Cost Benefit Analysis of Limited Reopening Relative to a Herd Immunity Strategy or Shelter in Place for SARS-CoV-2 in the United States

BACKGROUND: Fierce debate about the health and financial tradeoffs presented by different COVID-19 pandemic mitigation strategies highlights the need for rigorous quantitative evaluation of policy options. OBJECTIVE: To quantify the economic value of the costs and benefits of a policy of continued limited reopening with social distancing relative to alternative COVID-19 response strategies in the United States. DESIGN: We estimate the number and value of quality-adjusted life-years (QALY) gained from mortality averted, with a value of $125,000 per QALY, and compare these benefits to the associated costs in terms of plausible effects on US GDP under a policy of continued limited reopening with social distancing relative to a policy of full reopening toward herd immunity. Using the same QALY value assumptions, we further evaluate cost-effectiveness of a return to Shelter-in-Place relative to a policy of limited reopening. SETTING: United States MEASUREMENTS: QALY and cost as percent of GDP of limited reopening with continued social distancing relative to a strategy of full reopening aimed at achieving herd immunity; a limited reopening “budget” measured in the number of months before this strategy fails to demonstrate cost-effectiveness relative to a full reopening; a shelter-in-place “threshold” measured in the number of lives saved at which a month of sheltering in place demonstrates cost effectiveness relative to the limited reopening strategy. RESULTS: QALY benefits from mortality averted by continued social distancing and limited reopening relative to a policy of full reopening exceed projected GDP costs if an effective vaccine or therapeutic can be developed within 11.1 months from late May 2020. White House vaccine projections fall within this date, supporting a partial reopening strategy. One month of shelter-in-place restrictions provides QALY benefits from averted mortality that exceed the associated GDP costs relative to limited reopening if the restrictions prevent at least 154,586 additional COVID-19 deaths over the course of the pandemic. Current models of disease progression suggest that limited reopening will not cause this many additional deaths, again supporting a limited reopening strategy. LIMITATION: Limited horizon of COVID-19 mortality projections; infection fatality ratio stable across strategies, ignoring both the potential for ICU overload to increase mortality and the deployment of partially effective therapeutics to decrease mortality; effect on GDP modeled as constant within a given phase of the pandemic; accounts for age and sex distribution of QALYs, but not effect of comorbidities; only considers impact from QALY lost due to mortality and from changes in GDP, excluding numerous other considerations, such as non-fatal COVID-19 morbidity, reduced quality of life caused by prolonged social distancing, or educational regression associated with prolonged school closures and restrictions. CONCLUSIONS: A limited reopening to achieve partial mitigation of COVID-19 is cost effective relative to a full reopening if an effective therapeutic or vaccine can be deployed within 11.1 months of late May 2020. One additional month of shelter-in-place restrictions should only be imposed if it saves at least 154,586 lives per month before the development of an effective therapeutic or vaccine relative to limited reopening.