The Snorer with Heart Failure and a Normal Home Study
A 63-year-old man with obesity (BMI 36), treated hypertension, and HFrEF (LVEF 30%) is referred for daytime sleepiness and nocturnal dyspnea. His wife reports loud snoring and occasional gasping, but no witnessed apneas. A home sleep apnea test (HSAT) using a type 3 device reports an AHI of 4.1 events/h and an ODI of 6/h; the study is technically adequate with 7.2 hours of recording. He has paroxysmal nocturnal dyspnea twice weekly and awakens with a choking sensation. Medications include carvedilol, sacubitril/valsartan, and furosemide. Neck circumference is 44 cm; modified Mallampati 3.
1. How do you interpret the HSAT result, and what is the next diagnostic step?
Read the worked answer
A negative HSAT in a patient with high pretest probability does not exclude OSA. The pretest probability here is very high: obesity, large neck, Mallampati 3, witnessed gasping, nocturnal choking, and HFrEF. HSAT also systematically underestimates events because it measures airflow and effort surrogates and cannot score hypopneas that require arousal, and it typically records total sleep time as time-in-bed rather than true sleep, which dilutes the AHI denominator.
The next step is in-laboratory attended polysomnography with EEG, EOG, EMG, airflow, effort, oximetry, ECG, and body position. In-laboratory PSG is specifically indicated when there is concern for central sleep apnea or Cheyne-Stokes respiration (likely in HFrEF with LVEF 30% and PND), when HSAT is negative but suspicion remains, and when comorbid conditions make HSAT unreliable. The PSG should be scored for central events and for a Cheyne-Stokes pattern, and the report should distinguish obstructive from central AHI because management diverges sharply.
2. The PSG shows an overall AHI of 28/h, of which 16/h are central events in a Cheyne-Stokes pattern, with a central apnea index of 16/h and nadir SpO2 78%. How does this change management compared with pure OSA?
Read the worked answer
This is predominant central sleep apnea with Cheyne-Stokes respiration (CSA-CSR) in the setting of HFrEF. The management priority is optimal guideline-directed heart failure therapy and evaluation for volume overload; CSR often improves with diuresis, afterload reduction, and, where indicated, cardiac resynchronization therapy. The patient should be referred to heart failure/cardiology.
CPAP is not the default answer. Adaptive servo-ventilation (ASV) improves the central apnea index and quality of life but the SERVE-HF trial showed increased cardiovascular mortality with ASV in patients with symptomatic HFrEF and LVEF <=45% with predominant CSA, so ASV is contraindicated in that population. Nocturnal oxygen and, in selected patients, other therapies may be considered. The exam trap is reflexively titrating positive pressure for an AHI of 28; here the type of event and the ejection fraction drive the decision. Document the LVEF, the predominant event type, and the heart failure optimization plan.
3. What if the same PSG had shown an AHI of 28/h with 26/h obstructive events and 2/h central events? Would your answer change?
Read the worked answer
Yes, substantially. With predominant OSA, the standard pathway applies: PAP therapy (CPAP) with titration, adherence support, and treatment of the OSA as a contributor to his nocturnal dyspnea and hypertension. In HFrEF, untreated OSA worsens afterload and sympathetic tone, so effective treatment is clinically meaningful. ASV would not be the first choice; the SERVE-HF concern applies to predominant CSA, not to OSA.
The teaching point is that the overall AHI alone is insufficient; the obstructive-to-central ratio determines therapy. A patient with AHI 28 can be a straightforward CPAP case or a heart-failure-driven CSA case depending entirely on event classification, which is why attended PSG with effort belts and ECG is required when central apnea is suspected. Also note that treatment-emergent central apnea can appear during CPAP titration, so a follow-up study may be needed if central events emerge on PAP.