Abstract 169: Comparative Effectiveness of Surgical v. Medical Management of Carotid Stenosis in the Elderly: The Cardiovascular Health Study

Abstract

Introduction: In randomized clinical trials(RCTs), carotid endarterectomy(CEA) reduced risk of death and stroke(D&S)compared to medical therapy(MedRx) in carefully selected patients and surgeons. We hypothesized the benefits of CEA among the elderly in real world practice would be more limited than in RCTs because patients are older and sicker, have shorter life expectancy, and have less experienced surgeons. Methods: We conducted a comparative effectiveness study of CEA v MedRx for carotid stenosis in the elderly using data from the Cardiovascular Health Study(CHS), a multicenter, community-based, prospective observational cohort study of participants(Pts) randomly selected from Medicare lists. CHS Pts had intensive evaluation of vascular events, risk factors, & comorbidities. Carotid stenosis was measured via ultrasound(US) at baseline, 5 and 10 years. D&S were ascertained though annual contact and all events were adjudicated. CEA was identified via annual contact and Medicare data. We assessed risk of D&S using Cox proportional hazards regression and propensity scores(PS) to adjust for factors associated with selection for CEA and risk of D&S. We focused on 487 Pts with ≥50% carotid stenosis by US and no prior CEA. Results: Overall, 89 Pts had CEA, and 398 MedRx during a median 12.1 years of follow-up. Half were men and 91% white. At baseline, mean age was 73.4 years, 76% had hypertension (HTN), 37% diabetes, 28% coronary heart disease(CHD), 7% prior stroke, and 8% TIA. CEA Pts were older and had more HTN, heart failure, atrial fibrillation, prior stroke, TIA, or CHD and worse self-rated health(v MedRx,all p<.05). In unadjusted analyses, rate of D&S was higher for CEA v MedRx at all time points: 30 days(7.8% v 0.2%), 5 years(39% v 27%), and 10 years(60% v 48%; Hazard Ratio(HR)=1.8; 95%CI: 1.3-2.3). The HR for CEA was unchanged(HR=1.8;CI: 1.2-2.6) with adjustment for PS, neurologic symptoms,and 13 other time varying covariates. Analyses assessing impact of CEA v MedRx on the adjusted, long term hazard of stroke alone also found worse outcomes for CEA(p<.05). Conclusions: Elderly CHS Pts with carotid stenosis managed with CEA (v MedRx) had higher long term adjusted risk of D&S and stroke alone. The benefits of CEA shown in younger, healthier patients in RCTs did not extend to the older, sicker CHS Pts, likely better reflecting the impact of CEA in real world practice. CEA in the elderly may result in worse outcomes compared to MedRx.