A systematic review and meta-analysis of adverse outcomes following non-buried versus buried kirschner wires for paediatric lateral condyle elbow fractures
Wormald JCR., Park CY., Eastwood DM.
© 2017, British Editorial Society of Bone and Joint Surgery. All rights reserved. Purpose Fractures of the lateral condyle of the humerus in children are a common injury. If displaced or unstable they may require surgical reduction and fixation with Kirschner wires (K-wires). K-wires are placed using either an open or closed technique. The decision to bury or leave the ends extending through the skin is surgeon-dependent and based on factors including post-operative infection risk, bony union and ease of wire removal. Methods We performed a systematic review and meta-analysis of non-buried versus buried K-wires for lateral condyle elbow fractures in children in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses methodology. A comprehensive search strategy included Medline, Embase and CINAHL via NICE Evidence from database inception to June 2017. Two authors independently reviewed, included or excluded articles, extracted data and assessed for quality with the ROBINS-I tool. We performed direct comparison meta-analysis for all adverse events, post-operative infection and failure of bony union. Results Three studies were analysed comprising of 434 participants. There was a significantly reduced relative risk of adverse events in the non-buried group, equating to approximately 45% reduced risk (RR 0.55, 95% confidence interval 0.34 to 0.88). There were no significant differences in risk of post-operative infection or failure of bony union. All three cost-analyses in the included studies observed savings with non-buried K-wires. Conclusion Non-buried K-wires for lateral condyle elbow fractures convey a lower risk of adverse events and may be more cost-effective compared with buried K-wires. Non-buried K-wires do not appear to increase the risk of infection or failure of bony union. These findings are limited by a high risk of bias due to inherent methodological flaws in the design of included studies.