[Evaluation of a protocol for predicting difficult airway in routine practice: interobserver agreement]

Area of Science:

• Anesthesiology
• Airway Management
• Clinical Prediction Models

Background:

• Predicting difficult airway intubation is crucial for patient safety during anesthesia.
• Routine assessment protocols aim to identify patients at risk for difficult intubation.
• Interobserver agreement and predictive model behavior are key metrics for protocol evaluation.

Purpose of the Study:

• To evaluate a protocol for routine assessment of difficult airway intubation risk factors.
• To measure interobserver agreement for various airway assessment factors.
• To assess the behavior of these factors in a predictive model for difficult intubation.

Main Methods:

• A cross-sectional study of 320 patients undergoing orotracheal intubation.
• Assessed interrater agreement for Mallampati score, thyromental distance, neck thickness, kyphosis, mouth size, macroglossia, and dental prosthesis.
• Compared assessments between preanesthesia examination and operating room by anesthetists and residents.
• Constructed a bivariate regression model to predict difficult intubation (≥3 attempts).

Main Results:

• Interobserver agreement (kappa indices) for most factors was poor (<0.6) between preoperative and operating room assessments.
• Factors like thyromental distance, small mouth, and kyphosis showed very low agreement (<0.21).
• No individual factor or combination with the Mallampati score significantly predicted difficult intubation.

Conclusions:

• Poor interobserver agreement exists for difficult airway prediction factors.
• Preoperative and intraoperative assessments by different clinicians show low agreement.
• Current individual predictive factors and their association with the Mallampati score are not effective for predicting difficult intubation.