Minimally invasive plating versus either open reduction and plate fixation or intramedullary nailing of humeral shaft fractures: a systematic review and meta-analysis of randomized controlled trials

Hypothesis The purpose of this study was to perform a meta-analysis comparing the clinical outcomes and complications between anterior humeral minimally invasive plate osteosynthesis (MIPO) and the 2 standard techniques, either open reduction and plating or humeral nailing. Methods We performed a systematic review of Medline, Embase, Scopus, and Google Scholar to identify relevant randomized controlled trials in the English- and German-language literature. Eligibility criteria included randomized controlled trials comparing at least 1 surgical intervention with MIPO and reporting the primary clinical outcome using a validated functional scoring system and description of complications. Publication bias was assessed by funnel plot, and the risk of bias was established using the Cochrane Collaboration's Risk of Bias Tool. Heterogeneity was assessed using χ and I statistics. Results Eight prospective randomized studies (N = 376) met the eligibility criteria and were included in the analysis. The pooled estimate for clinical outcome showed that MIPO resulted in a significantly better outcome (standardized mean difference, 0.366; 95% confidence interval, 0.16 to 0.571; P = .0001; I = 61%). The pooled estimate for all complications showed that the open reduction–internal fixation/nail group had a significantly higher complication rate (odds ratio, 0.507; 95% confidence interval, 0.285 to 0.905; P = .021; I = 97%). Conclusion Current evidence indicates the MIPO approach has better clinical outcomes with a lower rate of complications compared with alternative surgical techniques. However, the results of this meta-analysis are limited by problems inherent in the primary studies, including poor reporting of randomization protocols, as well as possible attrition bias and reporting bias, of the primary studies. Future publications may therefore change the trend of the pooled estimate in either direction