Eversion versus Conventional Carotid Endarterectomy: A Meta-analysis of Randomised and Non-randomised Studies

• Published in: European journal of vascular and endovascular surgery Vol. 42; no. 6; pp. 751 - 765
• Main Authors: Antonopoulos, C.N, Kakisis, J.D, Sergentanis, T.N, Liapis, C.D
• Format: Journal Article
• Language: English
• Published: England Elsevier Ltd 01-12-2011
• Subjects: Aged; Carotid disease; Carotid Stenosis - mortality; Carotid Stenosis - surgery; Carotid surgery; Confidence Intervals; Conventional endarterectomy; Endarterectomy, Carotid - methods; Eversion endarterectomy; Humans; Meta-Analysis; Numbers Needed To Treat; Odds Ratio; Postoperative Complications - prevention & control; Randomized Controlled Trials as Topic; Stroke - mortality; Stroke - prevention & control; Surgery; Survival Analysis; conventional carotid endarterectomy; Carotid disease; Carotid surgery; Conventional endarterectomy; PCCEA; risk difference; ESVS; odds ratio; primary closure carotid endarterectomy; Mantel–Haenszel; common carotid artery; SVS; transient ischaemic attack; NNT; Eversion endarterectomy; internal carotid artery; ECEA; patch carotid endarterectomy; MOOSE; CCA; Meta-Analysis; ECA; CCEA; OR; NASCET; CI; ICA; M–H; North American Symptomatic Carotid Endarterectomy Trial; PCEA; TIA; eversion carotid endarterectomy; RD; European Society for Vascular Surgery; Society for Vascular Surgery; Meta-analysis of Observational Studies in Epidemiology; confidence intervals; number needed to treat; external carotid artery
• ISSN: 1078-5884; 1532-2165
• DOI: 10.1016/j.ejvs.2011.08.012

Abstract Background To compare eversion (ECEA) and conventional (CCEA) carotid endarterectomy from randomised and non-randomised studies. Methods Pooled odds ratios (ORs) with 95% confidence intervals (95%CIs) and numbers needed to treat (NNTs) were appropriately calculated. A sub-analysis was performed on studies directly comparing ECEA vs. patch CEA (PCEA). Meta-regression analysis was performed to examine the effect of potentially meaningful patient-related, procedure-related and definition-related modifiers. Power calculations were also conducted. Results A total of 21 studies were deemed eligible (8530 ECEA and 7721 CCEA procedures), seven of which were randomised and 14 non-randomised. ECEA was associated with significant reduction in perioperative stroke (OR = 0.46, 95%CI: 0.35–0.62, NNT = 68, 95%CI: 56–96), death (OR = 0.49, 95%CI: 0.34–0.69, NNT = 100, 95%CI: 85–185) and stroke-related death (OR = 0.40, 95%CI: 0.23–0.67, NNT = 147, 95%CI: 115–270); the results were replicated at the sub-analysis on PCEA. Concerning long-term outcomes, ECEA presented with a significant reduction in late carotid artery occlusion (OR = 0.48, 95%CI: 0.25–0.90, NNT = 143, 95%CI: 100–769) and late mortality (OR = 0.76, 95%CI: 0.61–0.94, NNT = 40, 95%CI: 25–167); the sub-analysis on PCEA replicated only the finding on late mortality. Meta-regression analysis did not point to significant effects mediated by the examined modifiers. Power calculations suggested adequate statistical power. Conclusions ECEA compared to CCEA may be associated with a lower incidence in both short-term and long-term outcomes, which does not seem to be hampered by potentially meaningful modifiers.