;Dalgic O;Mueller P;Adee M;Xiao Y;Ladd MA;Linas BP;Ayer T
2019冠状病毒病（COVID-19）大流行影响了人类生活的各个领域，需要采取动态和快速的行动来抑制这种流行。我们的目标是为政策制定者开发一个开放获取的交互式工具，以及时做出决策，并评估不同非药物干预措施（不同强度和时间）对减少COVID-19传播的影响。方法我们开发了一个房室模型，COVID-19模拟器，以模拟COVID-19在50个州，DC和波多黎各的轨迹。该模型使用具有连续时间进展的易感、暴露、感染和隔离区室（SEIR模型）来定义。输入数据包括报告的特定州的每日病例、住院和死亡;通过临床研究估计的疾病流行病学参数、特定州的有效繁殖数、检测率和GPS位置数据报告的移动性变化。使用广义模拟退火对未观察到的参数进行校准，以匹配每日报告的病例和死亡。我们预测了不同级别社交距离措施（最低限制、当前干预、居家命令和封锁）下的未来COVID-19病例、COVID-19活跃病例、死亡、住院和重症监护病房入院情况。结果COVID-19模拟器可以帮助用户了解在不同时间点取消或增加限制的影响，包括死亡和所需的医院床位。在目前的干预水平下，预计到9月，美国每天的新病例将超过10万例;在许多州再次实施居家令是创造发病率下降轨迹的必要条件。如果政策不改变，高流行率州的死亡人数可能会在秋季急剧增加。结论COVID-19模拟器提供了一个互动平台，为控制COVID-19传播的政策决策提供信息。随着新数据的出现，模拟器定期更新，并将扩展到告知新的政策相关问题。
ObjectivesDynamic and fast actions are needed to suppress the coronavirus disease 2019 (COVID-19) pandemic, which has affected every sector of human life. Our objective was to develop an open-access, interactive tool for policy makers to inform timely decisions and evaluate the impact of different non-pharmaceutical interventions (of varied intensity and timing) on reducing the spread of COVID-19 in the.MethodsWe developed a compartmental model, the COVID-19 Simulator, to simulate the trajectory of COVID-19 in 50 states, DC, and Puerto Rico. The model is defined using Susceptible, Exposed, Infectious, and Recovered compartments (SEIR model) with continuous time progression. Input data included reported state-specific daily cases, hospitalizations, and deaths; disease epidemiology parameters estimated by clinical studies, state-specific effective reproduction numbers, testing rates, and changes in mobility as reported by GPS location data. Unobserved parameters were calibrated using generalized simulated annealing to match the daily reported cases and deaths. We projected future cases of COVID-19, active cases of COVID-19, deaths, hospitalizations, and intensive care unit admissions under different levels of social-distancing measures (minimal restrictions, current intervention, stay-at-home orders, and lockdown).ResultsThe COVID-19 Simulator can help users to understand the implications, including deaths and hospital beds needed, of removing or adding restrictions at different time points. Under current intervention levels, new cases per day are projected to surpass 100,000 by September in the US; implementing stay at home orders again in many states is necessary to create a downward trajectory in incidence. High prevalence states may see a dramatic increase in deaths by Fall if policies do not change.ConclusionsThe COVID-19 Simulator provides an interactive platform to inform policy decisions on controlling the spread of COVID-19. The simulator is updated on a regular basis as new data become available and will be extended to inform new policy-relevant questions.