CRT: When is it economical to use DES?
| Slideshow | DES Economics: Insights for 2010 and Beyond |
| David J. Cohen, M.D., M.Sc. Director of Cardiovascular Research, Saint Luke's Mid America Heart Institute Professor of Medicine, University of Missouri-Kansas City  |
He cited data from the 2003 SIRIUS trial, prospectively comparing the economic data of Cordis’ Cypher sirolimus DES to a control group using BMS, finding a $309 difference at the end of one year. For the DES, the in-hospital costs were $11,345 and the follow-up costs were $5,468, totaling $16,813. In contrast, the control group in-hospital costs were $8,464 and the follow-up costs were $8,040, totaling $16,504.
However, for an additional cost-effectiveness analysis, Cohen noted that the cost for avoided repeat revascularization was assessed at $1,650, while the cost for quality-adjusted life years gained was assessed at $27,000. As a result, Cohen said that the “difference in cost was well worth the benefit in terms of the patients, and the way they were treated in the SIRIUS trial.”
Subsequent to SIRIUS, Cohen et al also used data from the Medicare program to model the relationship between BMS and its restenosis or target vessel revascularization (TVR) rate and associated cost. In 2003, the model assumptions—mainly based on SIRIUS—were that DES would cause a TVR risk reduction of 75 percent; the incremental DES cost was $2,000 per stent; and 1.4 stents were used per patient. They found that DES were reasonably cost effective if BMS had a TVR rate of 14 percent or greater.
However, he also pointed out that certain considerations have changed since the 2003 model assumptions. For instance, in 2009, DES typically cost about $1,000 compared to $2,000 six years ago; 1.6 stents are used in patients, as opposed to 1.4 in 2003; clopidogrel duration has now been extended by at least nine months to its current guideline expectation of one year; and the current TVR rate reduction has dropped to 50 percent from 75 percent in 2003.
Therefore, taking these 2009 changes to clinical practice, the “threshold where DES becomes cost effective is a little lower,” Cohen explains. More recently, DES are reasonably cost effective if the BMS TVR rate is 11 percent or greater.
Based on these data, he says that “DES still are not cost saving in 2009. But we have broadened the number of patients for whom DES are cost effective, largely because the cost of the stents has reduced substantially over the past six years.”
Focusing on patient outcomes, Cohen noted that DES are only preferable to BMS if the excess risk of late stent thrombosis is more than 0.2 percent per 1,000 per year. Out to five years, the acceptable risk of late stent thrombosis is 0.1 percent per 1,000 per year.
Applying a cost-effectiveness model to these stent thrombosis considerations if a thrombo-resistant DES existed, Cohen applied the model assumptions of background stent thrombosis rate (early + late + very late) of 1.5 percent; and the cost of stent thrombosis of $17,000 per event (Reynolds et al, 2002). He concluded that even at a cost premium of $400 per stent, a thrombo-resistant DES would be reasonably cost effective if it reduced total stent thrombosis by 25 percent.
“As restenosis rates have continued to decline, reductions in stent thrombosis and clopidogrel duration are now more attractive targets for improved DES design rather than further reductions in restenosis,” he said.
Despite “substantial changes in practice patterns over the past six years,” most of the fundamental insights regarding the cost effectiveness of DES have changed very little, explained Cohen. For example:
- DES are cost saving for only a minority of the PCI population; and
- DES appear reasonably cost effective compared with BMS for a substantial fraction of PCI patients (predicted BMS TVR greater than 10-11 percent).
“Given current clinical data, the most promising targets for improving the cost effectiveness of DES relate to reductions in stent thrombosis or mandatory dual-antiplatelet duration,” Cohen concluded.