ABSTRACT
Short-term probabilistic forecasts of the trajectory of the COVID-19 pandemic in the United States have served as a visible and important communication channel between the scientific modeling community and both the general public and decision-makers. Forecasting models provide specific, quantitative, and evaluable predictions that inform short-term decisions such as healthcare staffing needs, school closures, and allocation of medical supplies. In 2020, the COVID-19 Forecast Hub (https://covid19forecasthub.org/) collected, disseminated, and synthesized hundreds of thousands of specific predictions from more than 50 different academic, industry, and independent research groups. This manuscript systematically evaluates 23 models that regularly submitted forecasts of reported weekly incident COVID-19 mortality counts in the US at the state and national level. One of these models was a multi-model ensemble that combined all available forecasts each week. The performance of individual models showed high variability across time, geospatial units, and forecast horizons. Half of the models evaluated showed better accuracy than a naive baseline model. In combining the forecasts from all teams, the ensemble showed the best overall probabilistic accuracy of any model. Forecast accuracy degraded as models made predictions farther into the future, with probabilistic accuracy at a 20-week horizon more than 5 times worse than when predicting at a 1-week horizon. This project underscores the role that collaboration and active coordination between governmental public health agencies, academic modeling teams, and industry partners can play in developing modern modeling capabilities to support local, state, and federal response to outbreaks. f
Subject(s)
COVID-19ABSTRACT
With the emerging COVID-19 crisis, a critical task for public health officials and policy makers is to decide how to prioritize, locate, and allocate scarce resources. To answer these questions, decision makers need to be able to determine the location of the required resources over time based on emerging hot spot locations. Hot spots are defined as concentrated areas with sharp increases in COVID19 cases. Hot spots place stress on existing healthcare resources, resulting in demand for resources potentially exceeding current capacity. This research will describe a value based resource allocation approach that seeks to coordinate demand, as defined by uncertain epidemiological forecasts, with the value of adding additional resources such as hospital beds. Value is framed as a function of the expected usage of a marginal resource (bed, ventilator, etc). Subject to certain constraints, allocation decisions are operationalized using a nonlinear programming model, allocating new hospital beds over time and across a number of geographical locations. The results of the research show a need for a value based approach to assist decision makers at all levels in making the best possible decisions in the current highly uncertain and dynamic COVID environment.
Subject(s)
COVID-19ABSTRACT
The COVID19 pandemic has highlighted the lack of resilience in supply chains, as global networks fail from disruptions at single nodes and connections. Through an overview of the existing vaccine and pharmaceutical supply chain publications focusing on resilience, as well as recent papers reporting modeling of resilience in supply chains across multiple fields, we find that models for supply chain resilience are few and most of them are focused on individual dimensions of resilience rather than on comprehensive strategy necessary for scaling up vaccine production and distribution in emergency settings. We find that COVID19 resulted in a wave of interest to supply chain resilience, but publications from 2020 are narrow in focus and largely qualitative in nature; evidence-based models and measures are rare. Further, publications often focus exclusively on specific portions of the specific supply chain of interest, excluding associated supporting networks, such as transportation, social and command and control (C2) necessary for vaccine production and equitable distribution. This lack of network analysis is a major gap in the literature that needs to be bridged in order to create methods of real-time analysis and decision tools for the COVID19 vaccine supply chain. We conclude that a comprehensive, quantitative approach to network resilience that encompasses the supply chain in the context of other social and physical networks is needed in order to address the emerging challenges of a large-scale COVID-19 vaccination program. We further find that the COVID-19 pandemic underscores the necessity of positioning supply chain resilience within a multi-network context and formally incorporating temporal dimensions into analysis through the NAS definition of resilience, plan, absorb, recover, adapt, to ensure essential needs are met across all dimensions of society.