Objective

The objectives of this study were to describe inpatient anticoagulation and bridging in patients with non-valvular atrial fibrillation (NVAF) and to identify whether differences exist in length of stay (LOS) among bridged versus non-bridged NVAF patients.

Design

Administrative claims data were used to select patients ≥18 years with a primary or secondary discharge diagnosis of NVAF and inpatient warfarin use from 1 July 2004 to 30 September 2009. Patients with valvular or transient causes of NVAF or pregnancy were excluded. Inpatient bridging was defined as receipt of an anticoagulant in addition to warfarin during the hospitalization. LOS was reported for non-bridged patients (warfarin only) and compared with three bridging regimens: low molecular weight heparin/pentasaccharide (LMWH/PS); unfractionated heparin (UFH); and two-agent bridging (LMWH/PS and UFH). Multivariate analyses were performed to evaluate the association between bridging and LOS, adjusting for demographic and clinical variables.

Results

Of 6340 NVAF patients, 48% received inpatient warfarin (mean LOS 5.5 days); among them, 64% received bridging therapy (mean LOS 6.3 days) [LMWH/PS 45% (mean LOS 5.6 days), UFH 36% (mean LOS 6.0 days), two-agent bridging 18% (mean LOS 8.4 days)]. Following multivariate analysis, relative to patients who received inpatient warfarin only, LOS was significantly higher for patients with UFH (19.3%) and patients with two-agent bridging (45.1%). Patients with pre-period warfarin, cancer, or diabetes mellitus who received bridging agents had significantly longer LOS than patients with those conditions who were not bridged.

Conclusion

LOS was longer for bridged than non-bridged patients. Further studies are needed to identify predictors of bridging and to explain why bridged NVAF patients had longer LOS.