Split-Night Titration Pressure Gap in a Patient with Heart Failure
A 63-year-old man with HFrEF (LVEF 30%) and central sleep apnea is admitted for a split-night study. Diagnostic portion: AHI 48/hr, central apnea index 21/hr, nadir SpO2 79%, 22% of total sleep time below 88%. The technologist transitions to titration at 02:10 after only 2.5 hours of diagnostic recording. During titration, CPAP at 8 cmH2O eliminates obstructive events but central apneas persist at 14/hr with Cheyne-Stokes pattern; at 10 cmH2O central index rises to 19/hr and the patient reports dyspnea. The study ends at 06:00. The ordering physician expects a single fixed-pressure recommendation by morning rounds.
1. Was the split-night protocol appropriately applied here, and what does the diagnostic adequacy issue mean for the titration recommendation?
Read the worked answer
Split-night criteria generally require a diagnostic portion showing moderate-to-severe OSA (AHI ≥ 20–40 depending on policy) with sufficient recording time (commonly ≥ 2 hours of diagnostic sleep) before transitioning. Here the diagnostic AHI of 48/hr meets severity thresholds, but only 2.5 hours of diagnostic time was captured and the recording is dominated by central events, not OSA. The key judgment: the diagnostic portion is borderline adequate for OSA severity but inadequate to characterize the central/Cheyne-Stokes component across sleep stages and supine positions. A fixed-pressure recommendation derived from 3.5 hours of titration in a patient with HFrEF and treatment-emergent central apnea is not defensible. The correct action is to flag the study as technically limited for pressure determination and recommend a full-night titration (or ASV evaluation per current guidance) rather than issuing a fixed CPAP number.
2. At 8 cmH2O obstructive events are controlled but central apneas persist at 14/hr and worsen at 10 cmH2O. How should the technologist and interpreting clinician reason about the pressure response?
Read the worked answer
The pattern is treatment-emergent/worsening central apnea with increasing positive pressure - a recognized phenomenon in HFrEF patients with Cheyne-Stokes respiration. Raising pressure further is likely to worsen the central index (already 19/hr at 10 cmH2O) and provoked dyspnea, so escalating CPAP is the wrong move. The reasoning chain: (1) obstructive events are controlled at 8 cmH2O, so the residual events are central, not a pressure-deficit problem; (2) the dose-response is inverse - more pressure, more central events; (3) residual central index of 14/hr is clinically significant and above typical treatment targets. The appropriate conclusion is that CPAP is not the right modality for the central component; document the pressure-response data and recommend reassessment for ASV or other central-apnea therapy, with cardiology coordination given LVEF 30%. The exam-relevant judgment is recognizing that 'control the obstructive events' does not equal 'the titration succeeded.'
3. What should the final report recommend, and what documentation gaps must be addressed before any pressure is prescribed?
Read the worked answer
The report should explicitly state: diagnostic AHI 48/hr with central apnea index 21/hr and significant nocturnal hypoxemia (22% TST < 88%, nadir 79%); split-night titration limited by short diagnostic time; obstructive events controlled at 8 cmH2O; dose-dependent worsening of central apnea with higher pressure; patient intolerance at 10 cmH2O. Recommendation: do not issue a fixed CPAP pressure from this study; proceed to full-night titration and evaluate for ASV, with attention to the LVEF 30% contraindication considerations for ASV in HFrEF per current standards. Documentation gaps to close: total diagnostic sleep time and stage distribution, supine vs. non-supine event breakdown, whether the central events occurred in NREM vs. REM, and the timing/nature of the dyspnea complaint. These gaps matter because they determine whether the central pattern is positional, stage-dependent, or truly treatment-emergent - each leading to a different clinical path.