RAPID: Harnessing the power of multiple models for outbreak management

对于许多最具破坏性或最令人担忧的病原体，例如导致COVID-19的SARS-CoV-2病毒，多个科学团体开发了定量模型来预测疾病动态并评估可能的干预措施。这些模型在预测和建议方面往往有很大差异，反映了不同的政策假设，以及生物和管理过程的科学、后勤和其他不确定性。这种不确定性可能对决策者构成挑战，阻碍干预规划和应对。因此，决策者可能选择依赖单一可信的咨询意见来源，或在出现共识时依赖共识，而不相信决策将是最好的。然而，不太理想的决策意味着更多的生命可能会丧失，或者使用的资源可能比需要的更多。 面对生物学、流行病学和操作上的不确定性，采用系统性战略，将使用多种模型制定政策的过程正规化，可以提高疫情应对政策的效力和效率。COVID-19疫情爆发是全球大多数国家的重大健康问题。这项工作的目的是直接解决这个问题，在真实的时间。这项工作还将为未来的疫情应对提供框架。 许多应对COVID-19大流行的模型正在开发中，或最近发布。该项目将开发多模型启发协议，嵌入一个强大的决策框架，正式承认我们对这种新型病原体的不确定性，并避免已知的偏见来源。对不确定性的充分认识和解释对于决策制定和公众沟通都至关重要。与国家有关的目标（例如，最大限度地减少死亡）和干预（例如，社会距离）将在此过程中进行评估。该项目将把正式的专家启发方法（通常用于从个别专家那里征求意见）与来自多个研究小组的建模分析相结合，以加强疫情管理的决策。各小组将预测不同干预措施下的疾病动态，并将使用决策分析对总体产出进行分析，以根据决策者的目标对干预措施进行评价。该项目将开展这项工作，以解决在COVID-19大流行期间面临不确定性的关键政策决策需求。该奖项反映了NSF的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

For many of the most damaging or worrying pathogens, such as the SARS-CoV-2 virus that causes COVID-19, multiple scientific groups develop quantitative models to forecast disease dynamics and assess possible interventions. These models often differ significantly in their projections and recommendations, reflecting different policy assumptions, as well as scientific, logistical, and other uncertainty about biological and management processes. Such uncertainty can be challenging for policymakers, hindering intervention planning and response. Policymakers may thus choose to rely on single trusted sources of advice, or on consensus where it appears, without confidence that decisions will be the best possible. However, less-than-optimal decisions mean more lives may be lost or more resources used than needed. In the face of biological, epidemiological, and operational uncertainties, systematic strategies to formalize the process of using multiple models to develop policy can improve the effectiveness and efficiency of policy responses to outbreaks. The COVID-19 pandemic outbreak is a major health issue for most countries in the world. This work is intended to directly address this current problem in real time. The work will also provide a framework for future outbreak response. Many models to address the COVID-19 pandemic are in development, or recently published. This project will develop multiple-model elicitation protocols, embedded in a strong framework for decision making, that formally acknowledges our uncertainty about this novel pathogen, and avoids known sources of bias. A full acknowledgment and accounting of uncertainty is critical both for decision making and for public communication. Nationally relevant objectives (e.g., minimizing deaths) and interventions (e.g., social distancing) will be assessed during this process. The project will merge formal expert elicitation methods (usually used to elicit opinions from individual experts) with modeling analyses from multiple research groups to enhance decision making for outbreak management. Groups will project disease dynamics under different interventions, and the ensemble of outputs will be analyzed using decision analysis to provide an evaluation of interventions against the policy makers’ objectives. The project will conduct this exercise to address key policy decision-making needs in the face of uncertainty during the COVID-19 pandemic.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.