Background: Surgical bypass of an occluded arterial segment is one of the mainstay treatments for patients with critical limb ischaemia (CLI). However, it was introduced without formal evaluation.
Objectives: To determine the effects of bypass surgery in patients with CLI.
Search strategy: The Cochrane Peripheral Vascular Diseases Group (PVD) searched their trials register (last searched November 2007) and the Cochrane Central Register of Controlled Trials (CENTRAL) in The Cochrane Library (last searched Issue 4, 2007). Principal trial investigators were also contacted.
Selection criteria: All randomised controlled trials (RCTs) of bypass surgery versus control or any other treatment.
Data collection and analysis: For the update one author and PVD editorial staff extracted data and assessed trial quality. Unpublished data were obtained from trial investigators. Data were analyzed using Peto odds ratio (OR) or weighted mean difference (fixed and random effects models).
Main results: Nineteen trials were identified. Eight involved a total of just over 1200 patients. Four trials compared bypass surgery with angioplasty (PTA) and one each with thromboendarterectomy, thrombolysis, exercise, and spinal cord stimulation. Four included patients with intermittent claudication (IC) and CLI, two were restricted to claudicants, and two to CLI. Vein grafts were used for distal reconstructions and synthetic prostheses for aorto-iliac or ilio-femoral bypasses. Six trials included mortality. In general, trial quality was good; blinding was not possible. Mortality and amputation rates did not differ significantly between bypass surgery and PTA; primary patency was significantly higher in the bypass group after 12 months (Peto OR 1.6, 95% CI 1.0 to 2.6) but not after four years (P = 0.14). In patients with lower CLI, surgery was associated with increased surgical complications (Peto OR 2.69, 95% CI 1.87 to 3.86) and longer hospital stays during the first year, mean stay 46.1 days (SD 53.9) compared with 36.4 days (SD 51.4) for those receiving PTA (P < 0.0001). Amputation rates were significantly lower in bypass compared with thrombolysis (Peto OR 0.2, 95% CI 0.1 to 0.6); mortality rates did not differ. Blood flow restoration was significantly greater in bypass than in thromboendarterectomy patients (Peto OR 9.2, 95% CI 1.7 to 50.6); mortality and amputation rates did not differ. Bypass surgery outcomes did not differ significantly from exercise or spinal cord stimulation.
Authors' conclusions: There is limited evidence for the effectiveness of bypass surgery compared with other treatments; no studies compared bypass to no treatment. Further large trials are required.