Preoperative anticipation and encountered difficult airway in children with osteogenesis imperfecta undergoing orthopedic surgery: a retrospective cohort study from April 2015 to March 2024

Abstract Background Children with osteogenesis imperfecta (OI) may present with difficult airway management, but the concordance between routine pre-anesthetic clinic (PAC) airway assessment and encountered intraoperative airway difficulty remains unclear. This study assessed concordance between PAC-documented anticipated difficult airway and encountered difficult airway in children with OI undergoing elective orthopedic surgery from April 2015 to March 2024 and described associated preoperative features and perioperative outcomes. Methods This retrospective cohort included 80 children with OI undergoing elective orthopedic surgery at a South Indian tertiary referral center. Reporting followed the STROBE statement and, given the partially verified concordance analysis, relevant STARD 2015 items. The primary concordance analysis included 37 patients with assessable reference outcomes: 35 undergoing direct laryngoscopy and tracheal intubation, and 2 managed with a supraglottic airway device who encountered definite difficult bag-mask ventilation. Concordance between PAC-documented anticipated difficult airway and encountered difficult airway was calculated among verified patients. Results PAC-documented anticipated difficult airway was present in 45 of 80 patients (56.2%). Encountered difficult airway occurred in 31 of 37 verified patients (83.8%). The concordance matrix was: true positive 21, true negative 6, false positive 0, and false negative 10. Among the verified subgroup ( n = 37), PAC assessment demonstrated sensitivity 67.7% (95% CI 50.1–81.4%), observed specificity 100.0% (95% CI 54.1–100.0%, zero false positives among verified patients), positive predictive value 100.0% (95% CI 83.9–100.0%), negative predictive value 37.5% (95% CI 18.5–61.4%), and false-negative rate 32.3% (95% CI 18.6–49.9%). Begg-Greenes correction, as an assumption-based sensitivity analysis, yielded a corrected false-negative rate of 32.7% under stated assumptions; however, the ignorable verification assumption may not be fully satisfied in this dataset. Mallampati Class III/IV, short neck, and blue sclera were associated with PAC-documented anticipated difficult airway in exploratory multivariable analysis. No intraoperative fractures or airway-related critical incidents occurred. Conclusions Among verified patients, no false-positive PAC assessments were observed, but the false-negative rate was 32.3%; approximately one-third of encountered difficult airways in this subgroup were not anticipated preoperatively. Ready access to videolaryngoscopy and difficult-airway adjuncts may be reasonable in similar settings despite reassuring PAC findings, pending prospective confirmation. Findings may not generalize to centers with milder OI phenotypes. Trial Registration Clinical Trials Registry–India: CTRI/2025/07/090079 (registered on July 2, 2025, before initiation of retrospective data abstraction). IEC: 246/2025, June 11, 2025.