Different treatments for 3- or 4-part proximal humeral fractures in the elderly patients: A Bayesian network meta-analysis of randomized controlled trials

BACKGROUND: Proximal humeral fractures are the third most common fracture in the body, and their incidence is rising year by year as the population ages. However, the treatment of the proximal humerus in parts 3 and 4 is still debatable, necessitating a network meta-analysis to determine the best treatment for each treatment modality. METHODS: We searched PubMed, Embase, Cochrane Library for randomized controlled trials on proximal humeral fractures up to June 21, 2022. We performed data extraction and literature quality assessment by two independent authors and extracted constant score and reoperation rate as indicators for evaluation. Stata software, Revman software, JAGS software and the R-based BlandAltmanLeh package, gemtc package and riags package were used to perform this Bayesian network meta-analysis. RESULTS: Following screening, 11 papers with a total of 648 participants were included in the analysis. The SUCRA values for the constant score were in the following order: RSA, IMN, Conservative, HA, and LP, and the SUCRA values for the reoperation rate were LP, HA, IMN, Conservative, and RSA. CONCLUSION: The elderly with 3- or 4-part proximal humeral fractures should consider RSA because it received the best evaluation ranking in terms of constant score and reoperation rate. SYSTEMATIC REVIEW REGISTRATION: https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42022341209, identifier: CRD42022341209.