The economic impact of COVID-19 interventions: A mathematical modeling approach

Abstract

Prior to vaccination or drug treatment, non-pharmaceutical interventions were almost the only way to control the coronavirus disease 2019 (COVID-19) epidemic. After vaccines were developed, effective vaccination strategies became important. The prolonged COVID-19 pandemic has caused enormous economic losses worldwide. As such, it is necessary to estimate the economic effects of control policies, including non-pharmaceutical interventions and vaccination strategies. We estimated the costs associated with COVID-19 according to different vaccination rollout speeds and social distancing levels and investigated effective control strategies for cost minimization. Age-structured mathematical models were developed and used to study disease transmission epidemiology. Using these models, we estimated the actual costs due to COVID-19, considering costs associated with medical care, lost wages, death, vaccination, and gross domestic product (GDP) losses due to social distancing. The lower the social distancing (SD) level, the more important the vaccination rollout speed. SD level 1 was cost-effective under fast rollout speeds, but SD level 2 was more effective for slow rollout speeds. If the vaccine rollout rate is fast enough, even implementing SD level 1 will be cost effective and can control the number of critically ill patients and deaths. If social distancing is maintained at level 2 at the beginning and then relaxed when sufficient vaccinations have been administered, economic costs can be reduced while maintaining the number of patients with severe symptoms below the intensive care unit (ICU) capacity. Korea has wellequipped medical facilities and infrastructure for rapid vaccination, and the public's desire for vaccination is high. In this case, the speed of vaccine supply is an important factor in controlling the COVID-19 epidemic. If the speed of vaccination is fast, it is possible to maintain a low level of social distancing without a significant increase in the number of deaths and hospitalized patients with severe symptoms, and the corresponding costs can be reduced.

Keywords: COVID-19; cost estimation; mathematical model; social distancing; vaccination.

Figure 1

(A) Number of weekly confirmed cases for age groups (0–9,10–19,20–29,30–39,40–49,50–59,60–69, 70+) and (B) Daily first, second and third vaccination doses. The solid curve indicates average second doses per week. The horizontal dashed lines represent daily vaccination doses (ν=100000, 200000, . . , 500000) used in the model simulation.

Figure 2

Schematic diagram for the proposed model.

Figure 3

The effects of vaccination on: (A) Confirmed cases, (B) cumulative confirmed cases, (C) death, and (D) hospitalized population in the ICU for C_ν=1,2,⋯, 5 and the SD level 0, 1, 2, 3. The dashed line in the bottom panels represents the capacity of ICU bed for COVID-19 patients in Korea. The circle on each curve represents the time at which the vaccination coverage rate reaches 80%.

Figure 4

Ratio of cost to GDP for C_ν=1, 2, ⋯, 5 for (A) admission to the treatment center for mild patients (B) home treatment for mild patients.

Figure 5

Ratio of cost to GDP for each SD level mitigation scenario when the vaccination coverage rate reaches 60, 70, or 80% in case of admission to the treatment center for mild patients for (A) τ = 0.79 and (B) τ = 0.6 for (Top) C_ν = 3. (Bottom) C_ν = 1, 2, ..., 5. (N/A indicates that social distancing easing was not implemented).

Figure 6

The ratio of total costs to GDP for τ=0.3 − 0.8 and for R_t corresponding to β that is varied as β × C_β for C_β=0.7 − 1.3 under SD LV 0, 1, 2, and 3, and the rollout speed C_ν=1, 2, 3, 4, 5. Red lines indicate the case when the maximum number of hospitalized patients with severe symptoms is the capacity of the intensive care unit, 2,800, for COVID-19 patients in Korea. It means that the maximum number hospitalized patients with severe symptoms is lower than capacity on the left side of the red curves.

Figure 7

Sensitivity index of β, τ, ρ, ν, the vaccine cost, and GDP loss rate for C_ν=1, 2, ..., 5 and each SD level.