Table 2. The results of subgroup analyses performed according to the potential sources of heterogeneity.
| Outcome | Form of disease | Geographical region | NOS score | Source of control | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| AML | MDS | Asia | Europe | Nor A | High | Medium | PB | HB | PB&HB | |
| Current smokers | ||||||||||
| No. of datasets | 10 | 5 | 2 | 6 | 7 | 11 | 4 | 8 | 5 | 2 |
| OR (95% CI) | 1.49 (1.31–1.69) | 1.52 (1.19–1.94) | 1.22 (0.85–1.74) | 1.31 (1.05–1.63) | 1.62 (1.41–1.86) | 1.47 (1.30–1.66) | 1.62 (1.23–2.14) | 1.44 (1.26–1.65) | 1.71 (1.37–2.13) | 1.32 (0.74–2.34) |
| P value | <0.001 | 0.001 | 0.29 | 0.02 | <0.001 | <0.001 | 0.001 | <0.001 | <0.001 | 0.35 |
| I 2 | 18% | 0% | 0% | 27% | 0% | 4% | 23% | 20% | 0% | 51% |
| Pub. bias | YES | YES | CBC | YES | YES | YES | NO | YES | NO | CBC |
| AOR (95% CI) | 1.43 (1.27–1.61) | 1.48 (1.17–1.86) | 1.09 (0.90–1.31) | 1.54 (1.36–1.75) | 1.43 (1.27–1.61) | 1.43 (1.25–1.63) | ||||
| Ever-smokers | ||||||||||
| No. of datasets | 16 | 7 | 4 | 9 | 10 | 15 | 8 | 13 | 8 | 2 |
| OR (95% CI) | 1.21 (1.12–1.31) | 1.32 (1.16–1.50) | 1.10 (0.95–1.28) | 1.24 (1.10–1.40) | 1.31 (1.19–1.44) | 1.29 (1.19–1.41) | 1.16 (0.94–1.43) | 1.30 (1.19–1.41) | 1.17 (0.97–1.41) | 1.16 (0.81–1.66) |
| P value | <0.001 | <0.001 | 0.21 | <0.001 | <0.001 | <0.001 | 0.01 | <0.001 | 0.10 | 0.429 |
| I 2 | 37% | 0% | 19% | 0% | 47% | 0% | 60% | 2% | 51% | 0% |
| Pub. bias | YES | NO | NO | NO | YES | NO | YES | NO | YES | CBC |
| AOR (95% CI) | 1.20 (1.11–1.30) | 1.29 (1.17–1.42) | 1.13 (1.01–1.26) | 1.06 (0.88–1.28) | ||||||
Abbreviations: No. of datasets, Number of datasets (papers); MDS, Myelodysplastic Syndrome; AML, Acute Myeloid Leukemia; Nor A, North America; NOS, Newcastle-Ottawa Scale; Pub. bias, Publication bias; CBC, cannot be calculated; OR, odds ratio; AOR, Adjusted OR; CI, Confidence interval; PB, Population-based; HB, Hospital-based.