AI-assisted videolaryngoscopy improves intubation success: results from a randomized simulation trial

Area of Science:

• Anesthesiology
• Medical technology
• Artificial intelligence in medicine

Background:

• Video laryngoscopy is crucial for airway management but faces challenges like depth perception loss and cognitive load.
• Existing video laryngoscopes may not succeed in all airway scenarios.

Purpose of the Study:

• To evaluate the efficacy of larynGuide, an AI software integrated with C-MAC video laryngoscopy, in improving tracheal intubation success rates.
• To assess the impact of AI on both easy and difficult airway simulations.

Main Methods:

• A prospective, randomized, 2x2 crossover, simulation-based study involving 74 medical professionals.
• Participants performed tracheal intubations with and without larynGuide AI overlay on a C-MAC video laryngoscope.
• Primary outcome: first attempt success rate. Secondary outcomes: overall success, intubation time, performance scores, dental injury incidence.

Main Results:

• The overall first attempt success rate was 89% with larynGuide versus 85% with standard C-MAC, with 72% higher odds of success using larynGuide (OR=1.72, p=0.022).
• The benefit was more pronounced in experienced participants (OR=3.37, p=0.016).
• Time to intubation was longer with larynGuide (29s vs 23s, p<0.01), and dental injury incidence was lower (22% vs 27%, p=0.095).

Conclusions:

• larynGuide AI software significantly improved tracheal intubation success rates in a simulated environment compared to standard C-MAC videolaryngoscopy.
• AI shows potential for enhancing airway management, with greater benefits observed in experienced professionals.
• Further clinical studies are warranted to confirm these findings and support wider clinical adoption.