The STOP-Bang screen & who should test

Eight yes-or-no questions, under a minute to answer, and a surprisingly good read on whether your sleep is quietly falling apart. This page walks through the STOP-Bang screen item by item — what it catches, what it misses, and the uncomfortable truth that a silent bedroom does not clear you of sleep apnea.

Why a Screen Exists at All

The numbers make the case. Roughly a billion adults worldwide are estimated to have obstructive sleep apnea, the large majority undiagnosed (Benjafield et al., Lancet Respiratory Medicine, 2019) — and the parent topic Sleep Apnea: The Under-Diagnosed Risk lays out why the condition hides so well. The bottleneck is diagnosis itself: lab sleep studies are limited capacity, and most primary-care visits never mention sleep. A cheap, fast paper tool — one a clinic can hand over in the waiting room — exists to solve the triage problem: deciding who is worth the time, money, and equipment of a sleep study. That is the whole job of STOP-Bang. It does not diagnose. It ranks.

The Eight Questions, Line by Line

STOP-Bang is an acronym; each letter is a yes/no question worth one point. The original was developed for pre-operative screening (Chung et al., Anesthesiology, 2008) and later refined (Chung et al., Chest, 2016). Here is what each item is really probing:

Notice the structure: four items are about symptoms, four are about risk factors. A person with no symptoms but four risk factors scores the same as a person with four symptoms and no risk factors — deliberately, because both profiles produce real disease.

Scoring: What the Tiers Mean

The conventional reading, from the validation literature, sorts scores into three tiers: 0–2 (low risk), 3–4 (intermediate), and 5–8 (high risk). The probability of moderate-to-severe OSA rises steeply through the range — the curve below is illustrative, but the shape is well supported by the meta-analyses in the sources.

Two honest caveats. First, the screen is designed to over-call: at the popular cutoff of 3+, sensitivity is high but specificity is modest, so plenty of people referred on a score of 3–4 turn out not to have apnea. That is a feature, not a bug — a screen should catch, and the sleep study sorts. Second, the tiers are probabilities, not boundaries: a person with a score of 2 and hard symptoms (morning headaches, unrefreshing sleep, treatment-resistant hypertension) has a higher pretest probability than the number alone suggests. The score refines judgment; it does not replace it.

Why "I Don't Snore" Does Not Rule Apnea Out

The most common self-dismissal in sleep medicine is also the flimsiest. Here is why the absence of snoring — or the absence of any noisy sign — settles nothing:

• 🤫 Snoring is one item of eight. The screen has seven other questions precisely because no single symptom is required. You can score 4 without snoring — and plenty of people do.
• 🙋 Hypopneas are quiet. Many events are shallow-breathing episodes rather than full airway closures. Airflow drops by half or more and oxygen follows — without the acoustic drama.
• 👩 Women snore less and get missed more. Across cohorts, women report snoring and witnessed pauses less often than men at the same AHI — a core reason their cases are under-flagged (see Not just obesity for the full profile).
• 🛌 You cannot hear yourself. The snoring item is validated against partner observation. If you sleep alone, the two most informative items (S and O) are effectively unscorable.
• 📐 Anatomy without acoustics. Young, lean people with narrow jaws and crowded throats can have severe disease with minimal snoring — the airway narrows to a slit and airflow falls without flapping tissue. The N item (neck circumference) exists for the people the S item misses.

The summary is blunt: self-reported "I don't snore" has a real false-reassurance rate. If the other items — fatigue, hypertension, weight, neck size, age — add up, the quiet bedroom proves nothing.

Who Should Test Even With a Low Score

The screen is one route in; the other is the condition list. Several diagnoses are so entangled with apnea that guidelines treat them as reasons to evaluate almost regardless of score:

• 💓 Resistant hypertension. Blood pressure that stays high on three drugs is a classic apnea marker — the Blood Pressure protocol flags it directly.
• 🫀 Atrial fibrillation. Apnea is a major reversible contributor to AF recurrence — see Apnea's metabolic toll.
• 🩸 Type 2 diabetes or prediabetes. Untreated apnea worsens glucose control independent of weight — the Glucose 101 page owns the blood-sugar side.
• 🚽 Nocturia and morning headaches. Two or more nightly bathroom trips and regular dawn headaches are underrated apnea signals.
• 🚩 Sleepiness that has no habit explanation. If the sleepiness red flags apply, evaluation is warranted even with a STOP-Bang of 2.
• 🚚 Occupational and pre-operative settings. Commercial drivers with sleepiness, and anyone heading into major surgery, are screened aggressively because the stakes of a missed case are immediate.

What Happens After a Positive Screen

A score of 3+, or a high-risk condition from the list above, routes to the same next step: a clinician conversation, then a sleep study. For most uncomplicated cases that means a home sleep test; for complex presentations — suspected central apnea, heart failure, neurological disease, or an inconclusive home result — a lab polysomnography. The trade-offs between the two are exactly what Home sleep tests vs lab polysomnography walks through. What matters now is the sequence: screen first, study second, and treat from your actual AHI number — not from a hunch.

Questions, Answered Briefly

• 📋 Can I score myself? Yes, with one fix: ask a bed partner (or record yourself) for the snoring and witnessed-pause items. Self-report alone underweights the two most informative questions.
• 🎯 What score means "get tested"? 3+ in most settings; 5+ strongly. Below that, let symptoms and the condition list — not the number — decide.
• ⌚ My smartwatch shows oxygen dips — does that count? A consumer SpO2 trace is not a sleep study and can both over- and under-call. But a consistent pattern of dips is a legitimate reason to get the real test.
• 🧮 I scored 8. Am I doomed? High scores mean high probability of moderate-to-severe disease — and unusually good treatment odds, because treatable apnea is among the most reversible diagnoses in medicine. The Beyond CPAP page covers what happens next.

Related Topics