The costs of drug-eluting coronary stents among Medicare beneficiaries

Background Clinical trials predict the higher initial costs of drug-eluting coronary stents (DES) compared to bare metal stents (BMS) would be partially offset by decreased costs during the first year after percutaneous coronary intervention (PCI). However, the costs of DES in nonexperimental settings are not well understood. Methods We used a random, nationwide sample of Medicare beneficiaries who had received either DES from April through December 2003 or BMS from July 2002 through December 2003. Propensity score methods separately matched DES recipients (n = 4,375) to historical and contemporary BMS controls. Costs were estimated for each patient based on all physician and institutional claims beginning with the PCI hospitalization through the subsequent year, with institutional charges converted to costs via cost center-specific cost-to-charge ratios, as reported to Medicare by each institution. Results Drug-eluting stent patients had higher 30-day costs compared to both historical controls (mean difference $2,131, 95% CI $1,726 to $2,516) and contemporary controls ($1,882, 95% CI $1,480 to $2,322), but at 1 year, the DES-BMS mean cost differences had diminished substantially ($647, 95% CI $-385 to $1,664 compared to historical controls; $-84, 95% CI $-1,202 to $1,018 compared to contemporary controls) and were no longer statistically significant. Conclusions Despite higher initial costs of DES, the lower follow-up costs during the first year resulted in relatively small cumulative cost differences between DES and BMS recipients. These differences are comparable to those modeled in economic studies predicting acceptable cost-effectiveness for drug-eluting stents.