Obstructive Sleep Apnea (OSA): What Counts as Treatment Failure

Defining OSA: The Diagnosis Baseline You Compare Against

OSA is confirmed when polysomnography or home sleep testing records an AHI of 5 or more events per hour alongside symptoms such as witnessed apneas, unrefreshing sleep, or daytime sleepiness. An AHI of 15 or higher meets diagnostic criteria with no additional symptoms required. The American Academy of Sleep Medicine (AASM) 2023 clinical practice guidelines establish these exact cut-points and classify disease as mild (AHI 5-14), moderate (AHI 15-29), or severe (AHI >30) [1].

Knowing the entry-level AHI matters because treatment failure is always measured relative to it. A patient who starts with an AHI of 48 and achieves a treated AHI of 18 has improved substantially but has not reached the goal of AHI <5. That gap is a form of partial failure, and it changes the clinical plan.

Why the 5-Events-Per-Hour Target Matters

Residual AHI >5 on therapy correlates with continued cardiovascular risk. The Sleep Heart Health Study (N=6,441) found that even mild untreated OSA (AHI 5-14) carries an adjusted odds ratio of 1.42 for hypertension compared with AHI <5 [2]. Accepting a residual AHI above 5 means accepting a measurable portion of that cardiovascular burden, which is why most guidelines refuse to declare treatment successful at higher thresholds.

Severity and Symptom Mismatch

Some patients with AHI >30 report minimal sleepiness, while others with AHI of 8 report disabling fatigue. The Epworth Sleepiness Scale (ESS) score provides a standardized symptom anchor. An ESS above 10 suggests clinically significant daytime sleepiness regardless of the objective AHI reading, and persistent ESS >10 on treatment is its own marker of treatment failure even when device data look acceptable [3].

What Counts as CPAP Treatment Failure

CPAP is the most prescribed first-line therapy for OSA, but failure is common and multidimensional. CPAP treatment failure exists when any one of three measurable conditions persists after a reasonable trial period of at least 90 days.

The Three Failure Criteria

1. Residual AHI >5/hr on device data. Modern CPAP machines record nightly AHI through auto-titrating algorithms. A 30-day download showing a median residual AHI above 5 events per hour means the device is not adequately controlling airway obstruction at the current pressure settings. Mask leak, pressure settings, and positional factors all contribute and must be ruled out before labeling this true CPAP failure.

2. Adherence below the CMS threshold. The Centers for Medicare and Medicaid Services define adequate CPAP use as at least 4 hours per night on more than 70% of nights over any consecutive 30-day period [4]. Patients who do not meet this standard cannot be evaluated for therapeutic efficacy because the dose delivered is insufficient. CPAP non-adherence affects 30-50% of patients within the first year of therapy [5].

3. Persistent symptoms despite acceptable device use. A patient may run 6 hours per night with a residual AHI of 3 yet still score 13 on the ESS. That combination points to coexisting disorders including periodic limb movement disorder, insufficient sleep syndrome, or narcolepsy, but it also constitutes failure of CPAP alone to resolve the clinical problem.

Common Reasons CPAP Fails

Mask interface is the single biggest driver of abandonment. Pressure-related central apneas (complex sleep apnea syndrome, now called treatment-emergent central apnea) appear in roughly 5-15% of CPAP initiators and require switching to adaptive servo-ventilation (ASV) or bilevel PAP [6]. Claustrophobia, nasal congestion, and bed-partner disruption account for a large share of the remainder.

Oral Appliance Therapy: When It Works and When It Falls Short

Mandibular advancement devices (MADs) are the primary CPAP alternative and carry AASM and American Academy of Dental Sleep Medicine endorsement for mild-to-moderate OSA and for patients who prefer them or cannot tolerate CPAP. A 2015 Cochrane review (26 trials, N>900) confirmed that MADs reduce AHI and daytime sleepiness compared with inactive controls, though with a smaller AHI reduction than CPAP [7].

Defining Oral Appliance Failure

Oral appliance therapy fails when post-titration polysomnography or home sleep testing still records an AHI above 5. The titration window is 6-12 weeks of incremental advancement. If the residual AHI remains above 5 at maximal tolerated advancement, a switch to a different modality is appropriate. Temporomandibular joint discomfort limits titration in roughly 8% of patients, and that mechanical barrier itself constitutes device failure [8].

Who Should Not Rely on an Oral Appliance Alone

Patients with severe OSA (AHI >30) achieve AHI <5 with MADs in only about 30-40% of cases, compared with 70-80% for CPAP at optimal pressures [7]. For this subgroup, oral appliance therapy should be framed as a step-down option when CPAP truly cannot be used, not a primary equivalent.

Surgical Options and Neurostimulation: Failure Thresholds and Eligibility

When PAP therapy and oral appliances both fail, surgery or implantable neurostimulation becomes the evidence-based path. The AASM 2023 guidelines specify that surgical candidacy evaluation is appropriate after documented PAP failure [1].

Hypoglossal Nerve Stimulation (HNS)

The Inspire Upper Airway Stimulation system is FDA-approved for adults with moderate-to-severe OSA (AHI 15-65/hr) who have failed CPAP and lack complete concentric collapse at the velum on drug-induced sleep endoscopy. The STAR trial (N=126, 12-month follow-up) reported a median AHI reduction from 29.3 to 9.0 events per hour, with 68% of participants meeting the composite response criterion of >50% AHI reduction and residual AHI <20 [9].

Failure of HNS is defined post-operatively the same way as any other therapy: residual AHI >5 and persistent ESS >10 at the 6-month programming optimization visit.

Upper Airway Surgery

Uvulopalatopharyngoplasty (UPPP) alone produces a surgical success rate (AHI reduction >50% and residual AHI <20) in about 50% of carefully selected patients, well below the rates achieved by PAP and HNS [10]. Multi-level surgery combining UPPP with tongue-base procedures improves that figure modestly. The AASM defines surgical failure using the same residual AHI and symptom thresholds applied to device therapy.

Weight Loss as Disease-Modifying Therapy: Tirzepatide and the SURMOUNT-OSA Data

OSA and obesity are deeply linked. Pharyngeal fat deposits narrow the upper airway, and adipose tissue in the chest wall reduces functional residual capacity. Each 10% reduction in body weight reduces AHI by approximately 26%, based on pooled data from the Sleep AHEAD trial and related cohorts [11].

FDA Approval of Tirzepatide (Zepbound) for OSA

In January 2024, the FDA approved tirzepatide (Zepbound, Eli Lilly) specifically for moderate-to-severe OSA in adults with obesity, making it the first pharmacologic agent approved for OSA as a primary indication [12]. This approval was anchored on the SURMOUNT-OSA trial.

SURMOUNT-OSA Trial Results

SURMOUNT-OSA enrolled 469 adults with moderate-to-severe OSA and obesity (BMI >30 kg/m²) across two parallel cohorts: one using CPAP alongside tirzepatide and one not using CPAP. At 52 weeks, tirzepatide 10 mg or 15 mg weekly reduced AHI by a mean of 27.4 events per hour in the non-CPAP cohort (62.8% reduction from a baseline AHI of 43.7) versus 4.8 events per hour with placebo. In the CPAP cohort, the reduction was 25.1 events per hour (62.2%) versus 5.0 with placebo [13]. Those are large, clinically meaningful numbers.

The table below summarizes how to apply SURMOUNT-OSA findings to the treatment-failure decision.

| Clinical scenario | Residual AHI on current therapy | Recommended next step | |---|---|---| | CPAP-adherent, residual AHI <5, ESS <10 | Controlled | Continue; annual review | | CPAP-adherent, residual AHI 5-15, ESS <10 | Partial failure | Pressure re-titration; add positional therapy | | CPAP non-adherent (<4 hrs/night), any AHI | Adherence failure | MAD or HNS evaluation; tirzepatide if BMI >30 | | CPAP failure + BMI >30 + moderate-severe OSA | Device and weight | Tirzepatide; re-assess need for PAP at 6 months | | CPAP failure + normal BMI | Device failure, no obesity | HNS if AHI 15-65; surgical evaluation |

Weight-Loss Treatment Failure

Weight-loss therapy for OSA fails when 12 months of lifestyle intervention or pharmacotherapy produces less than 5% total body weight loss, since that threshold rarely produces a clinically significant AHI reduction [11]. At that point, escalation to bariatric surgery or a different GLP-1/GIP agent is appropriate.

Positional and Adjunctive Therapy: Modest Evidence, Specific Indications

Positional OSA, defined as an AHI at least twice as high in the supine position compared with non-supine, affects roughly 50-60% of OSA patients. Positional therapy devices (wearable vibrotactile trainers such as the Night Shift or SPT) reduce supine sleep time and can cut overall AHI by 30-50% in this subgroup [14].

Positional therapy fails when post-device polysomnography shows no significant AHI change, or when the patient's non-supine AHI is itself above 5 (indicating non-positional disease). Combining positional therapy with an oral appliance occasionally achieves AHI <5 in patients who failed each alone.

Monitoring Protocols: How to Detect Failure Early

Treatment failure is not a single event. It develops over months, sometimes years, as anatomy changes, weight shifts, or adherence erodes. AASM and the American Thoracic Society recommend structured follow-up to catch failure before cardiovascular consequences accumulate [1].

Follow-Up Schedule

• Weeks 4-8: First CPAP download review. Confirm residual AHI <5 and mask leak below 24 L/min. ESS re-administered.
• Month 3: Full adherence review. If below CMS threshold, identify barriers and intervene.
• Month 6: Repeat home sleep test if clinical suspicion of residual disease despite good device data.
• Annual: Weight change assessment, ESS, device download. If weight has increased by more than 10%, re-titration is often needed.

Objective vs. Subjective Discordance

When device data show excellent control (residual AHI <5, good adherence) but the patient remains subjectively sleepy, the clinician must investigate alternatives. The AASM guideline states: "Residual excessive sleepiness after adequate treatment of OSA should prompt evaluation for other sleep disorders, circadian rhythm disorders, or medical comorbidities" [1]. Acting on device data alone without re-evaluating the patient is a process failure, not just a treatment failure.

Combination Therapy Strategies After First-Line Failure

Single-modality failure does not exhaust options. Combination approaches are supported by evidence and increasingly used.

CPAP Plus Weight Loss

Adding tirzepatide to ongoing CPAP in SURMOUNT-OSA reduced AHI by an additional 25 events per hour versus CPAP plus placebo [13]. This makes pharmacologic weight loss a genuine adjunct rather than a replacement therapy, particularly for patients who tolerate CPAP but remain symptomatic or want to reduce dependence on the device.

Oral Appliance Plus Positional Therapy

A 2020 randomized trial (N=50) published in CHEST found that combining a MAD with a positional device produced a response rate (AHI <5) of 64% compared with 48% for MAD alone in patients with positional OSA who had previously failed MAD monotherapy [14].

Bilevel PAP and ASV After CPAP Failure

Bilevel PAP (BPAP-S) is appropriate when pressure requirements exceed 15 cmH2O or when the patient cannot tolerate high fixed pressures. Adaptive servo-ventilation is reserved for treatment-emergent central apnea and must be avoided in patients with an ejection fraction below 45% due to the increased mortality seen in the SERVE-HF trial (N=1,325; hazard ratio 1.28 for cardiovascular death, P<0.01) [15].

When to Refer: Escalation Triggers at a Glance

Physicians managing OSA in primary care or telehealth should trigger a sleep specialist or surgical referral when:

• Residual AHI remains above 5 after two rounds of PAP pressure adjustment
• The patient has failed or refused both CPAP and an oral appliance
• AHI is 15-65/hr and the patient meets Inspire eligibility criteria
• There is clinical suspicion of complex sleep apnea (rising central events on CPAP download)
• BMI is above 30 and moderate-to-severe OSA persists despite at least 5% weight loss