Cost-effectiveness of targeting patients undergoing percutaneous coronary intervention for therapy with bivalirudin versus heparin monotherapy according to predicted risk of bleeding

Background-Although bivalirudin compared with unfractionated heparin with glycoprotein IIb/IIIa inhibitors reduces bleeding and hospitalization costs in patients undergoing percutaneous coronary intervention (PCI), little is known about the economic impact of bivalirudin versus heparin alone and at what threshold of procedural bleeding risk bivalirudin would be considered cost-effective. Methods and Results-A validated model was used to predict risk of major bleeding for 81 628 National Cardiovascular Data Registry (NCDR) CathPCI Registry patients from 2004 to 2006 who received unfractionated heparin only. Costs were derived from multiple sources including wholesale acquisition costs (for drugs) and single-center data (for PCI-related complications). Based on ISAR-REACT 3, we assumed that bivalirudin would reduce the risk of major bleeding by 33% compared with unfractionated heparin alone. A Markov model was used to estimate lost life expectancy associated with a major bleed. Major bleeding was predicted to occur in 2.2% of patients. Bivalirudin for all patients was estimated to increase costs by $571 per patient, yielding cost-effectiveness ratios of $287 473 per bleeding event averted and $1 173 360 per quality-adjusted life-year gained. Bivalirudin was cost saving for patients with a predicted bleeding risk >20% (0.16% of CathPCI population). At willingness-to-pay thresholds of $50K and $100K per quality-adjusted life-year gained, bivalirudin was cost-effective for patients with a bleeding risk ≥8% (2.5% patients) and ≥5% (7.9% patients), respectively. Conclusions-This decision-analytic modeling study demonstrates that for patients undergoing PCI, substitution of bivalirudin for unfractionated heparin monotherapy is projected to increase costs for virtually all patients and would be considered cost-effective for only a minority of patients with a high bleeding risk. From a policy standpoint, studies such as this, aimed at identifying the appropriate risk threshold for initiating treatment, may help in the development of informed guidelines for the use of expensive therapies.