Surgical treatment of osteogenesis imperfecta: a summary of the incidence of femoral implant-related complications in children with Sillence type I, III and IV.

Purpose: This study explored the incidence of IRCs used in the procedures of the femur in children with osteogenesis imperfecta (OI) and investigated the independent risk factors of IRCs.

Methods: Three hundred eight-eight cases of surgical data about children with OI were included, who were treated with plate, elastic nail, Kirschner wire and telescopic rod. The choice of different procedures depended on the age of children, the status of femur and the availability of devices. Patient demographics and major IRCs were recorded to compare the outcomes of the four procedures. Then, Cox proportional hazard regression was used to analyse the independent risk factors of IRC, and subgroup analysis was applied to further verify the above results.

Results: The total incidence of IRC in the four groups was 90.1% (191/212) for plate, 96.8% (30/31) for Kirschner wire, 87.7% (57/65) for elastic nail and 30.0% (24/80) for telescopic rod. The incidence of IRC in the telescopic rod was lower than that in plate, elastic nail and Kirschner wire (P < 0.001). Cox proportional hazard regression analysis confirmed that procedure was the independent risk factor of IRC (HR, 0.191; 95% CI, 0.126-0.288; P < 0.001), fracture (HR, 0.193; 95% CI, 0.109-0.344; P < 0.001) and deformity (HR, 0.086; 95% CI, 0.027-0.272; P < 0.001). In addition, age of surgery was the independent risk factor of fracture (HR, 0.916; 95% CI, 0.882-0.952; P < 0.001) and deformity (HR, 1.052; 95% CI, 1.008-1.098; P = 0.019). Subgroup analysis confirmed that age of surgery, gender, classification, preoperative state and angle did not affect the effect of telescopic rod on reducing the risk of IRCs.

Conclusions: In our cohort, lower incidence of IRCs was observed in telescopic rod group compared with plate, Kirschner wire and elastic nail. Procedure and age of surgery were independent risk factors of fracture. Likewise, procedure and age of surgery were independent risk factors of deformity, and procedure was independent risk factors of IRC.