Duke 2008 | ASHE | Equity and Efficiency in Health and Healthcare

Background. Since released into practice in 2003, drug-eluting stents (DES) have become a commonly used technology for percutaneous coronary intervention (PCI). Yet controversy exists regarding efficacy in general clinical practice, safety, and cost-effectiveness. We aim to assess the clinical and economic outcomes for 'all-comers' in a general interventional cardiology practice following the introduction of DES.

Methods. We identified all patients undergoing PCI from 2000- 2002 (pre-DES era) and from 2004 - 4/31/2006 (DES era) in a large institutional PCI registry. Data collected for these PCI episodes of interest included demographic, clinical, angiographic, procedural, and outcome information (in-hospital as well as follow-up adverse clinical events). Administrative data was used to estimate post-procedural length of stay (LOS) and associated total direct medical costs for PCI episodes (hospital and physician) in standardized, year 2007 constant-dollars. We used logistic and Cox proportional hazard models to estimate the risk of adverse events and generalized linear modeling to predict LOS and direct medical costs associated with procedural episode. Models estimated the incremental effect of era (pre-and post DES implementation) on outcomes adjusting for demographic, clinical, angiographic, and procedural characteristics. Propensity score and standard risk adjustments were used to account for these baseline differences.

Results. 3422 patients from the DES era (mean age 67, 69% male) were compared to 4303 patients from the pre-DES era (mean age 67, 70% male). Ninety percent of patients in the pre-DES era had bare-metal stents implanted, the rest atherectomy or balloon angioplasty only; whereas in the DES era 83% of patients had DES, 14% BMS, and 6% balloon or atherectomy only. Glycoprotein IIb/IIIa inhibitor use decreased from 65% pre-DES to 56% (p<0.001) and mean LOS decreased from 2.55 to 2.32 days (p=0.024). Clinical adverse event rates were similar between treatment eras (adjusted odds ratio for in-hospital myocardial infarction in DES era: 0.79; 95% CI 0.62, 1.00). During a median follow-up of 22 months (14 months in DES pts), the adjusted incidence of death or myocardial infarction was not different between cohorts, but clinically-indicated follow up procedures were significantly fewer in the DES era (hazard ratio for target vessel revascularization in DES era vs. pre-DES era: 0.58; 95% CI 0.50, 0.68). Models predict an adjusted mean LOS reduction of 0.41 days in the DES era and a total in-hospital cost savings of $-2,090 per procedural episode (95% bootstrapped CI of adjusted mean difference: -3,079, -1,121).

Conclusions. In a large unselected cohort of patients undergoing PCI at Mayo Clinic, the introduction of DES was associated with improved clinical outcomes during follow-up and overall reduced in-hospital costs. These data suggest costly new technologies can be introduced into a general practice setting while maintaining and improving patient outcomes at an incremental cost savings.