Table 2.

Results from random-effects regression for COVID-19 case fatality ratio (log)*

Predictors	Coef.	Std. Err.	(95% CI)	P-value
Vaccine coverage	-0.076	0.025	(-0.126, -0.027)	0.002
Government effectiveness score	-0.076	0.014	(-0.102, -0.049)	<0.001
Transport infrastructure quality index	0.075	0.019	(0.038, 0.113)	<0.001
Population aged 65 or older (%)	0.054	0.024	(0.008, 0.100)	0.022
Hospital beds per 1000 population	0.001	0.038	(-0.073, 0.075)	0.981
Gross domestic product per capita (log)	-0.042	0.016	(-0.073, -0.010)	0.009
Government response stringency index	-0.010	0.014	(-0.037, 0.017)	0.478
Time to containment policy (weeks)	0.045	0.018	(0.010, 0.080)	0.012
Total tests for COVID-19 per 100 people	0.001	0.001	(0.000, 0.003)	0.045
Confirmed cases per 100 people last week	0.017	0.015	(-0.013, 0.047)	0.267

*Note: The sample had 90 countries and 2200 country-week observations. The standard errors were clustered at the country level. The R-squared value was 0.51. For convenience of interpretation, we multiplied the following variables by 10: government effectiveness score, infrastructure quality index, and gross domestic product per capita (log). Thus the corresponding coefficient should be interpreted on the basis of a 0.1 incremental increase in these variables. In addition, we divided the original value of government response stringency index by 10; thus its coefficient should be interpreted on the basis of a 10 incremental increase. The model included the continent in which a country is located and country random effects.