How should I manage a patient with obstructive sleep apnea who refuses continuous positive airway pressure (CPAP) therapy, considering their heart failure treatment (including a sodium‑glucose cotransporter‑2 inhibitor and a loop diuretic) and elevated hematocrit?

Offer a Mandibular Advancement Device as First-Line Alternative

For a patient with OSA who refuses CPAP, you should recommend a custom-made mandibular advancement device (MAD) as the primary alternative therapy 1, 2. This is the most evidence-based non-CPAP option available and is specifically endorsed by guidelines for CPAP-intolerant or refusing patients.

Why MAD is the Best Alternative

The American College of Physicians explicitly recommends MADs as alternative therapy for patients who refuse CPAP or experience adverse effects, though this is a weak recommendation based on low-quality evidence 1. More importantly, the 2021 European Respiratory Society guideline analyzed 13 RCTs comparing CPAP to custom-made dual-block advancement devices—the highest level of evidence for any non-CPAP therapy 2. While CPAP showed superiority in reducing AHI, patient-related outcomes were equal between the two treatments 2.

Key Points About MAD Efficacy:

• Works best in patients with AHI scores between 18-40 events/hour 1
• Reduces AHI, though less effectively than CPAP
• Achieves similar patient-reported outcomes and quality of life improvements as CPAP 2
• Better long-term adherence than CPAP in many patients due to higher tolerance 3

Additional Considerations for This Specific Patient

The Heart Failure Connection

Your patient's heart failure and current treatment with SGLT-2 inhibitors is actually relevant. Recent evidence shows SGLT-2 inhibitors may independently improve OSA severity through diuresis, weight loss, and reduction in fluid retention 4, 5. A 2026 meta-analysis demonstrated SGLT-2 inhibitors significantly reduce AHI (MD = -5.52, p = 0.01), oxygen desaturation index, and BMI 5. The improvement in OSA appears more pronounced at 3 months, correlating with body weight and fat mass reduction, while body water content reduction shows sustained benefit at 6 months 4.

This means your patient may already be experiencing some OSA improvement from their current cardiac medications, though this alone is insufficient treatment.

The Elevated Hematocrit Issue

The elevated hematocrit likely reflects chronic hypoxemia from untreated OSA. This is a red flag for cardiovascular risk and makes treatment more urgent. While CPAP would be ideal for rapidly correcting nocturnal hypoxemia, a MAD can still provide meaningful improvement in oxygen saturation 1.

Algorithmic Approach to Management

1. Explore reasons for CPAP refusal - Address specific concerns (claustrophobia, mask discomfort, noise, partner issues) as some may be modifiable with different interfaces or settings 1

2. If CPAP truly refused, proceed with MAD:

   • Refer to dentist experienced in sleep medicine for custom-made dual-block device
   • Ensure proper titration and follow-up
   • Expect gradual improvement over weeks

3. Optimize concurrent therapies:

   • Continue SGLT-2 inhibitor and loop diuretic for synergistic OSA benefit 4, 5
   • Aggressively pursue weight loss if patient is obese—this has independent benefits for OSA and is strongly recommended 1

4. Consider positional therapy if patient has position-dependent OSA (can be assessed from sleep study data) - vibratory positional devices can be used alone or combined with MAD 2

5. Reserve surgical options (hypoglossal nerve stimulation, maxillomandibular advancement) for MAD failure, as these require careful patient selection and have higher risks 2, 3

What NOT to Do

• Do not prescribe pharmacologic agents for primary OSA treatment—evidence is insufficient and this is explicitly not recommended 1
• Do not recommend surgery as initial treatment—associated with risks and serious adverse effects without sufficient evidence of benefit 1
• Do not accept "no treatment"—untreated severe OSA in a heart failure patient significantly increases cardiovascular morbidity and mortality risk 6

Follow-Up Strategy

Monitor AHI reduction with follow-up sleep study at 3-6 months after MAD titration. Track subjective improvement in daytime sleepiness (Epworth Sleepiness Scale), hematocrit normalization, and blood pressure control. Recent data from multidisciplinary OSA clinics show significant reductions in AHI (-44.1%), ESS (-32.9%), and ODI (-67.2%) with tailored non-CPAP strategies 7.

If MAD fails or provides insufficient improvement, revisit CPAP discussion with emphasis on newer interfaces and support strategies, or consider referral to specialized sleep surgery center for evaluation of hypoglossal nerve stimulation 2, 7.