Purpose: To evaluate the evidence on screening and treating asymptomatic adults for carotid artery stenosis (CAS) for the U.S. Preventive Services Task Force (USPSTF).
Data Sources: PubMed/MEDLINE, the Cochrane Library, EMBASE, and trial registries through September 2013; reference lists of published literature; MEDLINE searches for trials were updated through March 2014.
Study Selection: Two investigators independently selected studies reporting on asymptomatic adults with CAS, including randomized, controlled trials (RCTs) of screening for CAS; RCTs of carotid endarterectomy (CEA) or carotid angioplasty and stenting (CAAS) versus medical treatment; RCTs of medications versus placebo added to current standard medical therapy; multi-institution trials or cohort studies reporting harms; relevant systematic reviews; and studies that attempted to externally validate risk stratification tools.
Data Extraction: One reviewer extracted data and a second checked accuracy. Two independent reviewers assigned quality ratings using predefined criteria.
Data Synthesis: No RCTs compared screening with no screening, CAAS with medical treatment, or assessed intensification of medical therapy. Given the specificity of ultrasound (range 88% to 94% for CAS ≥50% to ≥70%), its use in low-prevalence populations would yield many false-positive results. Only one fair-quality study attempted external validation of a risk stratification tool to distinguish persons who are more likely to have CAS; the tool's discrimination was inadequate (c-statistic for ≥50% CAS, 0.60; 95% CI, 0.56 to 0.64). Our meta-analyses of RCTs comparing CEA with medical therapy found an absolute risk reduction of 5.5 percent (95% CI, 3.9 to 7.0) for any nonperioperative stroke over approximately 5 years. Meta-analyses for perioperative (30-day) stroke or death after CEA found rates of 2.4 percent (95% CI, 1.7 to 3.1) using all trials of CEA, regardless of the comparator; and 3.3 percent (95% CI, 2.7 to 3.9) using cohort studies (7 studies; n=17,474). Rates of perioperative stoke or death after CAAS were similar or slightly higher. Other important potential harms of CEA or CAAS include nonfatal perioperative myocardial infarction (approximately 0.8% rate after CEA), cranial nerve injury, pulmonary embolism, pneumonia, local hematoma requiring surgery, and psychological harms (e.g., anxiety or labeling). Externally validated, reliable risk stratification tools that can distinguish persons with asymptomatic CAS who have increased or decreased risk for ipsilateral stroke or harms after CEA or CAAS are not available.
Limitations: Medical therapy in trials varied and often lacked treatments that are now standard. For this reason, and because advances in medical therapy have reduced the rate of stroke in persons with asymptomatic CAS in recent decades, the true reduction of stroke or composite reduction of cardiovascular events is unknown. Trials utilized highly selected surgeons. No trials focused on a population identified by screening in primary care. Harms may be underreported.
Conclusion: Current evidence does not sufficiently establish incremental overall benefit of CEA, CAAS, or intensification of medical therapy beyond current standard medical therapy. Potential for overall benefit is limited by low prevalence in the general asymptomatic population and by harms from screening and treatment. Evidence is insufficient to allow reliable risk stratification.