Table 4.

Crude incidence rates per 100 patient years and HRs of major adverse cardiovascular events stratified by Janus kinase inhibitors

	Overall patients					Cardiovascular risk patients
	Number of treatment episodes	Patient years of follow-up	Number of events	Incidence rate (95% CI)	HR* (95% CI)	Number of treatment episodes	Patient years of follow-up	Number of events	Incidence rate (95% CI)	HR* (95% CI)
Baricitinib	1416	2460.17	12	0.49 (0.25 to 0.85)	0.58 (0.26 to 1.26)	862	1543.25	11	0.71 (0.36 to 1.28)	0.54 (0.24 to 1.22)
Tofacitinib	1126	1836.92	18	0.98 (0.58 to 1.55)	1.36 (0.70 to 2.63)	671	1145.75	14	1.22 (0.67 to 2.05)	1.04 (0.48 to 2.23)
Upadacitinib	768	750.67	4	0.53 (0.15 to 1.36)	NR	459	469.50	4	0.85 (0.23 to 2.18)	NR

*HR of major adverse cardiovascular events were estimated from the Andersen-Gill model applying inverse probability weighting. In all Andersen-Gill models, the number of previous DMARD substances was used as covariate in addition to the treatment effect due to exceeding the mean standardised difference threshold. Tumour necrosis factor inhibitors were selected as reference category.

DMARD, biologic disease-modifying antirheumatic drug; NR, not reported due to a low number of events.