The Association Between the Utilization of Traction and Postoperative Complications Following Growing Rod Instrumentation for Early-onset Scoliosis.

BACKGROUND Preoperative and/or intraoperative traction have been proposed as adjunctive methods to limit complications associated with growth-friendly instrumentation for early-onset scoliosis (EOS). By gradually correcting the deformity before instrumentation, traction can, theoretically, allow for better overall correction without the complications associated with the immediate intraoperative correction. The purpose of this multicenter study was to investigate the association between preoperative/intraoperative traction and complications following growth-friendly instrumentation for EOS. METHODS Patients with EOS who underwent growth rod instrumentation before 2017 were identified from 2 registries. Patients were divided into 2 groups: preoperative traction group versus no preoperative traction group. A subgroup analysis was done to compare intraoperative traction only versus no traction. Data was collected on any postoperative complication from implantation to up to 2 years postimplantation. RESULTS Of 381 patients identified, 57 (15%) and 69 (18%) patients received preoperative and intraoperative traction, respectively. After adjusting for etiology and degree of kyphosis, there was no evidence to suggest that preoperative halo traction reduced the risk of any complication following surgical intervention. Although not statistically significant, a subgroup analysis of patients with severe curves demonstrated a trend toward a markedly reduced hardware failure rate in patients undergoing preoperative halo traction [preoperative traction: 1 (3.1%) vs. no preoperative traction: 11 (14.7%), P=0.083]. Nonidiopathic, hyperkyphotic patients treated with intraoperative traction were 61% less likely to experience any postoperative complication (P=0.067) and were 74% (P=0.091) less likely to experience an unplanned return to the operating room when compared with patients treated without traction. CONCLUSIONS This multicenter study with a large sample size provides the best evidence to date of the association between the use of traction and postoperative complications. Our results justify the need for future Level I studies aimed at characterizing the complete benefit and risk profile for the use of traction in surgical intervention for EOS. LEVEL OF EVIDENCE Level III.