Time trends in the published risks of stroke and death due to endarterectomy for symptomatic carotid stenosis

Background: Large randomised trials performed in the 1980s and early 1990s showed that carotid endarterectomy (CEA) is beneficial for patients with recently symptomatic severe stenosis. Some surgeons have argued that the operative risk of stroke and death has fallen over the last decade due to refinements in operative technique, and that the indications for surgery should therefore now be broadened. Yet, studies of routinely collected data report higher operative mortality than in the trials, and surgical case series without independent postoperative assessment by a neurologist may not provide reliable data on stroke risk. Methods: We performed a systematic review of all studies published between 1994 and 2001 inclusive that which reported the risks of stroke and death for symptomatic carotid stenosis, and compared the reported risks and patient characteristics with those in the ECST and NASCET and with our previous review of studies published prior to 1995. Pooled estimates of the operative risk of stroke and death were obtained by Mantel-Haenszel meta-analysis. Results: Of 383 studies published between 1994 and 2001, only 45 reported operative risks for patients with symptomatic stenosis separately. The pooled operative risk of stroke and death reported in studies published by surgeons only (4.2%, 95% CI = 2.9-5.5, 34 studies) was significantly lower (p < 0.0001) than that in the ECST and NASCET combined (7.0%, 95% CI = 6.2-8.0), whereas the pooled risk reported in studies that involved neurologists was similar (6.5%, 95% CI = 4.3-8.7, 11 studies, p = 0.6). In contrast, operative mortality in ECST and NASCET was significantly lower than in other studies published between 1994 and 2001. By comparison with our previous review, when stratified according to involvement of neurologists, we found no evidence of a reduction in published risks of death or stroke and death due to CEA between 1985 and 2001. Conclusions: There is no evidence of a systematic reduction over the last decade in the published risks of stroke and death due to CEA for symptomatic stenosis. Operative risks in studies with comparable outcome assessment are similar to ECST and NASCET. The surgical data from the large trials are still likely therefore to be applicable to routine clinical practice. Copyright (C) 2004 S. Karger AG, Basel.