Table 2 Final multivariate model for global COVID-19 mortality adjusted for response measures.

From: Global ecological analysis of COVID-19 mortality and comparison between “the East” and “the West”

 

Final model

Correction for spatial autocorrelation

b (95% CI) p-value

b (95% CI) p-value

Constant

1.963 (1.871–2.055)

p < 0.001

1.963 (1.869–2.057)

p < 0.001

Median age (years)

0.370 (0.238–0.501)

p < 0.001

0.370 (0.145–0.594)

p = 0.004

Obesity prevalence (%)

0.358 (0.260–0.455)

p < 0.001

0.358 (0.189–0.526)

p < 0.001

Previous cases of SARS (‘Yes’)

−0.311 (−0.552–−0.069)

p = 0.012

−0.311 (−0.712 to 0.090)

p = 0.118

EIU democracy index (0–10)

0.353 (0.232 to 0.475)

p < 0.001

0.353 (0.147 to 0.560)

p = 0.003

Political Stability and Absence of Violence (%)

−0.297 (−0.428 to −0.166)

p < 0.001

−0.297 (−0.481 to −0.113)

p = 0.004

Average stringency index (%)

0.160 (0.052 to 0.268)

p = 0.004

0.160 (0.035 to 0.285)

p = 0.016

Timeliness of stay−at−home requirements (days)

−0.035 (−0.129 to 0.059)

p = 0.458

−0.035 (−0.155 to 0.085)

p = 0.537

Model statistics

Observations

133

R2

0.686

Adjusted R2

0.669

Residual SE

0.476

F statistic (df)

39.05 (7, 125)

123.60 (7, 13)

p−value

p < 0.001

AIC

187.79

Sample−size adj. BIC

210.91

  1. In the final model, three observation (ie, Mongolia, Thailand, and Papua New Guinea) was removed because outlier, as described in the supplementary results. Regions are controlled as sampling clusters and the clustering effect was accounted for using cluster-robust standard errors (the clustered sandwich estimator) .
  2. EIU Economist intelligence unit; SARS Severe acute respiratory syndrome.