Drug-Induced Sleep Endoscopy: Finding Where Your Airway Collapses
A sleep study proves your airway collapses. DISE is how a surgeon sees where. The evidence that it improves outcomes is real and more mixed than you would expect.
This entire site is built on one premise: work out where your airway narrows before choosing something to fix it. Drug-induced sleep endoscopy is the clinical procedure that actually answers that question, and most people considering airway surgery have never heard of it.
The problem it solves
A sleep study tells you that your airway collapses, and how badly. It says nothing about where.
That matters enormously, because the airway can obstruct at several levels — the soft palate, the tonsils and lateral pharyngeal walls, the tongue base, the epiglottis — and surgical procedures target specific levels. Operating on the palate when the collapse is at the epiglottis is a poor trade for an irreversible procedure.
Examining someone awake is a weak substitute. Awake examination and the Müller manoeuvre — sucking in with the mouth and nose closed while a scope watches — ask an alert person with active muscle tone to simulate what happens when that tone disappears. It is an approximation of the wrong state.
What the procedure involves
You are sedated, usually with propofol, to a depth intended to reproduce sleep-related muscle relaxation. A flexible endoscope passes through the nose and the surgeon watches the airway collapse in real time, recording which structures obstruct, in which direction, and how completely.
It is a day procedure under sedation rather than general anaesthesia, and it carries the ordinary risks of sedation. Several scoring systems exist for recording what is seen, which is itself a known weakness — there is no single agreed classification, and that limits how comparable findings are between centres.
What the evidence shows
A systematic review pooled studies comparing DISE-guided surgery against awake examination or the Müller manoeuvre: 8 studies, 880 patients (Sleep and Breathing, 2024).
Where DISE came out ahead:
- Greater improvement in lowest oxygen saturation: 6.83 ± 3.7 against 3.68 ± 2.9 (p < 0.001) [Sleep Breath 2024]
- Greater reduction in oxygen desaturation index: 19.6 ± 11.2 against 12.6 ± 10.4 (p < 0.001) [Sleep Breath 2024]
- Greater reduction in Epworth sleepiness score: 6.72 ± 4.1 against 3.69 ± 3.1 (p < 0.001) [Sleep Breath 2024]
On surgical success, defined by Sher's criteria as a postoperative AHI under 20 with at least 50% improvement, the difference was significant only against the Müller manoeuvre — 64.04% versus 52.48% (p = 0.016) [Sleep Breath 2024].
And the finding that complicates the picture: AHI reduction was higher in the non-DISE group — 39.92 ± 24.7 against 30.53 ± 21.7 (p < 0.001) [Sleep Breath 2024].
That last result deserves stating plainly rather than being left out, which is what most summaries do. DISE-guided patients had better oxygenation outcomes, better sleepiness outcomes and a higher success rate against one comparator — but a smaller absolute fall in AHI.
The likely explanation is selection rather than inferiority: groups with higher baseline AHI have more room to fall, and DISE tends to be used in more carefully selected or less severe candidates. But the review's design cannot separate that from a genuine difference, and pretending the number is not there would be dishonest.
Tier 2. Consistent benefit on oxygenation and symptoms, a modest success-rate advantage over the weakest comparator, and an unexplained AHI result. The scoring-system heterogeneity limits it further. How we grade evidence.
Its known limit
Even DISE does not fully predict who responds.
Roughly one-third of hypoglossal nerve stimulation patients are incomplete responders despite careful selection that includes drug-induced sleep endoscopy (Op de Beeck et al., AJRCCM, 2021).
So knowing where the airway collapses is necessary and not sufficient. Response also depends on physiological traits — arousal threshold, loop gain, muscle compensation — that no endoscope can see. Why treatments work for some people and not others.
There is also the question of whether sedated collapse faithfully reproduces natural sleep. Propofol is not sleep, and the depth of sedation affects what you see. It is closer than an awake exam, which is the honest claim, rather than being equivalent to natural sleep.
When it is worth asking about
Before any airway surgery. This is the main case. If a surgeon proposes a palate procedure without having established where your airway actually collapses, "would drug-induced sleep endoscopy change the plan?" is a fair question. Irreversible surgery aimed at the wrong level is the outcome DISE exists to prevent. Surgical options for snoring.
Before hypoglossal nerve stimulation, where it is part of standard selection — complete concentric palatal collapse is generally a contraindication, and that is a DISE finding.
When an oral appliance has failed unexpectedly. DISE can show whether jaw advancement opens your airway at all, and has been studied as a way of positioning the device. Oral appliance titration.
Not before trying CPAP. CPAP works regardless of collapse level, so where the airway closes does not change the decision to try it. DISE is a question for people heading toward site-specific treatment.
What it will not do
It will not tell you whether to have surgery, only where surgery would have to act. And the prior question remains whether a procedure is the right step at all — position, weight, alcohol and nasal patency are reversible and come first.
No consumer product addresses any of this. Tape, strips and pillows do not treat obstructive sleep apnea, and no amount of anatomical detail changes that. Snoring versus sleep apnea · Nasal strips versus surgery.