Association of Simulated COVID-19 Policy Responses for Social Restrictions and Lockdowns With Health-Adjusted Life-Years and Costs in Victoria, Australia

Abstract

Countries have varied enormously in how they have responded to the COVID-19 pandemic, ranging from elimination strategies (eg, Australia, New Zealand, Taiwan) to tight suppression (not aiming for elimination but rather to keep infection rates low [eg, South Korea]) to loose suppression (eg, Europe, United States) to virtually unmitigated (eg, Brazil, India). Weighing the best option, based on health and economic consequences due to lockdowns, is necessary. To determine the optimal policy response, using a net monetary benefit (NMB) approach, for policies ranging from aggressive elimination and moderate elimination to tight suppression (aiming for 1-5 cases per million per day) and loose suppression (5-25 cases per million per day). Using governmental data from the state of Victoria, Australia, and other collected data, 2 simulation models in series were conducted of all residents (population, 6.4 million) for SARS-CoV-2 infections for 1 year from September 1, 2020. An agent-based model (ABM) was used to estimate daily SARS-CoV-2 infection rates and time in 5 stages of social restrictions (stages 1, 1b, 2, 3, and 4) for 4 policy response settings (aggressive elimination, moderate elimination, tight suppression, and loose suppression), and a proportional multistate life table (PMSLT) model was used to estimate health-adjusted life-years (HALYs) associated with COVID-19 and costs (health systems and health system plus gross domestic product [GDP]). The ABM is a generic COVID-19 model of 2500 agents, or simulants, that was scaled up to the population of interest. Models were specified with data from 2019 (eg, epidemiological data in the PMSLT model) and 2020 (eg, epidemiological and cost consequences of COVID-19). The NMB of each policy option at varying willingness to pay (WTP) per HALY was calculated: NMB = HALYs × WTP - cost. The estimated most cost-effective (optimal) policy response was that with the highest NMB. Estimated SARS-CoV-2 infection rates, time under 5 stages of restrictions, HALYs, health expenditure, and GDP losses. In 100 runs of both the ABM and PMSLT models for each of the 4 policy responses, 31.0% of SARS-CoV-2 infections, 56.5% of hospitalizations, and 84.6% of deaths occurred among those 60 years and older. Aggressive elimination was associated with the highest percentage of days with the lowest level of restrictions (median, 31.7%; 90% simulation interval [SI], 6.6%-64.4%). However, days in hard lockdown were similar across all 4 strategies. The HALY losses (compared with a scenario without COVID-19) were similar for aggressive elimination (median, 286 HALYs; 90% SI, 219-389 HALYs) and moderate elimination (median, 314 HALYs; 90% SI, 228-413 HALYs), and nearly 8 and 40 times higher for tight suppression and loose suppression, respectively. The median GDP loss was least for moderate elimination (median, $41.7 billion; 90% SI, $29.0-$63.6 billion), but there was substantial overlap in simulation intervals between the 4 strategies. From a health system perspective, aggressive elimination was optimal in 64% of simulations above a WTP of $15 000 per HALY, followed by moderate elimination in 35% of simulations. Moderate elimination was optimal from a GDP perspective in half of the simulations, followed by aggressive elimination in a quarter. In this simulation modeling economic evaluation of estimated SARS-CoV-infection rates, time under 5 stages of restrictions, HALYs, health expenditure, and GDP losses in Victoria, Australia, an elimination strategy was associated with the least health losses and usually the fewest GDP losses.