Burden of Covid-19 restrictions: National, regional and global estimates

Abstract

Background: A growing literature has documented the high global morbidity, mortality and mental health burden associated with the current Covid-19 pandemic. In this paper, we aimed to quantify the total utility and quality of life loss resulting from Covid-19-related government restrictions imposed at the national, regional and global levels.

Methods: We conducted quality of life online surveys in France, India, Italy, UK and the United States of America between June 21st and September 13th 2021, and used regression models to estimate the average quality of life loss due to light and severe restrictions in these countries. We then combined estimated disutility weights from the pooled sample with the latest data on Covid-19 restrictions exposure in each country to estimate the total disutility generated by restrictions at the national, regional and global level. We also embedded a discrete choice experiment (DCE) into the online survey to estimate average willingness to pay to avoid specific restrictions.

Findings: A total of 947 surveys were completed. Thirty-five percent of respondents were female, and 69.5% were between 18 and 39 years old. The weighted average utility weight was 0.71 (95% CIs 0.69-0.74) for light restrictions, and 0.65 (0.63-0.68) for severe restrictions. At the global scale, this implies a total loss of 3259 million QALYs (95% 3021, 3496) as of September 6th, 2021, with the highest burden in lower and upper middle-income countries. Utility losses appear to be particularly large for closures of schools and daycares as well as restaurants and bars, and seem relatively small for wearing masks and travel restrictions.

Interpretation: The results presented here suggest that the QALY losses due to restrictions are substantial. Future mitigation strategies should try to balance potential reductions in disease transmission achievable through specific measures against their respective impact on quality of life. Additional research is needed to determine differences in restriction-specific disutilities across countries, and to determine optimal policy responses to similar future disease threats.

Funding: No funding was received for this project.

Keywords: Cost-effectiveness; Covid-19; QALY; Quality of life; Restrictions.

Figure 1

Mean utility weights for light restrictions, severe restrictions and paraplegia by country. Light restrictions: wearing masks in public spaces, restricted access to bars and restaurants, limited international travels. Severe restrictions: wearing masks in public spaces, restricted access to bars and restaurants, limited international travels. Mandatory home office, remote schooling and the inability to hold private meetings. Sampling weights were used to make the data representative of each country in terms of the overall distribution of age, sex, and educational attainment. 95% confidence intervals around estimated mean utility weights are shown on top of the bar charts.

Figure 2

Utility weights by age group and gender. Figure 2 shows estimated average utility weights by age (Panel A) and gender (Panel B). Age estimates are based on the weighted pooled sample and include both light and severe restrictions. Gender estimates were computed separately for each country. 95% confidence intervals around estimated mean utility weights are shown on top of the bar charts.

Figure 3

Months of light or severe restrictions by country between Jan 21, 2020 and Sept 6, 2021. Figure 3 shows the number of months with light or severe restrictions between January 21, 2020 and September 6, 2021. Data source: Oxford Covid-19 Governmental Response Index Tracker (https://github.com/OxCGRT/covid-policy-tracker). Stringency indices between 20 and 60 were coded as light restrictions; stringency indices > 60 were coded as severe restrictions.  Map created by authors using the World Bank's International Boundaries shapefile available at https:\\datacatalog.worldbank.org/search/dataset/0038272.

Figure 4

Estimated WTP per year (% of income) for avoiding specific restrictions. Figures show estimated population-weighted WTP for avoiding each restriction as a proportion of incomes. A relative income loss of -0.1 implies that on average respondents are willing to give up 10% of their incomes to avoid the specific measure. Estimates are based on random utility logistic regression. For France and Italy, the median monthly salary used was Euro 2000. For India, the UK and the US, median annual salaries used as reference point were RP 260,000, UKP 30,000 and USD 50,000, respectively. 95% confidence intervals around estimated mean willingness to pay are shown on top of the bar charts.

Figure 5

Respondent Approval of Government Measures. Figures show country-specific approval ratings of government actions taken in response to Covid-19. As part of the online survey, subjects were asked “What is your overall view on governmental pandemic restrictions during the past 18 months?” The first panel shows the results from the pooled sample, the remaining 5 panels show country-specific results.