Long Term Outcomes of Carotid Endarterectomy in the Elderly

描述(申请人提供)：颈动脉内膜切除术(CEA)，预防中风的手术，是常见的，但潜在的风险。随机对照试验(RCT)显示，与药物治疗相比，在精心挑选的患者(PT)和外科医生中，CEA降低了死亡和中风的风险。然而，有症状的PTS(有中风或短暂性脑缺血发作)需要存活2-3年才能实现CEA的好处，而无症状的患者需要存活5年才能受益。由于在社区实践中接受CEA的PT比RCT中的PTS年龄更大、病情更重，预期寿命更短，CEA对这些PT的好处可能更有限。CEA的潜在益处在无症状PTT中最为明显，他们占接受该手术者的四分之三。尽管CEA被广泛使用，但CEA在社区实践中的长期结果以及CEA与老年人药物治疗的相对好处尚不清楚。本申请的具体目的是：1)通过比较社区实践中使用基于人群的纽约颈动脉手术(NYCAS)的Medicare队列研究进行的CEA的长期结果与RCT中报道的那些，来确定CEA RCT的普适性；2)使用两种方法评估外科(CEA)与在社区实践中对老年人颈动脉疾病的医疗管理的有效性：2A)比较在NYCAS接受CEA的Medicare PTS与基于人群的心血管健康研究(CHS)中有颈动脉疾病的类似PTS的长期结果；2)比较接受手术(CEA)和内科治疗的CHS患者(CEA)和接受内科治疗的老年人(CEA)的长期预后；3)根据年龄、合并症、RCT资格、性别、种族/族裔和症状状态，分析CEA在老年人中的长期预后差异，并建立和验证用于预测预后的多元临床模型；以及4)确定影响经风险调整、卒中相关生存率的独立护理过程、外科医生和医院因素，以提供最佳实践和政策参考。NYCAS有详细的PT、临床、流程、外科医生和医院数据。4000名NYCAS PT的CEA后5年和10年内的死亡和中风将使用国家死亡指数进行评估，并对Medicare Part A Claims确定的中风入院情况进行图表审查。CHS有详细的临床基线和长期结果数据。NYCAS的结果将与随机对照试验进行比较。418例患有颈动脉疾病的CHS患者接受药物治疗的长期结果将与1672例类似的NYCAS CEA患者进行配对分析(1：4)进行比较(“两项研究之间”的分析)。在一项“在研究范围内”的分析中，69名患有CEA的CHS PT将与418名使用倾向评分进行医疗管理的CHS PT进行比较。多变量回归将被用来确定在NYCAS(N=4000)的长期结果的PT、护理过程、外科医生和医院预测因素。

DESCRIPTION (provided by applicant): Carotid endarterectomy (CEA), surgery to prevent stroke, is common but potentially risky. Randomized controlled trials (RCTs) showed that among carefully selected patients (Pts) and surgeons, CEA reduced the risk of death and stroke compared to medical therapy. However, symptomatic Pts (with strokes or transient ischemic attacks) needed to survive for 2-3 years to realize the benefits of CEA, and asymptomatic patients needed to survive for 5 years to benefit. Because Pts undergoing CEA in community practice are older and sicker than those in the RCTs and have shorter life expectancy, the benefits of CEA for these Pts may be more limited. The potentially reduced benefit of CEA would be most marked among asymptomatic Pts who comprise three-quarters of those having the procedure. Despite its, widespread use, the long term outcomes of CEA in community practice and the relative benefits of CEA v. medical therapy in the elderly is unknown. The specific aims of this application are to: 1) Determine the generalizability of the CEA RCTs by comparing the long term outcomes of CEA performed in community practice using the population-based, New York Carotid Artery Surgery (NYCAS) Medicare cohort study to those reported in the RCTs; 2) Assess the effectiveness of surgical (CEA) v. medical management of carotid disease among the elderly in community practice using 2 methods: 2A) Comparing long term outcomes of Medicare Pts who underwent CEA in NYCAS to similar Pts with carotid disease in the population-based Cardiovascular Health Study (CHS) who did not have surgery; and 2B) Comparing long term outcomes of elders with carotid disease in CHS who were surgically (CEA) v. medically managed using propensity score methods; 3) Examine differences in long term outcomes of CEA in the elderly according to: age, comorbidities, RCT eligibility, gender, race/ethnicity, and symptom status, and develop and validate a multivariate clinical model to predict outcomes; and 4) Identify the independent process of care, surgeon, and hospital factors that influence risk-adjusted, stroke- specific survival to inform best practices and policies. NYCAS has detailed Pt, clinical, process, surgeon, and hospital data. Deaths and strokes 5 and 10 years after CEA for 4000 NYCAS Pts will be assessed using the National Death Index and performing chart review of stroke admissions identified by Medicare Part A claims. CHS has detailed clinical baseline and long term outcomes data. Outcomes in NYCAS will be compared to the RCTs. Long term outcomes of 418 CHS Pts with carotid disease who were medically managed will be compared(1:4) in matched analysis to 1672 similar NYCAS CEA Pts ('Between Study' analysis). In a 'Within Study' analysis, 69 CHS Pts who had CEA will be compared to 418 who were medically managed using propensity scores. Multivariate regression will be used to identify Pt, process of care, surgeon, and hospital predictors of long term outcomes in NYCAS (N=4000).