Table 3.

Persistence rates of OACs among OAC naïve patients

	Apixaban		Rivaroxaban		Dabigatran		VKA
	N=413		N=1275		N=656		N=9303
Persistence of index medication by the end of follow-up (n, %)
Persistent	364	88.1%	999	78.4%	461	70.3%	7461	80.2%
Non-persistent	49	11.9%	276	21.7%	195	29.7%	1842	19.8%
Type of change in therapy (n, %)	N non-persistent=49		N non-persistent=276		N non-persistent=195		N non-persistent=1842
Discontinued therapy	31	63.3%	212	76.8%	126	64.6%	1367	74.2%
Switched therapy	18	36.7%	64	23.2%	69	35.4%	475	25.8%
Persistence at different time points
At 3 months
Percentage*, 95% CI	91.5	88.5 to 94.0	86.3	84.3 to 88.1	84.1	81.2 to 86.8	93.9	93.4 to 94.4
Number of patients at risk†	297		945		496		8121
Number of patients censored‡	82		158		57		630
At 6 months
Percentage*, 95% CI	88.2	84.5 to 91.3	80.7	78.4 to 83.0	74.2	70.6 to 77.6	87.0	86.2 to 87.7
Number of patients at risk†	150		603		351		6189
Number of patients censored‡	220		448		148		2005
At 12 months
Percentage*, 95% CI	82.8	76.8 to 87.9	73.1	70.0 to 76.2	66.7	62.6 to 70.7	77.8	76.8 to 78.7
Number of patients at risk†	21		227		189		3321
Number of patients censored‡	343		782		279		4324
At end of follow-up
Percentage*, 95% CI	82.8	76.8 to 87.9	67.6	62.9 to 72.2	62.5	57.5 to 67.6	70.6	68.9 to 72.3
Number of patients at risk†	0		0		0		0
Number of patients censored‡	364		999		461		7461

*100% minus the cumulative incidence of non-persistence, accounting for the competing risk of death.

†Number of patients at who remain ‘at risk’ of non-persistence after the specified time point.

‡Number of patients censored before the specified time point due to non-persistence, leaving the database, end of the study period or death.

OAC, oral anticoagulants.