Background: Stroke is the third leading cause of death and the leading cause of disability in the United States. Carotid stenosis (CS) is responsible for approximately 30% of these strokes. Carotid endarterectomy (CEA) is a well-established procedure to treat CS shown to be efficacious. An emerging management, carotid artery stenting (CAS) is considered a less invasive alternative to CEA and shown to be equally efficacious in patients with high-risk for surgery in one clinical trial directly comparing CAS to CEA. However, many methodological issues in the trial made the results controversial. Also, it was not clear whether high-risk patients can be accurately identified and whether high-risk patients have higher complication rates after CEA.
Aim 1: Demonstrate differences in the overall incidence of death and stroke between trial populations and Medicare patients undergoing CEA/CAS.
Aim 2: Identify subgroups and predictors for high risk after CAS.
Aim 3: Compare outcomes after CAS to CEA in high-risk patients.
Aim 4: Assess if the effectiveness of CAS versus CEA differ among subgroups.
Aim 5: Assess the impact of unmeasured confounding in a well-designed comparative effectiveness study comparing CAS/CEA versus medical management.
Aim 6: Investigate the potential of linking imaging data with medical record and claims data and write a white paper.
Study Design: Retrospective cohort studies linking two registries and Medicare data.
Data Source: 1) The Center for Medicare and Medicaid Services (CMS) CAS registry; 2) the Society of Vascular Surgery (SVS-VR) data; 3) Medicare claims data; and 4) imaging/ clinical data from Brigham and Women’s Hospital (BWH) and Beth Israel Deaconess Medical Center (BIDMC) vascular labs.
Participants, setting and expected sample size: Study population will consist of: 1) 3,300 CAS patients and 5,200 CEA patients included in the SVS-VR from 2005-2008; 2) 7,000 CAS patients included in the CMS-CAS registry from 2005-2008; 3) all Medicare patients who had CAS or CEA; and 4) patients who had carotid imaging tests at BWH or BIDMC. We will further require that said patients be: a) greater than 65 years of age at the time of CAS/CEA or medications and b) eligible for Medicare at least 12 months prior to the date of the index procedure (to allow covariate assessment during the 12 months prior to the procedure). In addition, greater sample size in all Medicare CAS versus CEA data will be obtained using propensity score calibration (PSC).
Exposures: CAS, CEA, medical treatment
Outcome: Study outcomes include all-death as well as hospitalization for stroke and myocardial infarction. To define these events, we will use previously validated definitions with very high specificity.
Statistical methods: Survival analyses and logistic regression will be used to describe and predict outcomes among subgroups. We will use Cox regression to compare CAS and CEA after adjusting for confounders using high dimensional propensity score (hd-PS) or multiple imputation.
Comparator groups: CAS, CEA, medical treatment
Expected Outputs: Scientific Reports
Expected Date of Project Completion: 9/30/2011