Published online before print February 13, 2003, doi: 10.1161/01.STR.0000059757.63250.AC
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Stroke. 2003;34:592.)
© 2003 American Heart Association, Inc.
Letters to the Editor |
Potential Pitfalls in Applying Carotid Endarterectomy Perioperative Mortality and Morbidity Rates
Department of Endocrinology, The Queen Elizabeth Hospital, Woodville, South Australia, Australia
To the Editor:
Recent guidelines for carotid endarterectomy (CE) recommend a requirement for very low perioperative complication rates (
3.0%).1 For symptomatic patients, with 70% to 90% stenoses, however, a CE stroke perioperative complication rate of 6.0% is likely to show a 16.5% decrease in risk for nonfatal stroke or death at 2 years.2 The 2 major trials, the NASCET2 and ECST,3 had perioperative complication rates of 5.8 (95% CI, 3.3 to 8.3) and 9.1% (95% CI, 6.1 to 12.0), respectively. There was no significant difference between CE complication rates for the 2 trials (P=0.14), and the NASCET rate is significantly higher (P<0.01) than the present recommendation of 3.0%. The importance of the CE complication rate, with respect to any net patient benefit, has prompted calls for the routine availability of relevant data on both surgeons and hospitals. This is particularly relevant given comments that CE perioperative complication rates quoted for published studies may be atypically low due to a reporting bias, and that analysis of Medicare data finds CE 30-day complication rates that range between 5% and 11%.4
If we assume that the true underlying rate for CE perioperative complications is 6.0%, and that figure would ultimately be achieved by an institution, if it performed a large enough series of CE procedures (n=1000), the expected range of perioperative complications, in smaller random samples, can be calculated using a technique referred to as Monte Carlo sampling.5 Accordingly, when the recommended sample size of 100 is used, the results show that only perioperative complication rates 
10% are statistically distinct (P<0.05, 1-tail). Thus, within an institution, CE complication rates of 7%, based on small sample sizes, are likely to occur about 40% of the time, even though the true, but as yet undefined, CE perioperative complication rate is 6%. Alternatively, if the true institution CE complication rate is 10%, identical calculations establish that approximately 19% of all samples will have a rate 7%. These findings challenge the present assumption of predicting future CE complications rates, based on retrospective small sample analyses.
Although the NASCET and ECST trials provide some information about the risks and benefits of CE, considerable uncertainty still persists on how to effectively advise individual patients about the potential risks and benefits at any single institution. Moreover, one study found that 20% of surveyed physicians were unaware of the existing CE complication rates, while 20% of accredited surgery residence programs acknowledged a failure to implement any form of systematic CE audit.4 Notwithstanding these latter limitations, it has also been previously reported that patients rarely conceptualize interventions, such as CE, as a straightforward question of risk reduction.6 In summary, if the absolute benefits of CE are dependent on accurate data concerning the perioperative complication rate, then for many individual patients appropriate clinical advice as to the risk versus benefit trade-off cannot be reliably provided.
References
1. Biller J, Feinberg WM, Castaldo JE, et al. Guidelines for carotid endarterectomy. Stroke. 1998; 29: 554–562.
2. The North American Symptomatic Carotid Endarterectomy Trial Collaborators. Beneficial effect of carotid endarterectomy in symptomatic patients with high-grade carotid stenosis. N Engl J Med. 1991; 325: 445–453.[Abstract]
3. The European Carotid Surgery Trialists’ Collaborative Group. MRC European Carotid Surgery Trial: interim results for symptomatic patients with severe (70–99%) or with mild (0–29%) carotid stenosis. Lancet. 1991; 337: 1235–1243.[CrossRef][Medline] [Order article via Infotrieve]
4. Goldstein LB, Moore WS, Robertson JT, Chaturvedi S. Complication rates for carotid endarterectomy: a call to action. Stroke. 1997; 28: 889–890.Editorial.
5. Noreen EW. Computer-Intensive Methods for Testing Hypotheses. New York: J Wiley & Sons; 1989.
6. Fitzgerald SP, Phillipov G. Patient attitudes to commonly promoted medical interventions. Med J Aust. 2000; 172: 9–12.[Medline] [Order article via Infotrieve]
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||