TABLE 2.
Subgroup analysis performed according to patient and study characteristics considered as potential sources of heterogeneity for outcomes.
| Outcome | Subgroups (No. studies) | HR (95% CI) | I2, % | P-value |
| Ischemic | Hospital control (4) | 1.29 (1.11, 1.50) | 95 | 0.0008 |
| stroke | Population control (3) | 1.94 (1.02, 3.69) | 96 | 0.04 |
| HR (4)1 | 1.19 (1.15, 1.22) | 0 | <0.00001 | |
| OR (3) | 2.18 (1.54, 3.08) | 89 | <0.0001 | |
| Hemorrhagic | Hospital control (1) | 1.28 (1.16, 1.41) | N/A | <0.00001 |
| stroke | Population control (3) | 2.74 (1.49, 5.03) | 89 | 0.001 |
| HR (2) | 1.48 (1.05, 2.08) | 79 | 0.02 | |
| OR (2)1 | 3.87 (3.26, 4.60) | 0 | <0.00001 |
1Based on fixed effect meta-analysis. N/A, not applicable; CI, confidence interval; HR, hazard ratio.