Background:Kawasaki disease (KD) is a systemic inflammatory condition occurring predominantly in children. Coronary artery bypass grafting (CABG) is performed in the presence of inflammation and aneurysms of the coronary arteries. The objectives of our study were to assess which CABG strategy provides better graft patency and long-term outcomes. Methods:A systematic review using Medline, Cochrane and Scopus databases was performed by February 2020, incorporating a network meta-analysis, performed by random-effect model within a Bayesian framework, and pooled prevalence of adverse outcomes. Hazard ratios (HR) and corresponding 95% credible intervals (CI) were calculated by Markov chain-Monte Carlo methods. Results:Among 581 published reports, 32 studies were enrolled, including 1191 patients undergoing CABG for KD. Graft patency of internal thoracic arteries (ITA), saphenous veins (SV) and other arteries (gastroepiploic artery and radial artery) were compared. ITAs demonstrated the best patency rates at long-term follow-up (HR 0.33, 95% CI:0.17-0.66). Pooled prevalence of early mortality after CABG was 0.28% (95% CI:0.00-0.73%,I²=0%,tau²=0), with 63/1108 and 56/1108 patients, respectively, undergoing interventional procedures and surgical re-interventions during follow-up. Pooled prevalence was 3.97% (95% CI:1.91-6.02%,I²=60%,tau²=0.0008) for interventional procedures and 3.47% (95% CI:2.26-4.68%,I²=5%,tau²<0.0001) for surgical re-interventions. Patients treated with arterial, venous and mixed (arterial plus second venous graft) CABG were compared to assess long-term mortality. Mixed CABG (HR 0.03,95% CI: 0.00-0.30) and arterial CABG (HR 0.13, 95% CI: 0.00-1.78) showed reduced long-term mortality compared with venous CABG. Conclusions:CABG in KD is a safe and effective procedure. Use of arterial conduits provides better patency rates and lower mortality at long-term follow-up.