Table 3.

The significance (p-value) of the effect modifiers, seasonal effect and SCCS assumptions on the analyses, and the final model configurations utilised to calculate the incidence rate ratios for each thrombotic event.

	Arterial thrombosis	Cerebral venous thrombosis	Splanchnic thrombosis	Venous thromboembolism
Seasonal effect	0.010	0.897	0.003	<0.001
Dose dependence	0.512	0.346	0.650	0.049
Age effect modifier	0.077	0.319	0.137	0.470
Ethnicity effect modifier	0.879	0.295	0.504	0.639
Gender effect modifier	0.107	0.428	0.230	0.493
Event dependent exposure	<0.001	0.075	0.010	<0.001
Event dependent count	<0.001	0.067	<0.001	<0.001
Event dependent impact	0.160	–a	0.329	<0.001
Final model implemented	Event dependent exposure modelb	Standard SCCS model	Event dependent exposure model	Adaption to the event dependent exposure model for high event-related mortalityc

^aThere were too few counts to calculate this number.

^bEvent dependent exposure model developed by Farrington et al. [45], [46].

^cAdaption to the event dependent exposure model for high event-related mortality developed by Ghebremichael-Weldeselassie et al. [33].