Why the Tube Does Not Follow the View: A Video-based Analysis of Tube Delivery Difficulty during Videolaryngoscopy

Background and Aims: Videolaryngoscopy (VL) improves glottic visualisation, yet successful endotracheal intubation may remain challenging. Although this discrepancy is well recognised, the underlying patterns of tube-delivery difficulty (TDD) remain poorly described. Using recorded VL intubations, we aimed to describe patterns of TDD and assess whether these patterns could be reproducibly identified during VL-assisted intubation using a structured classification system. Materials and Methods: In this retrospective observational study, 176 VL recordings obtained during routine clinical practice were reviewed. Videos with adequate visualisation of the glottic inlet and complete recording of the tube-delivery phase were included. TDD was defined as any deviation from smooth tube advancement despite an adequate glottic view. Tube-delivery events were categorised into predefined phenotypes based on trajectory and site of obstruction. Two independent reviewers assessed TDD and phenotype classification. Interrater reliability was evaluated using Cohen’s kappa. Results: Of 176 screened recordings, 150 were included. TDD was identified in 67 (44.7%) cases. The most frequent pattern was excessive posterior trajectory towards the oesophageal inlet (29.9%), followed by subglottic advancement failure after passage through the vocal cords (22.4%), posterior impingement at the arytenoid level (17.9%) and lateral misplacement into the pyriform fossa (14.9%). Interrater agreement was substantial for identification of TDD (observed agreement 88.7%; κ = 0.74) and video-quality grading (86.1%; κ = 0.68) and moderate for phenotype classification (78.0%; κ = 0.61). Conclusion: VL recordings identified distinct, reproducible patterns of TDD. Recognition of these patterns may improve understanding of tube-delivery dynamics and may assist in documentation and future research in difficult airway management.