Multivariable prediction models for difficult direct laryngoscopy: Systematic review and literature metasynthesis.

ABSTRACT

CONTEXT The difficult airway is an important scenario in anaesthesia due to the impact of its potential complications, and the difficulty in predicting its presence in current clinical practice. METHODS: Systematic review of articles in English and Spanish retrieved from MEDLINE (Ovid), LILACS and EMBASE up to March 2018. The search strategy was defined by the authors. The reviewers uploaded the studies to specially designed tables in order to qualitatively analyse the results of each paper. RESULTS: A total of 3602 studies were identified. Thirty-four of these were included in the qualitative review. The most commonly used definition of difficulty was the Cormack-Lehane 3 or 4 classification, with a weighted mean incidence of 7.23%. The most relevant finding was the methodological weaknesses in obtaining these scales. CONCLUSIONS: Available prediction models show limited discrimination, and weaknesses were detected in the methodology used to develop these prediction rules.