TABLE 2.
Pooled effect for cohort and cross-sectional studies, and subgroup analysis for cross-sectional studies by population group, study location, sample size, proportion of male, study quality.
| No. studies | Odds ratio (95% CI) |
I 2 | P-value for heterogeneity | P-value for subgroup differences | Weight (%) | |
| Study design | ||||||
| Cohort studies | 3 | 1.62 (1.02, 2.56)* | <0.1% | 0.795 | . | |
| Cross-sectional studies | 17 | 1.55 (1.30, 1.84) | 89.5% | <0.001 | . | |
| Population group | < 0.001 | |||||
| Community residents | 5 | 1.77 (1.16, 2.70) | 87.9% | <0.001 | 30.70 | |
| Contacts | 3 | 1.12 (0.88, 1.43) | 35.2% | 0.214 | 18.33 | |
| Refugees | 2 | 2.88 (1.93, 4.29) | 39.4% | 0.199 | 11.48 | |
| Elderly/patients with other chronic diseases | 5 | 1.13 (0.95, 1.35) | 61.7% | 0.034 | 35.24 | |
| Health care worker | 2 | 5.27 (1.52, 18.20) | 63.1% | 0.100 | 4.25 | |
| Study location | <0.001 | |||||
| Europe | 1 | 1.22 (1.04, 1.43) | 8.37 | |||
| Asia | 8 | 1.33 (1.03, 1.72) | 71.1% | 0.001 | 44.49 | |
| North America | 7 | 1.69 (1.25, 2.29) | 93.8% | <0.001 | 44.56 | |
| Africa | 1 | 9.33 (3.74, 23.28) | 2.57 | |||
| Sample size | 0.001 | |||||
| 100–1,000 | 6 | 1.41 (0.78, 2.55) | 82.8% | <0.001 | 24.89 | |
| 1,000–5,000 | 8 | 1.99 (1.49, 2.66) | 83.3% | <0.001 | 49.11 | |
| >5,000 | 3 | 1.11 (1.03, 1.20) | 47.2% | 0.151 | 26.00 | |
| Proportion of male | 0.010 | |||||
| <40% | 4 | 2.36 (1.19, 4.66) | 82.6% | 0.001 | 17.76 | |
| 40–50% | 7 | 1.42 (0.99, 2.05) | 91.1% | <0.001 | 42.51 | |
| ≥ 50% | 5 | 1.63 (1.03, 2.57) | 88.8% | <0.001 | 31.12 | |
| Unknown | 1 | 1.01 (0.89, 1.14) | 8.61 | |||
| Risk of bias | 0.658 | |||||
| Low | 12 | 1.53 (1.25, 1.86) | 91.1% | <0.001 | 80.67 | |
| Moderate | 5 | 1.78 (0.93, 3.42) | 80.7% | <0.001 | 19.33 |
*This effect is Risk Ratio.