With the publication of randomized controlled trials on carotid endarterectomy, the appropriate indications for this surgery (“when to operate”) are becoming better defined. From a public health perspective, however, the benefits of carotid surgery are convincing only when surgical morbidity and mortality rates are very low. Thus, an equally important question needs to be considered: “Who should operate in carotid disease?”

In this issue of American Family Physician, Biller and Thies¹ review evidence showing that carotid endarterectomy can be highly effective in preventing stroke. Patients with high-grade internal carotid artery stenosis (greater than 70 percent) that is ipsilateral to the side of previous symptoms or nondisabling stroke benefit dramatically from the combined use of surgery and medical therapy compared with the use of medical therapy alone. In patients with severe carotid artery stenosis, fewer than eight carotid endarterectomies are required to prevent one stroke. The benefits of carotid endarterectomy are not as dramatic in symptomatic patients with lesser degrees of stenosis (50 to 69 percent), and surgery may not be beneficial in patients with less than 50 percent stenosis.

The benefits of surgery in asymptomatic patients have been controversial. In these patients, carotid endarterectomy may prevent only nondisabling strokes rather than disabling ones, and the clinical significance of a small risk reduction for such an outcome is questionable.² A recent meta-analysis³ of all randomized controlled trials conducted in asymptomatic patients indicated that the number of surgeries needed to prevent one stroke over three years is close to 50.¹

Although some authorities speculate that certain subgroups of asymptomatic patients might benefit from surgery, we have no idea who these patients are. Many experts believe that surgery is not indicated for patients without symptoms, regardless of the severity of carotid stenosis.² Treatment must always be individualized, but risk-factor reduction and education about the warning signs of ischemic stroke, rather than surgery, are appropriate measures in most asymptomatic patients.

When carotid surgery is an appropriate recommendation, who should be performing the operation? In all categories of carotid disease, low perioperative complication rates are crucial for the benefits of surgery to be realized. The surgeons who participated in the carotid surgery trials were high-volume experienced technicians with low mortality rates for the procedures they performed. For example, the 30-day mortality rate was 0.6 percent in the North American Symptomatic Carotid Endarterectomy Trial (NASCET)⁴ and 0.1 percent in the Asymptomatic Carotid Atherosclerosis Study (ACAS).⁵ In fact, the risk of surgery in ACAS was lower than the risk of angiography.

Several population-based studies have demonstrated that the low complication rates cited for the NASCET and the ACAS are almost impossible to achieve outside of centers of excellence such as those in which these trials were performed.^6–9 Mortality rates in American and Canadian hospitals are substantially higher than those reported in the surgical trials, and no evidence indicates that selective referral to surgeons with a proven track record of excellence in carotid surgery is taking place.⁹ Physicians and, indeed, patients have a right to know the performance data on carotid surgery for local surgeons and hospitals. However, a recent national survey found that only 19 percent of referring physicians had this information.¹⁰

Performance rates for carotid endarterectomy rose with publication of the NASCET and ACAS results in the 1990s.⁹ Referring physicians and surgeons are now better aware of the appropriate indications for carotid surgery. Yet, endarterectomies continue to be performed for inappropriate reasons. Even when patient selection is appropriate, selection of the surgeon can influence outcome. For example, investigators in one study found that low surgeon volume (fewer than six cases per year) was associated with higher complication rates and suggested that a minimum volume threshold for this procedure should be established.⁷

One study¹¹ found that the use of a citywide audit, combined with an education campaign and ongoing audit, improved the appropriateness and safety of carotid endarterectomies. Public health pressures will probably continue to affect our access to performance data such as these. For now, it is wise for referring physicians to seek out surgeons and hospitals with a proven track record in performing successful, safe carotid surgery. Selective referral to regionalized centers of excellence may be one model for this approach.

What does the future hold? Yet another trial, the Asymptomatic Carotid Surgery Trial, is ongoing. This trial may help to answer questions about the suitability of surgery in patients without symptoms. In addition, non-invasive forms of imaging the carotid arteries, such as magnetic resonance angiography, may replace conventional angiography and lower the overall risk of intervention. Nonsurgical alternatives, such as carotid angioplasty and stenting, may play a role in selected patients. At present, however, carotid angioplasty and stenting should be performed only in research studies. Finally, the results of the Aspirin in Carotid Endarterectomy (ACE) trial¹² indicated that 30-day perioperative morbidity and mortality following carotid endarterectomy were significantly reduced by the use of low doses of aspirin (80 or 325 mg) compared with high doses (650 or 1,300 mg). High-dose aspirin is likely better than no aspirin at all; however, patients scheduled for carotid artery surgery should be placed on low-dose aspirin prior to surgery and remain on such a dose indefinitely.

Risk-factor reduction, aspirin in an appropriate dosage and selective referral to surgeons with proven track records of excellence are measures presently available to family physicians. In particular, it is up to referring physicians to wisely select “who should operate.”