A decision tool to inform the optimal use of non-pharmaceutical interventions during the COVID-19 pandemic

PROJECT SUMMARY/ABSTRACT As the prospect for the elimination of COVID-19 in the near future remains uncertain, non-pharmaceutical interventions (NPIs) such as limiting social gatherings, quarantine after exposure to the virus, and school closure, will continue to play important roles in mitigating the morbidity and mortality associated with the pandemic. Since these interventions impose immense economic, social, and health-related costs, their use should be recommended only when epidemic control benefits outweigh their adverse consequences. Our overall objective in this proposal is to develop an analytical decision tool to optimize the use of NPIs based on latest information related to the local epidemiology of COVID-19, the effectiveness of different NPIs, and the population's stated disutility associated with these interventions. This decision tool is structured to provide a transparent mechanism to communicate the rationale for the current policy regarding the use of NPIs and the conditions under which the policy would change. To develop our decision tools, this proposal has three specific aims: 1) to develop state-level decision models that identify the optimal combination of NPIs, in real-time, and based on the projected loss in the quality-adjusted life-years (QALYs) and the disutility borne by the population under various combinations of NPIs under various combinations of NPIs; 2) to design, conduct, and analyze discrete-choice experiments to estimate the disutility weights of different NPIs as borne by population members due to social, economic, and health consequences of these programs; and 3) to estimate the societal tolerance for loss in QALYs due to existing infectious diseases without triggering NPIs. This tolerance threshold can be estimated using historical data related to past pandemic and seasonal influenza and will serve as a benchmark to decide when the burden of COVID-19 is low enough to lift all NPIs, at least for a short term. The research proposed in this project is innovative as it develops a novel, principled approach to consolidate real-time data from three different sources to optimize the use of NPIs: 1) COVID-19 cases, hospitalizations, and deaths as projected by existing and new predictive models of COVID-19 pandemic, 2) effectiveness of various NPIs in breaking the transmission of SARS-CoV-2, and 3) disutility weights of NPIs directly elicited from target populations. The proposed research is significant because it meets the critical needs of policymakers to identify evidence-based and real-time recommendations regarding the efficient use of NPIs to contain the burden of COVID-19. The methods and decision tools developed as part of this project could also be used in responding to other existing and future infectious threats where NPIs are employed.