BACKGROUND:Complicated urinary tract infections (cUTI) are universal reasons for hospitalization, and highly likely to develop into sepsis or septic shock. Carbapenem antibiotics with potentially higher efficacy or with fewer and milder side effects have increased in popularity, but evidence is limited by a scarcity of randomized controlled trials (RCTs) comparing different carbapenem antibiotics for cUTI. Network meta-analysis is a useful tool to compare multiple treatments when there is limited or no direct evidence available. OBJECTIVE:The aim of this study is to compare the efficacy and safety of different carbapenems with alternative antibiotics for the treatment of cUTI. METHODS:Pubmed, Medline, CENTRAL, and Embase were searched in November 2018. Studies of cUTI patients receiving carbapenem were included. We performed network meta-analysis to estimate the risk ratio (RR) and 95% credible interval (CrI) from both direct and indirect evidence; traditional meta-analysis was also performed. Primary outcomes were clinical and microbiological treatment success. RESULTS:A total of 19 studies and 7380 patients were included in the analysis. Doripenem (DOPM) was associated with lower clinical treatment success rates than other carbapenems. Although the efficacy of other carbapenems by RRs with 95% CrIs did not show statistical differences, the cumulative rank probability indicated that meropenem/vaborbactam (MV), ertapenem (ETPM), and biapenem (BAPM) had higher clinical and microbiological treatment success rates; imipenem/cilastatin (IC) and MV showed higher risk of adverse events (AEs). CONCLUSIONS:MV was associated with higher treatment success rates for cUTI, especially for cUTI caused by carbapenem-resistant uropathogens, but also with higher risk of AEs. Our findings suggest MV as a first-choice treatment of carbapenem-resistant cUTI. ETPM, BAPM, and meropenem (MEPM) is another reasonable choice for cUTI empiric therapy.