TABLE 2.
Summary effects and heterogeneity obtained from the meta‐analysis of the association of diabetes with clinical outcomes in hospitalized COVID‐19 patients
| Outcome | Summary effects | |||||||
|---|---|---|---|---|---|---|---|---|
| REDL | Heterogeneity a | Heterogeneity variance estimates | ||||||
| Effect measure | OR (95% CI) | Tests of overall effect | Cochran's Q | H | I2 b | P value | tau≤ c | |
| COVID‐19 severity | OR | 3.39 (2.14‐5.37) | P < .0001, z = 5.206 | 40.21 | 2.11 (95% CI, 1.48‐2.71) | 77.6% (95% CI, 54.2%‐86.4%) | P < .0001 | 0.3706 |
| ARDS | OR | 2.55 (1.74‐3.75) | P < .0001, z = 4.772 | 11.05 | 1.49 (95% CI, 1.00‐2.42) | 54.7% (95% CI, 0%‐79.9%) | P = .05 | 0.1200 |
| Mortality | OR | 2.44 (1.93‐3.09) | P < .0001, z = 7.425 | 48.47 | 1.48 (95% CI, 1.111‐1.845) | 54.6% (95% CI, 19%‐70.6%) | P = .001 | 0.1586 |
| Need for mechanical ventilation | OR | 3.03 (2.17‐4.22) | P < .0001, z = 6.524 | 29.93 | 1.58 (95% CI, 1.06‐2.07) | 59.9% (95% CI, 11.3%‐76.7%) | P = .003 | 0.1974 |
Abbreviations: ARDS, acute respiratory distress syndrome; H, relative excess in Cochran's Q over its degrees of freedom; I2, proportion of total variation in effect estimate due to between‐study heterogeneity (based on Cochran Q); OR, odds ratio; Q, heterogeneity measures were calculated from the data with CI based on noncentral chi‐square (common effect) distribution for Cochran Q test; REDL, DerSimonian‐Laird random‐effects method.
Heterogeneity measures were calculated from the data with 95% CI based on gamma (random effects) distribution for Q.
Values of l2 are percentages.
Heterogeneity variance estimates (tau≤) were derived from REDL.