ABSTRACT
BACKGROUND: The optimal revascularization technique in diabetic patients is an important unresolved question. PURPOSE: To compare long-term outcomes between the revascularization techniques of percutaneous coronary intervention (PCI) and coronary artery bypass grafting (CABG). DATA SOURCES: English-language publications in PubMed, the Cochrane Central Register of Controlled Trials, Ovid, and EMBASE between 1 January 1990 and 1 June 2014. STUDY SELECTION: Two investigators independently reviewed randomized, controlled trials comparing PCI (with drug-eluting or bare-metal stents) with CABG in adults with diabetes with multivessel or left main coronary artery disease. DATA EXTRACTION: Study design, quality, patient characteristics, length of follow-up, and outcomes were extracted. For duplicate publications, outcomes were obtained from the publication with the longest follow-up. DATA SYNTHESIS: 40 studies were combined using a Bayesian network meta-analysis that accounted for the variation in stent choice. The primary outcome, a composite of all-cause mortality, nonfatal myocardial infarction, and stroke, increased with PCI (odds ratio [OR], 1.33 [95% credible interval {CrI}, 1.01 to 1.65]). Percutaneous coronary intervention resulted in increased mortality (OR, 1.44 [CrI, 1.05 to 1.91]), no change in the number of myocardial infarctions (OR, 1.33 [CrI, 0.86 to 1.95]), and fewer strokes (OR, 0.56 [CrI, 0.36 to 0.88]). LIMITATIONS: Study design and length of follow-up were heterogeneous, and results were driven primarily by a single study. Costs and nonvascular complications of the interventions were not examined. CONCLUSION: Coronary artery bypass grafting seems to be the preferred revascularization technique in diabetics, especially if long-term survival is anticipated. However, because of residual uncertainties and increased risk for stroke with CABG, clinical judgment is required when choosing a revascularization technique in patients with diabetes. PRIMARY FUNDING SOURCE: Fonds de recherche du Québec-Santé.
