0–90 day risk of a first VTE among patients with first hospital admission for CAB and matched hospitalised controls and population controls, Northern Denmark, 1992–2010.

Abbreviations: CI, confidence interval.
1Computed by conditional logistic regression. Controls matched for age, sex and calendar time act as reference group. All odds ratio estimates are controlled for matching factors. Estimates for “All study subjects” are adjusted for cancer, cardiovascular disease, other comorbidity, and recent hospital contact. Subgroup analyses are adjusted for any comorbidity.
2Not adjusted for comorbidity due to few events. A (−) denotes that the odds ratio could not be calculated because no events occurred among population controls.