Table 2.

Frequency of supra-therapeutic INR and hemorrhage among warfarin users by eGFR category

	eGFR>60 (n=819)	eGFR=45–59 (n=223)	eGFR=30–44 (n=115)	eGFR<30 (n=116)	P trend
INR ≥4
No. of INRs	20953	6027	3286	3359
INRs ≥4	1029 (4.9)	292 (4.8)	199 (6.1)	249 (7.4)	<0.001c
Major Hemorrhaged
No. of Events	68	22	18	29
Person-years	1160.5	331.5	169.7	140.1
Incidence rate	5.6 (4.6–7.4)	6.6 (4.3–9.9)	10.6 (6.5–16.4)	20.7 (14.3–29.3)
INR at event <4	51 (75.0)	19 (86.4)	11 (61.1)	10 (37.0)
INR at event ≥4	17 (25.0)	3 (13.6)	7 (38.9)	17 (63.0)	<0.001

eGFR,estimated glomerular filtation rate; INR: international normalized ratio

Note: All 1,273 warfarin pharmacogenetics cohort patients on warfarin therapy with target INR of 2–3. Incidence rate given as rate (95% confidence interval); INR given as number (percentage). eGFR categories expressed in ml/min/1.73m².

^bp is significant at α =0.05 and denote differences across kidney function categories

^cThe p-values were obtained by using the generalized estimating equation with the autoregressive lag-1 covariance structure to account for multiple INR measurements from the same patient as the density of the INRs differ across the patients during clinical care.

^d137 major hemorrhages (2 INRs not available); 91 occurred at INR <4 and 44 occurred at INR ≥4. INR not available for 2 events. Breslow-Day test for interaction of Rate Ratio over kidney function strata p=0.002. Breslow-Day test for interaction of Risk Difference over kidney function strata p=0.003.