Surgical treatment of patellar dislocation: A network meta-analysis of randomized control trials and cohort studies

BACKGROUND: Currently, there are many surgical options for patellar dislocation. The purpose of this study is to perform a network meta-analysis of the randomized controlled trials (RCTs) and cohort studies to determine the better treatment. METHOD: We searched the Pubmed, Embase, Cochrane Central Register of Controlled Trials, Web of Science, clinicaltrials.gov and who.int/trialsearch. Clinical outcomes included Kujala score, Lysholm score, International Knee Documentation Committee (IKDC) score, redislocation or recurrent instability. We conducted pairwise meta-analysis and network meta-analysis respectively using the frequentist model to compare the clinical outcomes. RESULTS: There were 10 RCTs and 2 cohort studies with a total of 774 patients included in our study. In network meta-analysis, double-bundle medial patellofemoral ligament reconstruction (DB-MPFLR) achieved good results on functional scores. According to the surface under the cumulative ranking (SUCRA), DB-MPFLR had the highest probabilities of their protective effects on outcomes of Kujala score (SUCRA 96.5 %), IKDC score (SUCRA 100.0%) and redislocation (SUCRA 67.8%). However, DB-MPFLR (SUCRA 84.6%) comes second to SB-MPFLR (SUCRA 90.4%) in Lyshlom score. It is (SUCRA 70%) also inferior to vastus medialis plasty (VM-plasty) (SUCRA 81.9%) in preventing Recurrent instability. The results of subgroup analysis were similar. CONCLUSION: Our study demonstrated that MPFLR showed better functional scores than other surgical options.