Accuracy, intra- and inter-rater reliability of three scoring systems for the glottic view at videolaryngoscopy

An accurate and reproducible recording of laryngoscopic view at tracheal intubation is an important aspect of anaesthetic practice. Unlike direct laryngoscopy, in which the view achieved by the line of sight directly relates to the ease of intubating the trachea, videolaryngoscopy can create a situation in which the view is good, but intubation difficult or impossible. Communicating this to a subsequent anaesthetist is important. We compared three scoring systems: Cormack and Lehane; POGO (percentage of glottic opening); and the Fremantle score, as used by 74 critical care doctors rating 30 anonymised videos of videolaryngoscopic intubations. Accuracy (degree of agreement of score with an expert panel assessment) was higher for POGO (75.5%) and the Fremantle score (73.9%) than for Cormack and Lehane (65.4%; p < 0.001). Intra-rater reliability (mean free marginal Kappa for ordinal scores and mean Cronbach's Alpha for continuous score) was higher for Fremantle score (0.796) and Cormack and Lehane (0.773) than POGO (0.693). Inter-rater reliability for Fremantle score (0.618) and POGO (0.614) were similar and higher than the inter-rater reliability of Cormack and Lehane 0.464 (p < 0.001). The higher accuracy and inter-rater reliability of POGO and the Fremantle score suggest they are preferable to Cormack and Lehane for use when documenting videolaryngoscopy. The additional information about ease of intubation conveyed by the Fremantle score may support its routine use in recording videolaryngoscopic intubation. © 2017 The Association of Anaesthetists of Great Britain and Ireland