Abstract 314: Impact of Spontaneous Bleeding Events on Health State Utility in Patients with Atrial Fibrillation: Results from the ENGAGE AF-TIMI 48 Trial
BACKGROUND: Anticoagulation in AF patients involves consideration of the benefit of stroke risk reduction versus increased bleeding risk. The impact of bleeding events on health state utility has not been clearly understood.
METHODS: We used prospectively-collected data from the ENGAGE AF-TIMI 48 Trial to estimate the impact of different bleeding events on utility scores derived from the EQ-5D. EQ-5D data were collected at randomization and 3 month intervals for up to 48 months; at least two EQ-5D assessments were available for 15,618 participants. Spontaneous bleeding events were identified prospectively and adjudicated according to prespecified criteria. Patients who experienced a bleed and provided at least one EQ-5D assessment before and after the event were included in the analysis. To maximize specificity, patients were excluded if they had more than 1 bleeding event or any other major cardiovascular events in addition to bleeding. Longitudinal growth curve models were used to estimate changes in utilities after the following mutually exclusive bleeding event categories: intracranial hemorrhage (ICH; n=25); major GI (n=294); major non-GI (n=236); clinically relevant non-major (CRNM) (n=1,414); and minor (n=783).
RESULTS: There were reductions in EQ-5D utility after the event for all categories of bleeding that were in parallel with the severity of bleeding (Table). The estimated decrease in utility was greatest for ICH (-0.079), followed by major non-GI (–0.045), GI (–0.030) and CRNM (–0.009). In addition, a significant linear improvement over time was observed for major non-GI bleeds (and a similar albeit non significant improvement for major GI bleeds) such that EQ-5D scores returned to near pre-event levels one year after the bleeding event. No post-event improvement was observed for ICH, CRNM and minor bleeds, however.
CONCLUSION: Spontaneous bleeding events are associated with a negative impact on health state utility, with greatest immediate impact for major bleeds. Health state utility remains low 1 year after an ICH, but gradually returns to baseline 1-year after other major bleeding events. Future cost-effectiveness analysis of strategies to prevent stroke in AF should consider both the economic and health status impact of bleeding events to inform health care decision making.
Author Disclosures: E.A. Magnuson: B. Research Grant; Significant; Daiichi Sankyo, Eli Lilly, Astra Zeneca, Edwards Lifesciences, Medtronic. E. Honoraria; Modest; Astra Zeneca. K. Wang: None. H. Li: None. W.J. Kwong: A. Employment; Significant; employed by Daiichi Sankyo. E.M. Antman: B. Research Grant; Significant; Daiichi Sankyo. C.T. Ruff: G. Consultant/Advisory Board; Modest; Daiichi Sankyo, Bristol-Myers Squibb, Boehringer Ingelheim. S.A. Murphy: B. Research Grant; Significant; Daiichi Sankyo. R.P. Giugliano: B. Research Grant; Significant; Daiichi Sankyo, Merck, Johnson & Johnson, Sanofi Aventis, Astra Zeneca. D. Speakers; Modest; Bristol-Myers Squibb, Daiichi Sankyo, Merck, Sanofi Aventis. G. Consultant/Advisory Board; Modest; Daiichi Sankyo, Janssen Pharmaceuticals, Merck. D.J. Cohen: B. Research Grant; Significant; Astra Zeneca, Daiichi Sankyo, Eli Lilly, Boston Scientific, Edwards Lifesciences, Medtronic, Biomet. G. Consultant/Advisory Board; Modest; Eli Lilly, Medtronic, Abbott Vascular. G. Consultant/Advisory Board; Significant; Astra Zeneca.
- © 2014 by American Heart Association, Inc.