Non-operative versus operative management of open fractures in the paediatric population: A systematic review and meta-analysis of the adverse outcomes.

Open fractures are at greater risk of infection and delayed bone healing. Guidelines to reduce these risks exist for adult open fracture management but not for paediatric open fractures, where there is considerable practice variability. This systematic review evaluates the quality of the evidence and clinical outcomes for paediatric open fractures treated non-operatively versus operatively. A PROSPERO-registered, PRISMA-compliant systematic review and meta-analysis. Databases searched were MEDLINE, Embase and Web of Science (WoS) up to March 2019. Clinical studies reporting adverse outcomes of both non-operative and operative management of paediatric open fractures were included. Data extracted included demographics, treatment methods and outcomes. The primary outcome was infection (osteomyelitis and/or wound infection). Secondary outcomes were abnormalities of bone healing (delayed union, malunion and nonunion) and re-fracture. The ROBINS-I tool was used to assess risk of bias. Seventeen studies reporting 1093 open fractures were included. Non-operatively managed injuries had a lower risk of osteomyelitis (RR 0.33 [95%CI 0.12-0.86]), wound infection (RR 0.47 [95%CI 0.22-0.97]) and nonunion (RR 0.27 [95%CI 0.09-0.80]). Gustilo-Anderson (GA) III injuries had the highest incidence of osteomyelitis (10.7%) with no difference in outcome between operative and non-operative groups (RR 0.67 [95%CI 0.22-2.03]). Tibial GA III fractures were associated with a lower risk of osteomyelitis than femoral fractures: adverse effects were seen in the operative group. All studies were retrospective, observational and at high risk of bias. The quality of the evidence relating to paediatric open fractures is low, and findings are limited by significant methodological flaws in the literature. GA I and II fractures were commonly managed non-operatively and associated with a lower infection rate. Operative intervention was more frequent in GA III fractures, where the risk of infection was highest. Operative management of GA III fractures was not associated with a lower infection risk compared to non-operative management. Robust prospective multi-centre studies are needed to explore further the most effective management of paediatric open fractures and to inform guideline development.