Skip to main content
. 2020 Dec 21:M20-6558. doi: 10.7326/M20-6558

Figure. Infections, costs, and economic efficiency, by COVID-19 mitigation strategy.

Figure. Infections, costs, and economic efficiency, by COVID-19 mitigation strategy. 1timeLT = 1-time laboratory testing; CampusClosed = campus remains closed with only online education; COVID-19 = coronavirus disease 2019; DesigIsol = student isolation in a separate, college-sponsored location; ExtSocDist = extensive social distancing; ICU = intensive care unit; LT = laboratory testing; Masks = mask-wearing policies; MinSocDist = minimal social distancing; NoIntervention = campus operates as it did before COVID-19 without any mitigation interventions; NPI = nonpharmacologic intervention; ResIsol = residence isolation in student dorm room; RLT = routine LT; RLTqX = RLT every X days. A-C. The number and source of infections among students (A) and faculty (B) for each strategy, and total costs (C). On the left are the NoIntervention and CampusClosed strategies. The 4 broad NPI strategies (MinSocDist, ExtSocDist, Masks, and combined ExtSocDist and Masks) are further stratified by the use and frequency of LT, ranging from no LT, where those who report symptoms associated with COVID-19 are asked to isolate in their residence for 10 d; to 1 test for those who report symptoms to confirm placement in isolation; to RLT for all students and faculty at the start of the semester; to RLT among asymptomatic students and faculty at 3-, 7-, or 14-d intervals. Infections decrease as strategies increase in intensity, from MinSocDist to the ExtSocDist+Masks strategy. In each case, adding LT further decreases infections. Among students, most infections are from other students (A). Among faculty, depending on the strategy, most infections are from the community and other faculty (B). In strategies without RLT, hospital and ICU costs account for >50% of total costs (C). In strategies with RLT, testing accounts for >50% of total costs. Cost per test was $10. D. The efficiency frontier (cost per infection prevented) for COVID-19 mitigation strategies. The efficiency frontier represents the relationship between infections prevented (vertical axis) and total costs (horizontal axis). NoIntervention is shown in the open red circle on the lower left. Without RLT or testing at the semester start, regardless of isolation approach, there is clustering (ovals) of strategies involving MinSocDist (triangles), ExtSocDist (circles), Masks (diamonds), and ExtSocDist+Masks (squares). Unshaded ovals represent strategies where masks are not incorporated, and beige ovals represent clustering of strategies where masks are incorporated. More infections are prevented when masks are used. Symbols on the solid black line represent economically efficient strategies. The slope of the solid line represents the incremental cost per infection prevented for each strategy, compared with the next less costly efficient strategy. Testing at 14-, 7-, or 3-d intervals prevents additional infections, but at a substantially increased cost per infection prevented.

1timeLT = 1-time laboratory testing; CampusClosed = campus remains closed with only online education; COVID-19 = coronavirus disease 2019; DesigIsol = student isolation in a separate, college-sponsored location; ExtSocDist = extensive social distancing; ICU = intensive care unit; LT = laboratory testing; Masks = mask-wearing policies; MinSocDist = minimal social distancing; NoIntervention = campus operates as it did before COVID-19 without any mitigation interventions; NPI = nonpharmacologic intervention; ResIsol = residence isolation in student dorm room; RLT = routine LT; RLTqX = RLT every X days. A-C. The number and source of infections among students (A) and faculty (B) for each strategy, and total costs (C). On the left are the NoIntervention and CampusClosed strategies. The 4 broad NPI strategies (MinSocDist, ExtSocDist, Masks, and combined ExtSocDist and Masks) are further stratified by the use and frequency of LT, ranging from no LT, where those who report symptoms associated with COVID-19 are asked to isolate in their residence for 10 d; to 1 test for those who report symptoms to confirm placement in isolation; to RLT for all students and faculty at the start of the semester; to RLT among asymptomatic students and faculty at 3-, 7-, or 14-d intervals. Infections decrease as strategies increase in intensity, from MinSocDist to the ExtSocDist+Masks strategy. In each case, adding LT further decreases infections. Among students, most infections are from other students (A). Among faculty, depending on the strategy, most infections are from the community and other faculty (B). In strategies without RLT, hospital and ICU costs account for >50% of total costs (C). In strategies with RLT, testing accounts for >50% of total costs. Cost per test was $10. D. The efficiency frontier (cost per infection prevented) for COVID-19 mitigation strategies. The efficiency frontier represents the relationship between infections prevented (vertical axis) and total costs (horizontal axis). NoIntervention is shown in the open red circle on the lower left. Without RLT or testing at the semester start, regardless of isolation approach, there is clustering (ovals) of strategies involving MinSocDist (triangles), ExtSocDist (circles), Masks (diamonds), and ExtSocDist+Masks (squares). Unshaded ovals represent strategies where masks are not incorporated, and beige ovals represent clustering of strategies where masks are incorporated. More infections are prevented when masks are used. Symbols on the solid black line represent economically efficient strategies. The slope of the solid line represents the incremental cost per infection prevented for each strategy, compared with the next less costly efficient strategy. Testing at 14-, 7-, or 3-d intervals prevents additional infections, but at a substantially increased cost per infection prevented.