ASV Requested for Central Sleep Apnea With an Ejection Fraction of 28%
A 71-year-old man with ischemic cardiomyopathy (left ventricular ejection fraction 28%, NYHA class III), treated with sacubitril-valsartan, carvedilol, and spironolactone, is referred for initiation of adaptive servo-ventilation. Diagnostic polysomnography shows an AHI of 42/hr with 86% central events, a Cheyne–Stokes pattern in both NREM and REM, a nadir SpO2 of 81%, and 30% of total sleep time below 90%. The supine central apnea index is 38/hr. CPAP at 8 cm H2O was tried 4 months earlier for 3 weeks with no improvement, though adherence data were never reviewed. His cardiologist reports no diuretic change and an ejection fraction that has been 25–30% for 2 years. The patient is desperate for relief, and the referring sleep physician has written "proceed with ASV titration tonight, EPAP minimum 3 cm H2O."
1. Should the technologist proceed with the ASV titration? Discuss the evidence and the contraindication.
Read the worked answer
No. This is a case where the correct action is to stop and contact the ordering physician before any device is applied.
The relevant evidence is the SERVE-HF trial, which enrolled patients with heart failure with reduced ejection fraction and predominantly central sleep apnea and found that adaptive servo-ventilation increased all-cause and cardiovascular mortality compared with control. That result produced a clear contraindication: ASV should not be used in patients with symptomatic heart failure with reduced ejection fraction, conventionally defined as an ejection fraction of 45% or less, who have predominantly central sleep apnea. This patient has an ejection fraction of 28%, NYHA class III symptoms, and 86% central events. He sits squarely inside the contraindication.
An order from a physician does not transfer the responsibility to apply a device that is contraindicated. The technologist's role is to stop, document the ejection fraction, the event composition, and the guideline basis for concern, and reach the ordering physician by phone that night. If the physician insists, ask for the documentation of that decision in the record and escalate to the medical director; do not proceed silently.
The patient's desperation is understandable and should be acknowledged directly, but it does not change the risk. What he needs is a plan, not a night on a contraindicated device.
2. What alternatives and prerequisites should be recommended, and how would you distinguish treatment-emergent central apnea from primary central sleep apnea here?
Read the worked answer
Prerequisites before choosing any device. The prior CPAP trial is not interpretable: 3 weeks is short and adherence and residual event data were never reviewed. Obtain the download to see whether the pressure actually controlled obstructive events, whether leak was high, and whether the patient used the device at all. If CPAP was never effectively delivered, the central events may never have had a fair trial of treatment.
Separate the two entities. Treatment-emergent central apnea appears when obstructive events are eliminated on CPAP but central apneas persist or emerge, typically resolving over weeks to months as the patient stabilizes. Primary central sleep apnea with Cheyne–Stokes in a patient with a low ejection fraction is a different process driven by circulatory delay and instability of ventilatory control. The distinction matters because treatment-emergent central apnea often resolves with continued effective PAP, whereas Cheyne–Stokes in heart failure needs heart failure optimization and non-ASV options.
Alternatives to recommend. First, optimize guideline-directed heart failure therapy with the cardiology team, since improving cardiac function is the most direct lever on Cheyne–Stokes. Second, consider nocturnal oxygen, which can reduce central event frequency and improve saturation modestly. Third, consider CPAP for any coexisting obstructive component. Fourth, screen for other contributors such as opioids, atrial fibrillation, and prior stroke. Bilevel with a backup rate can be considered case by case, but only with cardiology input and not as a reflex substitution for ASV.
Documentation. Record that ASV was declined, why, who was notified, and what was recommended, so the decision is visible to the next clinician.
3. Three months later, after heart failure optimization, the ejection fraction is 48% and central apnea persists with an AHI of 32/hr on no therapy. How does your recommendation change?
Read the worked answer
The contraindication no longer applies. With an ejection fraction above 45%, the population in which ASV was shown to be harmful no longer describes this patient, so ASV becomes a reasonable option rather than a prohibited one. Document the current ejection fraction with the date and source; an ejection fraction is a time-stamped value, not a permanent attribute.
Before titrating, confirm the rest of the picture. Verify that the ejection fraction improved on current therapy and is stable, confirm that central events still predominate on a current study rather than relying on the older tracing, review symptoms and adherence history, and re-screen for opioids and other central apnea drivers. Confirm the Cheyne–Stokes pattern persists, because ASV is designed for periodic breathing with central events.
Titration approach. Proceed with an in-laboratory ASV titration as ordered, starting with a low EPAP (around 3–5 cm H2O) and letting the device deliver the variable pressure support needed for hypopneas and central apneas. Titrate to eliminate central and obstructive events, keep SpO2 at 90% or above, and avoid excessive pressure support that can cause discomfort or central event exaggeration. Document the settings, the residual event index in supine REM, and the leak.
Follow-up. Arrange a download at 4 weeks and repeat echocardiography at intervals agreed with cardiology, since the safety margin depends on maintaining an ejection fraction above 45%. If cardiac function deteriorates again, the ASV decision must be revisited, and that plan should be written into the report rather than assumed.