Difficult facemask ventilation remains poorly classified. Existing definitions rely on expert opinion and subjective rating of surrogates. This lack of standardization has led to highly variable reported incidences and inconsistencies in clinical practice and research.
This secondary analysis of the prospective observational MASCAN study aims to develop and validate a data-driven objective classification system for difficult facemask ventilation and to determine its diagnostic performance and calibration. Facemask ventilation was facilitated after anaesthesia induction in all patients. An independent observer systematically assessed potential indicators for difficult facemask ventilation that serve as candidate predictor variables for the fitting of a diagnostic multivariable logistic regression model and simplified score to classify difficult facemask ventilation. Cross-validated LASSO regression will be used for variable selection. The area under the receiver operating characteristic curve (AUROC) and calibration curves will be used to quantify the diagnostic performance and calibration, and optimal decision thresholds will be defined.