Diagnostic Performance of Airway Ultrasound for the Assessment of Difficult Laryngoscopy: A Systematic Review and Meta-Analysis.

Objective: This study aimed to review and appraise the evidence regarding airway ultrasound assessment in predicting difficult laryngoscopy in adult patients.

Design: A systematic review of the literature was conducted according to the Cochrane collaboration guidelines and the recommendations for the systematic review and meta-analysis of diagnostic studies. Observational studies that evaluated the diagnostic performance of airway ultrasound for the prediction of difficult laryngoscopy were included for consideration.

Setting: Literature searches were performed in 4 databases (PubMed [Medline], Embase, Clinical Trials, and Google Scholar) to identify all observational studies using any ultrasound technique to assess difficult laryngoscopy. The search terms included "sonography," "ultrasound," "airway," "difficult airway," "difficult laryngoscopy," "Cormack," "risk factors," "ultrasound at the point of care," "difficult ventilation," "difficult intubation" and others, combined with sensitive filters. The search was done for studies performed in the last 20 years in English or Spanish.

Participants: Adult patients older than 18 years old under general anesthesia for an elective procedure. Evident anatomic airway abnormalities, obstetric populations, those using an alternative imaging method besides ultrasound, and animal studies were excluded.

Interventions: Preoperative bedside ultrasound measuring distances and ratios from the skin to different reference points, such as the ratio of the hyomental distance in a neutral position (HMDN) and hyomental distance in extension (HMDR), HMDN, and the skin-to-epiglottis distance (SED), the preepiglottic area, and tongue thickness, among others.

Measurements And Main Results: A total of 24 studies evaluated the prediction of a difficult laryngoscopy using airway ultrasound. The diagnostic performance and the number of ultrasound parameters reported in the studies were variable. Meta-analysis was performed for 3 measurements consistently included in most studies. The SED and the HMDR ratio presented a sensitivity of 75% and 61%, respectively, and a specificity of 86% and 88%, respectively. The ratio of the preepiglottic distance to the epiglottic distance at the midpoint of the vocal cords (pre-E/E-VC) presented the best performance for predicting a difficult laryngoscopy (sensitivity: 82%, specificity: 83%, diagnostic odds ratio: 22.2).

Conclusion: With the currently available evidence, the 3 commonly used point-of-care ultrasound measures used to identify difficult laryngoscopy, (SED, HMDR, and pre-E/E-VC), showed better sensitivity and similar specificity to clinical measures. Future studies and more data may change the authors' confidence in these conclusions, given the wide variability of measurements noted in studies.