Age- and severity-adjusted treatment of proximal humerus fractures in children and adolescents—A systematical review and meta-analysis

BACKGROUND: Fractures of the proximal humerus in patients under the age of 18 years show a low incidence; existing clinical studies only comprise small patient numbers. Different treatment methods are mentioned in the literature but a comparison of the outcome of these methods is rarely made. Up to now, no evidence-based algorithm for conservative and operative treatment is available. The aim of this systematic review with meta-analysis was therefore to gather the best evidence of different treatment methods and their associated functional outcome, complication rates, rates of limb length discrepancies and radiological outcome. METHODS AND FINDINGS: The OVID database was systematically searched on September 30th in 2016 in order to find all published clinical studies on the subject of proximal humerus fractures of patients ≤18 years. Exclusion criteria were previously defined. The Coleman Methodology Score was used to evaluate the quality of the single studies. 886 studies have been identified by the search strategy. 19 studies with a total of 643 children (mean age: 11.8 years) were included into the meta-analysis with a mean Coleman Methodology Score of 71 ± 7.4 points. 18 of the 19 studies eligible for inclusion were retrospective ones, of the best quality available (mean follow-up ≥ 1 year, mean follow-up rate ≥ 65%). 56% of the patients were male. Proximal humerus fractures were treated conservatively in 41% and surgically in 59% of the cases (Elastic Stable Intramedullary Nailing (ESIN): 31%; K-wires: 20%; 8% other methods, e.g. plate osteosynthesis, olecranon traction). The overall success rate (good/excellent outcome) for all treatment methods was 93%. The success rate of ESIN (98%) and of K- wire fixation (95%) was significantly higher (p = 0.01) than the success rate of conservative treatment options (91%). A subgroup analysis of severely displaced fractures (Neer grade III/IV, angulation ≥ 20°) resulted in a change of success rates, to the disadvantage of conservative treatment methods (conservative treatment 82%, ESIN 98%, K-wires 95%; p < 0.001). Complication rates did not differ to a significant extent. 9% of the complications occurred in the patients treated by K-wire fixation, 8% if a conservative treatment option was chosen and 7% in the fractures that were stabilized by ESIN. A change from a one-nail technique to a two-nail technique reduced the complication rate of ESIN significantly. Follow-up X- rays without residual deformity could be found in 96% of the patients treated by ESIN, a rate which was higher than in the patients treated conservatively (93%) or by K-wire fixation (88%). The rate of arm length discrepancies at final follow- up was lower if the fractures were stabilized by ESIN (4%) than if they were treated conservatively (9%) or by K-wires (19%). An evaluation of age-dependent treatment options was performed. CONCLUSIONS: By performing this meta-analysis an evidence-based treatment algorithm could be introduced to treat the fractures according to the severity of displacement and according to the patient's age. For severely displaced fractures ESIN is the method of choice, with the best clinical and radiological outcome.