Preoperative Airway Assessment: Purpose, Techniques, and Limitations
Purpose of Preoperative Airway Assessment
The primary purpose of preoperative airway assessment is to identify patients at risk for difficult airway management—not just difficult intubation, but also difficult mask ventilation and supraglottic device placement—to enable advance planning for maintaining oxygenation and ventilation. 1
The assessment serves to:
- Detect potential difficult direct laryngoscopy before induction
- Evaluate the probability of effective ventilation using supraglottic airway devices (oropharyngeal airway, laryngeal mask airway)
- Determine whether tracheal intubation is essential or if alternative strategies are appropriate
- Guide the decision between awake intubation, supraglottic device use, or inhalational induction with maintained spontaneous ventilation 1
Critical caveat: Preoperative assessment anticipates difficult airways in only 50% of patients who subsequently prove difficult to intubate 2. This fundamental limitation means clinicians must always prepare for the unexpected difficult airway regardless of assessment findings 2.
Conventional Bedside Airway Assessment
Standard bedside screening tests include:
Most Commonly Used Tests
- Modified Mallampati classification (most widely studied—105 studies) 3
- Thyromental distance (52 studies) 3
- Mouth opening test/interincisor distance (34 studies) 3
- Upper lip bite test (30 studies) 3
- Sternomental distance (18 studies) 3
- Wilson risk score (6 studies) 3
- Mentohyoid distance 4
Performance Characteristics
The upper lip bite test demonstrates the most favorable diagnostic accuracy among bedside tests, with sensitivity of 0.67 and specificity of 0.92 for difficult laryngoscopy—significantly higher sensitivity than mouth opening (0.22) 3. For difficult intubation specifically, the modified Mallampati test showed sensitivity of 0.51, significantly better than mouth opening (0.27) or thyromental distance (0.24) 3.
Recent evidence suggests the ratio of height to thyromental distance (RHTMD) may outperform traditional single parameters, achieving an AUC of 0.84 with adjusted odds ratio of 6.5, superior to both modified Mallampati (aOR 4.2) and sternomental distance (aOR 3.8) 5.
Limitations of Clinical Airway Assessment
Fundamental Diagnostic Inadequacy
All conventional bedside screening tests demonstrate relatively low sensitivities with high variability, though specificities are consistently higher 3. This creates a critical clinical problem: many difficult airways are missed by routine assessment 3.
Specific Limitations
Sensitivity failures:
- Modified Mallampati and thyromental distance are unreliable for predicting difficult tracheal intubation 2
- Bedside tests lack accuracy overall 2
- Routine assessment alone cannot reliably predict all cases of intraoperative difficult intubation 6
Timing of recognition: Unanticipated difficult intubation frequently becomes apparent only during laryngoscopy itself, characterized by poor visualization (Cormack-Lehane grade III-IV in 74.8% of difficult cases vs. 3.0% of controls) 6. The difficulty is often not evident until the blade is in the mouth 6.
Prediction vs. reality mismatch: Even when associated factors like ASA III status and airway deformity are identified, the discriminatory ability of combined clinical models remains modest (AUC 0.685) 6.
Clinical Implications
No single bedside test reliably excludes difficult intubation 7. The experience of the anesthesiologist plays an important role in clinical decision-making, as predicting difficulty remains "an imperfect science" 1.
Emerging Adjuncts
Point-of-care ultrasound shows promise as a supplement to clinical examination:
- Skin-to-epiglottis distance (SED) achieved AUC 0.912 with 85% sensitivity and 91% specificity in one study 4
- Provides dynamic anatomical assessment impossible with clinical examination alone 7
- Meta-analysis identified DSE (distance from skin to epiglottis) as most associated with difficult laryngoscopy 7
AI-based models demonstrate statistical superiority (facial image analysis: 80-90% sensitivity vs. Mallampati's 39%; deep learning AUC 0.84) but lack external validation and proven workflow integration 8.
Key Clinical Pitfall
The most dangerous assumption is that a normal preoperative assessment excludes difficult airway. Since 50% of difficult airways are unanticipated 2, every anesthetic plan must include contingency strategies for failed intubation and ventilation, regardless of assessment findings. The focus should remain on maintaining oxygenation and ventilation, not necessarily achieving intubation 1.