The STOP-Bang score is an eight-item screening tool for estimating an adult's risk of obstructive sleep apnea (OSA). It is a risk-stratification tool, not a diagnosis, a measure of apnea severity, or a substitute for appropriate sleep testing 12.
Clinicians may use the score to decide who needs closer evaluation or extra perioperative attention. Its meaning depends on where it is used, which scoring rule is applied, and how common OSA is in the population being assessed 3.
What STOP-Bang stands for
STOP-Bang combines four symptom or medical-history items with four demographic or body-measurement items. Each “yes” receives one point, producing a raw score from 0 to 8 14.
| Letter | One point is added for |
|---|---|
| S: Snoring | Loud snoring, such as snoring louder than talking or heard through a closed door |
| T: Tiredness | Often feeling tired, fatigued, or sleepy during the day |
| O: Observed apnea | Someone having observed you stop breathing during sleep |
| P: Pressure | Having high blood pressure or receiving treatment for it |
| B: Body mass index | BMI greater than 35 kg/m² |
| A: Age | Being older than 50 |
| N: Neck circumference | Neck circumference greater than 40 cm, or about 15.75 inches |
| G: Gender | “Male gender” in the published questionnaire |
Forms may update the wording of the symptom questions or provide local instructions for measurements. Use the version supplied by the clinician or health system rather than mixing questions and cutoffs from different online calculators. If you do not know whether you snore or stop breathing because nobody observes your sleep, report that uncertainty instead of assuming the answer must be “no.”
How the score is interpreted
A raw score of 3 or more is a commonly used screen-positive threshold. It was chosen to catch a large share of people who have clinically important OSA, not to confirm that everyone above the cutoff has it 13.
A later two-step approach separates the raw score into three groups:
| Raw score | Common two-step interpretation |
|---|---|
| 0 to 2 | Lower risk of moderate-to-severe OSA in the studied adult populations |
| 3 to 4 | Intermediate risk that may need a second scoring step |
| 5 to 8 | Higher risk of moderate-to-severe OSA in the studied adult populations |
For a raw score of 3 or 4, the two-step method classifies risk as higher when at least two of the four STOP items are positive and any one of these is also present: BMI greater than 35 kg/m², neck circumference greater than 40 cm, or the questionnaire's male-gender item 4.
These are validated screening approaches, not universal diagnostic rules. A preoperative form may use a simple cutoff of 3 or more, while another service may use the two-step method or a different validated adaptation. Follow the rule attached to the form and the purpose for which it is being used.
Why one cutoff cannot give you a diagnosis
Screening performance involves a tradeoff between sensitivity and specificity:
- Sensitivity describes how often the screen is positive among people who truly have OSA at the apnea threshold being studied.
- Specificity describes how often the screen is negative among people who do not have OSA at that threshold.
A 2015 meta-analysis found that a cutoff of 3 or more had 94% sensitivity for moderate-to-severe OSA in sleep-clinic patients and 91% sensitivity in surgical patients. Specificity was only 34% and 32%, respectively 3. In practical terms, this cutoff missed relatively few cases in those studies but also flagged many people who did not have moderate-to-severe OSA.
Raising the cutoff or requiring a particular combination of items generally improves specificity but lowers sensitivity. Fewer people without OSA will screen positive, but more people with OSA may screen negative 45.
The chance that a result is a true positive also depends on the starting population. A person already referred to a sleep clinic because of loud snoring and severe sleepiness has a different pretest likelihood than an adult completing a questionnaire during routine intake. This is why a percentage attached to a score in one study should not be treated as your personal probability.
STOP-Bang also does not measure airflow, breathing effort, oxygen, carbon dioxide, sleep stages, or the number of breathing events per hour. It cannot tell you:
- whether you have OSA
- how mild or severe it is
- whether breathing events are obstructive or central
- whether another sleep, breathing, or medical condition explains your symptoms
- which treatment would be appropriate
The American Academy of Sleep Medicine (AASM) recommends against using a questionnaire or clinical prediction tool to diagnose adult OSA without polysomnography or an appropriate home sleep apnea test 2.
The setting changes what the score is for
Before surgery
STOP-Bang was developed and first validated in adults attending preoperative clinics. The Society of Anesthesia and Sleep Medicine recommends screening tools such as STOP-Bang to identify surgical patients with suspected OSA because many arrive without a diagnosis and there may be limited time for testing before a procedure 15.
In this setting, the score helps the perioperative team combine suspected OSA risk with the planned procedure, anesthesia, pain medicines, other health conditions, and postoperative monitoring needs 5. It does not confirm OSA or tell you to cancel a procedure. Give the completed form and any prior sleep-study or PAP information to the surgical and anesthesia teams, then follow their plan.
In a sleep clinic or symptom evaluation
People referred to a sleep clinic usually have symptoms, risk factors, or a prior abnormal finding, so their starting likelihood differs from that of an unselected population 3. STOP-Bang can organize those clues, but the clinician still needs the full history and an appropriate test. A low score should not end the evaluation when there is persistent concern about witnessed breathing pauses, gasping or choking, habitual loud snoring, or excessive daytime sleepiness 2.
In adults without a recognized concern
Screening every adult is a different question from evaluating someone who presents with symptoms. In 2022, the U.S. Preventive Services Task Force found insufficient evidence to determine the balance of benefits and harms of screening the general adult population for OSA. Its statement covers adults without recognized symptoms or whose symptoms have not been raised as a concern, and it does not apply to people who present for evaluation of suspected OSA 6.
The finding is not a recommendation to ignore symptoms. It means evidence from selected surgical and sleep-clinic groups cannot by itself prove that routine population-wide screening improves health outcomes.
Important limits for particular people
The standard STOP-Bang questionnaire was designed for adults. Do not transfer its cutoffs to children or teenagers. The USPSTF population-screening recommendation also excludes pregnancy and occupational fitness evaluations, which require their own evidence and clinical or regulatory pathways 6.
The published “G” item uses the wording “male gender.” That binary item does not provide a validated scoring rule for transgender, nonbinary, or gender-diverse people. Current perioperative guidance notes that hormone-related body changes and the uncertain roles of sex and gender may affect the score, and that its accuracy in transgender populations is unclear 7.
There is no evidence-based instruction to automatically score that item according to current gender, sex recorded at birth, hormone use, or whichever choice produces the higher score. Ask the clinician using the form how the local pathway handles the item. Regardless of how it is recorded, a binary point should not override concerning symptoms or a clinician's judgment.
Younger age, a BMI at or below 35, a neck circumference at or below 40 cm, or not receiving the male-gender point can all lower the raw score. None of those answers rules out OSA. They show why the questionnaire is a compact risk model rather than a complete assessment.
What to do with your result
Share the score together with the answers that produced it. The pattern can matter clinically even when the total is the same, as shown by scoring variants that give added weight to particular combinations 4. Also report symptoms, relevant medical conditions, medicines, prior sleep tests, and whether sleepiness affects driving or work.
A clinician may recommend polysomnography or a home sleep apnea test. For uncomplicated adults with signs and symptoms that indicate increased risk of moderate-to-severe OSA, the AASM supports either polysomnography or a technically adequate home test. If a home test is negative, inconclusive, or technically inadequate and concern remains, the guideline recommends polysomnography 2.
In-lab polysomnography is generally preferred when suspected OSA occurs with significant cardiorespiratory disease, possible respiratory-muscle weakness, awake or suspected sleep-related hypoventilation, chronic opioid use, a history of stroke, or severe insomnia 2.
Do not use a score to begin, stop, or change PAP, oxygen, an oral appliance, medicine, or another treatment on your own. The score does not establish the diagnosis or mechanism those treatments are meant to address.
If you are sleepy enough that you may doze while driving, do not drive. Arrange another way to travel or stop in a safe place if sleepiness begins on the road. Caffeine may briefly increase alertness but does not reliably prevent brief losses of consciousness in severe sleep deprivation 8.
A sleeping partner's report of repeated pauses, gasping, or choking deserves clinical follow-up, but STOP-Bang is not an emergency-triage tool. Call emergency services if someone is unresponsive, cannot breathe, or is not breathing normally 9.
The bottom line
STOP-Bang is useful because it turns eight common clues into a consistent adult OSA risk screen. Its value is in deciding what deserves further attention, especially before surgery. The score cannot diagnose OSA, assign personal severity, or make a low-risk label more important than persistent symptoms.
