Incidence and Risk Factors for Instrumentation-related Complications After Scoliosis Surgery in Pediatric Patients With NF-1

Study Design. Retrospective cohort study. Objective. To assess the incidence and risk factors of instrumentation-related complication (IRC) in pediatric patients surgically treated for neurofibromatosis type 1 (NF-1) dystrophic scoliosis. Summary of Background Data. Surgical management including growing rods technique and early definitive fusion has been recommended to avoid progression of NF-1 scoliosis. However, no study has yet investigated the incidence and risk factors of IRC in these surgical interventions. Methods. Data of 59 pediatric NF-1 patients who had been surgically treated for dystrophic scoliosis were retrospectively reviewed. All of their clinical and radiographic data were collected. We evaluated potential risk factors, including age, sex, curve type, preoperative Cobb angle, kyphosis angle and spinal length, and surgical procedure. The univariate analysis and multivariate logistic regression analysis were performed to identify the risk factors associated with IRC. Results. Seventeen (28.8%) patients suffered 19 IRCs, including seven cases of curve progression, three cases of screw dislodgement, three cases of adding-on phenomenon, two cases of rod breakage, two cases of proximal junctional kyphosis, one case of cap loosing, and one case of pedicle cutting. The univariate logistic regression analysis revealed age<9, kyphosis >= 50 degrees, and growing-rod techniques the significant risk factors (P<0.05). Binomial logistic regression analysis demonstrated two indepen-dent risk factors of IRC, including kyphosis >= 50 degrees (OR: 8.23; P = 0.025) and application of growing-rod technique (OR: 8.75; P = 0.032). Conclusion. 17/59 (28.8%) patients suffered IRCs and age less than 9 years, kyphosis more than 508 and application of growing-rod are three risk factors for IRC after surgical treatment of NF-1 dystrophic scoliosis. Identification of these risk factors aids in stratifying preoperative risk to reduce IRC incidence. In addition, the results could be used in counseling patients and their families during the consent process.