What is the purpose and interpretation of the STOP-BANG (Snoring, Tiredness, Observed apnea, high blood Pressure, Body mass index, Age, Neck circumference, Gender) score in assessing obstructive sleep apnea (OSA)?

STOP-BANG Score: Purpose and Interpretation

The STOP-BANG questionnaire is a validated screening tool for obstructive sleep apnea (OSA) that consists of eight dichotomous yes/no items (Snoring, Tiredness, Observed apnea, high blood Pressure, BMI >35, Age >50, Neck circumference >40cm, male Gender), with scores ranging from 0-8, but it should not be used as a standalone diagnostic tool due to its high sensitivity (>90%) but low specificity for detecting OSA. 1, 2

Scoring and Risk Stratification

The STOP-BANG score stratifies patients into risk categories based on their total points:

• Low risk (0-2 points): Probability of moderate-to-severe OSA is only 18%, with a 95% probability of excluding OSA at scores <2 3, 4
• Intermediate risk (3-4 points): Requires additional criteria for classification; a score ≥2 plus BMI >35 kg/m² indicates high risk 4
• High risk (5-8 points): Probability of moderate-to-severe OSA increases to 60% at scores 7-8, with specificity of 91% for OSA at scores ≥6 5, 3, 4

Diagnostic Performance Characteristics

The questionnaire demonstrates the following test characteristics when compared against polysomnography:

• Sensitivity: 93% for moderate-to-severe OSA (AHI ≥15) and 100% for severe OSA (AHI ≥30) at a cutoff score ≥3 4
• Specificity: Only 36% at high-risk cutoffs, resulting in substantial false positives 1
• Negative predictive value: 90% for moderate-to-severe OSA and 100% for severe OSA, making it excellent for ruling out disease 4
• Positive predictive value: Poor at 50.6% for moderate-to-severe OSA, limiting its ability to confirm diagnosis 6

Clinical Application Algorithm

For patients with suspected OSA, follow this approach:

1. Score 0-2: Low probability of OSA; consider alternative diagnoses for symptoms 3, 4

2. Score 3-4: Intermediate risk requiring further evaluation:

   • If BMI >35 kg/m², treat as high risk and proceed to sleep testing 4
   • Consider adding overnight pulse oximetry to improve risk stratification 3

3. Score 5-8: High probability of moderate-to-severe OSA; proceed directly to diagnostic sleep testing 5, 4

Mandatory Diagnostic Testing

All patients with STOP-BANG scores ≥3 require confirmatory testing with either polysomnography (gold standard) or home sleep apnea testing (HSAT) before initiating treatment, as screening tools alone cannot definitively diagnose OSA. 1, 2

The American Academy of Sleep Medicine explicitly states that clinical questionnaires produce too many false negatives to serve as diagnostic instruments for individual patients 1

Critical Limitations and Pitfalls

Do not rely solely on STOP-BANG for diagnosis without objective sleep testing. 2 Key limitations include:

• Low specificity leads to overdiagnosis, with 61 false negatives per 1,000 patients at 87% prevalence 1
• Population bias: Validated primarily in middle-aged, obese males in surgical settings, which may reduce accuracy in women, younger patients, and non-surgical populations 1, 7
• Geographic variation: Diagnostic accuracy varies by region, with notably lower performance (AUC 0.52) in East Asian populations compared to other regions (AUC >0.80) 8
• BMI cutoff: The original BMI >35 kg/m² threshold performs adequately even in Asian populations; lower cutoffs do not improve performance 6

Comparison to Other Screening Tools

The STOP-BANG outperforms alternative questionnaires:

• Berlin Questionnaire: Lower accuracy (AUC 0.50-0.64) with excessive false negatives 1
• Epworth Sleepiness Scale: Poor diagnostic accuracy (AUC 0.42-0.85) with sensitivity as low as 27% 1
• Neither alternative achieves clinically useful likelihood ratios (≥10 for positive, ≤0.1 for negative) 1

Practical Clinical Use

The STOP-BANG is most valuable for:

• Preoperative screening to identify patients requiring perioperative precautions for unrecognized OSA 5
• Triaging patients in sleep clinics to prioritize those needing urgent diagnostic testing 8
• Ruling out OSA in low-scoring patients (0-2), given the excellent negative predictive value 3, 4