Introduction: Schemas to identify bleeding-related hospitalizations in claims data differ in both the billing codes used and coding positions allowed.Objective: To assess agreement among commonly used bleeding-related hospitalization coding schemas for claims analyses of nonvalvular atrial fibrillation (NVAF) patients.Methods: Within MarketScan claims data, we identified adults with NVAF newly started on an oral anticoagulant (OAC) from 1/2012-6/2015. Billing code schemas developed by Cunningham et al., Yao et al. and the US Food and Drug Administration (FDA) Mini-Sentinel program were used to identify bleeding-related hospitalizations as a surrogate for major bleeding. Bleeds were subcategorized as intracranial hemorrhage (ICH), gastrointestinal (GI) or other. Schema agreement was assessed by comparing incidence, rates of events/100 person years (PYs) and Cohen’s κ statistic.Results: We identified 151,738 new users of OACs with NVAF (median; 25%, 75% range CHA2DS2-VASc score=3 (2, 4) and HASBLED score=3 (2, 3)). The Cunningham, Yao and FDA schemas identified bleeding-related hospitalizations in 2,845 (1.9%), 7,065 (4.7%) and 4,027 (2.7%) patients (corresponding rates=3.45, 8.65 and 4.90 events/100-PYs). Agreement across all schemas was weak to moderate (κ=0.47-0.66) for total bleeding (Figure). Near perfect agreement (κ=0.99) was observed with the Yao and FDA schemas for ICH-related hospitalizations; but agreement was weak when comparing Cunningham to the Yao or FDA schemas (κ=0.52-0.53). Yao and FDA schema agreement was moderate (κ=0.62) for GI bleeding, but weak upon comparing Cunningham to the Yao or FDA schemas (κ=0.44-0.56). For other bleeds, agreement across all schemas was minimal at best (κ=o.14-0.38).Conclusions: There appears to be varying levels of agreement among the 3 frequently used billing code schemas for identifying bleeding-related hospitalizations in NVAF patients.