Intraoperative Antibiotic Use in Patients With Early-onset Scoliosis: Current Practices and Trends

Abstract

Background: Best Practice Guidelines (BPGs) regarding antibiotic prophylaxis in early-onset scoliosis (EOS) patients were published in September 2019. Recommendations included using intravenous cefazolin and topical vancomycin for all index procedures, plus gram-negative coverage for neuromuscular patients. Guideline adherence is unknown. This study aimed to characterize antibiotic prophylaxis at the time of index growth-friendly procedures and assess changes in practice patterns over time. Materials and Methods: This retrospective review of data collected through a multicenter study group included EOS patients undergoing index growth-friendly procedures between January 2018 and March 2021, excluding revisions, lengthenings, and tetherings. Demographics, clinical measurements, intraoperative antibiotics, and 90-day complications were recorded. Descriptive and univariate statistics were utilized. Antibiotic prophylaxis from April 2018 through September 2019 and October 2019 through March 2021 were compared with evaluate change after BPG publication. Results: A total of 562 patients undergoing growth-friendly procedures were included. The most common scoliosis types included neuromuscular (167, 29.7%), syndromic (134, 23.8%), and congenital (97, 17.3%). Most index procedures involved magnetically controlled growing rods (417, 74%) followed by vertical expandable prosthetic titanium rib or traditional growing rods (105, 19%). Most patients received cefazolin alone at index procedure (310, 55.2%) or cefazolin with an aminoglycoside (113, 20.1%). Topical antibiotics were used in 327 patients (58.2%), with most receiving vancomycin powder. There was increased use of cefazolin with an aminoglycoside after BPG publication (16% vs. 25%) (P=0.01). Surgical site infections occurred in 12 patients (2.1%) within 90 days of index procedure, 10 pre-BPGs (3%), and 2 post-BPGs (0.9%), with no significant difference in surgical site infection rate by type of antibiotic administered (P>0.05). Conclusions: Historical variability exists regarding antibiotic prophylaxis during index growth-friendly procedures for EOS. There continues to be variability following BPG publication; however, this study found a significant increase in antibiotic prophylaxis against gram-negative bacteria after BPG publication. Overall, greater emphasis is needed to decrease variability in practice, improve compliance with consensus guidelines, and evaluate BPG efficacy. Level of Evidence: Level III—retrospective.