Published online before print October 2, 2003, doi: 10.1161/01.STR.0000096458.53961.60
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Stroke. 2003;34:2573.)
© 2003 American Heart Association, Inc.
Original Contributions |
Editorial Comment—Identifying Risk Factors for Perioperative Outcomes After Carotid Endarterectomy: The Story Continues
School of Public Health, University of Alabama at Birmingham, Birmingham, Alabama
School of Public Health, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina
Given that the perioperative stroke and death rate associated with carotid endarterectomy (CEA) ranges from 2% to 8%,1–5 identification of subgroups of patients with differential risk is critically important. Given this relatively high average rate, an absolute difference of 2% to 4% in the perioperative risk of CEA could easily occur and may be sufficient to change the positive overall efficacy of CEA to negative. The growing literature5–11 suggests that substantial differences by patient characteristics exist in perioperative risk associated with CEA.
The study by Tu and colleagues11 is an important contribution to our understanding of the perioperative risk associated with CEA, reporting risk factors for 30-day stroke and death associated with CEA performed in Ontario, Canada, from 1994 through 1997. This report is unique for several reasons. The first has to do with statistical power. The proportion of patients suffering events is relatively low, so a very large sample size is required for there to be a sufficient number of patients with 
events
to permit appropriate statistical analyses to identify risk factors with reasonable precision. With a sample size of >6000 procedures and 361 events, the present study is among the largest studies to date, therefore providing the most precise estimates of the impact of risk factors on perioperative stroke and death. Second, although the study was retrospective, a major focus of effort was the standardization of procedures for chart abstraction and assessment of risk factors; thus, this data collection effort was likely more robust than many surgeon- or institution-specific retrospective chart reviews. Indeed, many of the definitions were the same as those used in North American Symptomatic Carotid Endarterectomy Trial (NASCET), so there was more standardization than in most other published CEA series. Finally, the linkage through the Canadian Institute for Health Information system and the Ontario Registered Persons Database makes it unlikely that major stroke events during the 30-day follow-up period went undetected. This feature is particularly important because in many locations privacy regulations restrict the ability to conduct record linkage to evaluate nonfatal outcomes in retrospectively identified cohorts without consent.
These results are presented in a format that can be used easily by practicing clinicians. A simple score, calculated by adding the number of risk factors present (history of stroke or transient ischemic attack, presence of contralateral occlusion, history of atrial fibrillation, history of congestive heart failure, and history of diabetes), was shown to be associated with the differential risk from a rate of 3.3% for those with no risk factors to a rate of 9.5% and higher for those with 
2 risk factors. This simple checklist can be applied quickly in the counseling of patients.
Rothwell et al6 conducted a systematic review of CEA studies published from 1981 to 1996 that reported perioperative risk data by 1 clinical or angiographic characteristics. This effort carefully reviewed the literature to select only studies meeting strict criteria; of 126 studies reviewed, only 35 met the criteria. The selected studies included a variety of study designs ranging from retrospective case series to prospective randomized clinical trials. Although overall a high-quality systematic review according to the Oxman and Guyatt12 index, the review of potential studies was likely problematic because of a lack of standardization in the definition of many variables of interest. Rothwell et al found cerebral versus ocular transient ischemic attack, age >75 years, systolic hypertension, female sex, and peripheral vascular disease to be significant independent predictors of perioperative stroke and death.6 It is notable that these risk factors differ substantially from those identified in the present study, in which differences in risk were found by symptomatic status, atrial fibrillation, contralateral carotid occlusion, congestive heart failure, and diabetes.11 The differences between these 2 reports could arise from differences in populations or methods, publication bias, or chance, all of which underscore the importance of more work in this area.
Although the present report has many strengths, limitations exist. The registry was not designed to tell us about the perioperative stroke and death risk of the patients who were screened but did not have CEA. There can be differential use of the procedure in patients with and without the risk factor. Selection bias is one of the most important sources of bias in observational studies. It is difficult to control for confounding by indication in these types of studies, so variables (such as history of diabetes) that are associated with the use (or nonuse) of CEA can influence the outcome measures. This type of bias can affect the estimates of the 30-day stroke and death rate and the direction of the effect.
In interpreting these results, we should also remember that this report focuses on the increased risk associated with CEA and does not address the equally likely differences between subgroups of patients in the benefit of the procedure during the subsequent postoperative period. An equal effort could be directed at determining those patients who, without CEA, would be at higher-than-average risk of subsequent stroke and whose stroke risk could be substantially reduced by the surgery. For example, patients with a contralateral carotid occlusion were found to be at greater perioperative risk of stroke and death; however, this may be a subgroup for which the successful completion of the procedure that maintains patent flow through a single remaining carotid artery warrants the increased risk through potentially greater reduction of events over the subsequent follow-up period. Conversely, age and degree of carotid stenosis were not identified as major predictors of perioperative events. However, that the surgery can be performed safely in young patients with a low level of carotid stenosis may not be warranted given the low likelihood of subsequent events during the follow-up. Thus, the counseling of patients should focus on both the risk of the procedure that was described by Tu et al11 and the likely benefit of the procedure. This difference in both risk and benefit is best described by randomized clinical trials and underscores why information arising from registry efforts should be interpreted with caution.
This report by Tu and colleagues11 clearly and precisely describes the risks associated with the conduct of endarterectomy. In the ever-growing CEA literature describing the differential risk of subpopulations, the report is exceptional and can serve as the basis for the counseling of patients before the procedure. Differences in perioperative risk faced by specific patients are important components in the decision as to whether to proceed with CEA.
 |
References |
|---|
 Top References |
|---|
1. Rothwell PM, Slattery J, Warlow CP. A systematic review of the risks of stroke and death due to endarterectomy for symptomatic carotid stenosis. Stroke. 1996; 27: 260–265.
2. Rothwell PM, Slattery J, Warlow CP. A systematic comparison of the risks of stroke and death due to carotid endarterectomy for asymptomatic and asymptomatic stenosis. Stroke. 1996; 27: 266–269.
3. Young B, Moore WS, Robertson JT, Toole JF, Ernst CB, Cohen SN, Broderick JP, Dempsey RJ, Hosking JD, for the ACAS Investigators. An analysis of perioperative surgical mortality and morbidity in the Asymptomatic Carotid Atherosclerosis Study. Stroke. 1996; 27: 2216–2224.
4. Ferguson GG, Eliasziw M, Barr HWK, Clagett GP, Barnes RW, Wallace C, Taylor DW, Haynes B, Finan JW, Hachinski VC, Barnett HJM, for the NASCET Collaborators. The North American Symptomatic Carotid Endarterectomy Trial: surgical results in 1415 patients. Stroke. 1999; 30: 1751–1758.
5. Bond R, Narayan SK, Rothwell PM, Warlow CP, for the ECST Collaborative Group. Clinical and radiographic risk factors for operative stroke and death in the European Carotid Surgery Trial. Eur J Vasc Endovasc Surg. 2002; 23: 108–116.[CrossRef][Medline] [Order article via Infotrieve]
6. Rothwell PM, Slattery J, Warlow CP. Clinical and angiographic predictors of stroke and death from carotid endarterectomy: systematic review. BMJ. 1997; 315: 1571–1577.
7. Goldstein LB, Samsa GP, Matchar DB, Oddone EZ. Multicenter review of preoperative risk factors for endarterectomy for asymptomatic carotid artery stenosis. Stroke. 1998; 29: 750–753.
8. Urbach DR, Bell CM. The effect of patient selection on comorbidity-adjusted operative mortality risk: implications for outcomes studies of surgical procedures. J Clin Epidemiol. 2002; 55: 381–385.[CrossRef][Medline] [Order article via Infotrieve]
9. Frawley JE, Hicks RG, Woodforth IJ. Risk factors for peri-operative stroke complicating carotid endarterectomy: selective analysis of a prospective audit of 1000 consecutive operations. Aust N Z J Surg. 2000; 70: 52–56.[Medline] [Order article via Infotrieve]
10. Ballotta E, DaGiau G, Renon L. Is diabetes mellitus a risk factor for carotid endarterectomy? A prospective study. Surgery. 2001; 129: 146–152.[CrossRef][Medline] [Order article via Infotrieve]
11. Tu JV, Wang H, Bowyer B, Green L, Fang J, Kucey D, for the Participants in the Ontario Carotid Endarterectomy Registry. Risk factors for death or stroke after carotid endarterectomy: observations from the Ontario Carotid Endarterectomy Registry. Stroke. 2003; 34: 2568–2575.
12. Oxman A, Guyatt G. Validation of an index of the quality of review articles. J Clin Epidemiol. 1991; 44: 91–98.[Medline] [Order article via Infotrieve]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||