Classification of 23 selected studies according to their methodologic design
| Class of study . | No. of studies . | Cohorts . | Patients . | Events . | IR for thrombosis, % (95% CI) . | Homogeneity test, P . |
|---|---|---|---|---|---|---|
| Retrospective6,59-67 | 10 | 17 | 16 861 | 1069 | 6.3 (6.0-6.7) | < .001 |
| Prospective68-75 | 8 | 12 | 1157 | 80 | 6.9 (5.5-8.6) | .034 |
| Registries*4,76-79 | 5 | 7 | 3 313 471 | 83 016 | 2.5 (2.4-2.5) | < .001 |
| Class of study . | No. of studies . | Cohorts . | Patients . | Events . | IR for thrombosis, % (95% CI) . | Homogeneity test, P . |
|---|---|---|---|---|---|---|
| Retrospective6,59-67 | 10 | 17 | 16 861 | 1069 | 6.3 (6.0-6.7) | < .001 |
| Prospective68-75 | 8 | 12 | 1157 | 80 | 6.9 (5.5-8.6) | .034 |
| Registries*4,76-79 | 5 | 7 | 3 313 471 | 83 016 | 2.5 (2.4-2.5) | < .001 |
95% CI indicates 95% confidence interval; and IR, incidence rate.
Studies determined on the basis of record linkage of epidemiologic registries and/or disease/hospitalization databases.