Table 7.
Random-effects meta-regression analysis of the crude mortality risk at the last follow-up date in the extended analysis (February 12, 2021) for 73 countries.
| MODEL 1: a priori model | MODEL 2: bootstrap variable selection | |||||||
|---|---|---|---|---|---|---|---|---|
| Mortality risk (per 1,000 population) | Mortality risk (per 1,000 population) | |||||||
| Covariates | β | SE | P-value | OR | β | SE | P-value | OR |
| Intercept | −5.5394 | 4.3667 | – | – | −4.8819 | 3.0459 | – | – |
| Sociodemographic factors | ||||||||
| GDP per capita ($1,000 USD, 2019) | −0.0227 | 0.0155 | 0.147 | 0.98 | −0.0215 | 0.0087 | 0.016 | 0.98 |
| Urban population (%) | – | – | – | – | −0.0327 | 0.0095 | 0.001 | 0.97 |
| Population density (pop per km2) | ||||||||
| Elderly dependency ratio (% of adults) | – | – | – | – | – | – | – | – |
| Proportion over 65 years (%) | −0.0078 | 0.0525 | 0.882 | 0.99 | −0.1930 | 0.0747 | 0.012 | 0.82 |
| Proportion over 80 years (%) | – | – | – | – | 0.4775 | 0.2220 | 0.036 | 1.61 |
| Proportion overweight (%) | – | – | – | – | 0.0676 | 0.0140 | <0.0001 | 1.07 |
| Pandemic–related factors | ||||||||
| Time since 1st case (days) | – | – | – | – | −0.0323 | 0.0079 | <0.0001 | 0.97 |
| Time since 100 cases (days) | – | – | – | – | 0.0324 | 0.0065 | <0.0001 | 1.03 |
| Time since 1st death (days) | −0.0006 | 0.0128 | 0.961 | 1.00 | – | – | – | – |
| Testing coverage (n. tests per 10,000 pop) | 0.0080 | 0.0048 | 0.100 | 1.01 | 0.0100 | 0.0032 | 0.003 | 1.01 |
| Health system strength | ||||||||
| Healthcare workers (n. per 1,000 pop) | 0.0523 | 0.0636 | 0.414 | 1.05 | – | – | – | – |
| Hospital beds (n. per 1,000 pop) | −0.1164 | 0.0992 | 0.245 | 0.89 | – | – | – | – |
| Health expenditure (% of GDP) | – | – | – | – | 0.2644 | 0.0649 | <0.0001 | 1.30 |
| Out–of–pocket health expenditure (%) | – | – | – | – | – | – | – | – |
| Cultural characteristics | ||||||||
| Individualism vs. collectivism | 0.0273 | 0.0114 | 0.020 | 1.03 | – | – | – | – |
| Uncertainty avoidance | 0.0421 | 0.0098 | <0.0001 | 1.04 | – | – | – | – |
| Indulgence vs. restraint | – | – | – | – | – | – | – | – |
| Long–term vs. short–term orientation | – | – | – | – | 0.0186 | 0.0073 | 0.014 | 1.02 |
| Power distance | – | – | – | – | 0.0113 | 0.0075 | 0.139 | 1.01 |
| Masculinity vs. femininity | – | – | – | – | – | – | – | – |
| Political characteristics | ||||||||
| Polity (democracy vs. authoritarianism) | 0.1022 | 0.0437 | 0.023 | 1.11 | 0.1122 | 0.0307 | 0.001 | 1.12 |
| pseudo R2: 41% | pseudo R2: 74% | |||||||
| AIC:246.6 BIC: 270.6 | AIC:195 BIC:224 | |||||||
Dependent variables were log transformed rates. Random-effects meta-regression was used to explore the impact of cultural characteristics on fatalities while adjusting for important predefined covariates. Pseudo-R-squared value represent the proportion of heterogeneity explained by predictors included in the model. OR, Odds ratio; AIC, Akaike information criterion; BIC, Bayesian information criterion. Bold font indicates a statistically significant association with outcome at p < 0.05.