Split-Night Decision at 01:10 With Incomplete REM Sampling
A 46-year-old man (BMI 38, neck circumference 18.5 in) is being studied for loud snoring, witnessed apneas, and dozing at traffic lights; Epworth Sleepiness Scale score is 16. Laboratory policy permits conversion to a split night when accepted criteria are met and at least 3 hours remain for titration. At 01:10, after 2 hours 10 minutes of recording, the technologist scores 92 minutes of sleep: AHI 27/hr, all obstructive; supine AHI 61/hr; REM sleep only 6 minutes, all supine coincidentally; SpO2 nadir 79%; 24% of total sleep time below 88%; no arrhythmia. The patient must leave at 05:30 for a scheduled work shift and cannot return for 6 weeks. He has never used CPAP, wears full dentures, and reports chronic nasal congestion. The attending physician is not on site and is reachable only by phone.
1. Should the technologist convert to CPAP titration now or continue the diagnostic portion? Justify the decision using explicit criteria and identify the interpretive limitation of the data collected so far.
Read the worked answer
Convert, and document the rationale.
Criteria applied. Split-night conversion is appropriate when the diagnostic portion captures at least 2 hours of recording with an AHI of 40/hr or greater, or an AHI of 20–40/hr accompanied by a clinical modifier such as severe oxyhemoglobin desaturation, arrhythmia during sleep, or marked daytime sleepiness. This patient meets the second pathway on three counts: AHI 27/hr, nadir 79%, and 24% of sleep below 88% with reported drowsy driving.
Interpretive limitation. Only 6 minutes of REM were captured, so REM-related severity is almost certainly underestimated; the true whole-night AHI may be higher, and the REM supine requirement for a definitive titration has not yet been sampled. That limitation argues for completing the titration in supine REM rather than for continuing a purely diagnostic night, because supine AHI of 61/hr plus desaturation already establishes severity.
Practical constraints. Roughly 4 hours 20 minutes remain, which is enough for mask fitting and at least 3 hours of titration if conversion happens promptly. A 6-week delay to repeat testing is a real harm given the driving risk.
Documentation. Note the AHI, the desaturation burden, the abbreviated REM sampling, the reasoning for conversion, and the phone contact with the attending. Also plan how dentures will be handled for mask fitting and how nasal congestion will be addressed, since both affect titration quality.
2. If you convert, outline the titration plan: starting pressure, adjustment rules, endpoints, and a contingency if pressure climbs above 15 cm H2O with events persisting.
Read the worked answer
Start and interface. Begin at 4–5 cm H2O with a full-face or nasal mask chosen after inspecting the nares for congestion; if a nasal interface is used, plan for a chin strap or a switch to oronasal if mouth leak appears. Decide and document whether dentures stay in, because edentulous anatomy changes mask seal and pharyngeal geometry.
Adjustment rules. Increase pressure by 1 cm H2O at intervals of at least 5 minutes when obstructive events, flow limitation, or arousals persist. Larger steps (up to 3 cm H2O) are acceptable when events are frequent and desaturating, but avoid stacking pressure during wake or during a leak-ridden epoch.
Endpoints. Elimination of obstructive apneas, hypopneas, and flow limitation in supine REM; SpO2 maintained at 90% or above; no respiratory-related arousals; acceptable leak. A single pressure that holds in supine REM is the target because it will cover non-REM and lateral sleep.
Contingency above 15 cm H2O. Stop escalating and interrogate the data rather than the number: verify leak at mask and mouth, confirm the patient is actually asleep and supine, check for positional worsening, and confirm the events are obstructive rather than central. If obstructive events persist at 15 cm H2O with a clean leak and confirmed supine REM, this suggests a high pressure requirement, and bilevel PAP should be discussed with the attending rather than pushing CPAP higher in an unmonitored setting. Record the highest pressure tolerated, the residual event index at each level, and the reason titration stopped.
3. The night ends at 13 cm H2O: supine REM AHI 2/hr, nadir SpO2 91%, but there are 6 minutes at 86–88%, and the patient accumulated only 3 hours 10 minutes of titration sleep. What do you write in the report and what follow-up do you recommend?
Read the worked answer
Report content. State the prescribed pressure as 13 cm H2O with the interface and humidification used, and explicitly note that the endpoint was reached in supine REM with a residual AHI of 2/hr and a nadir of 91%. Report the 6 minutes at 86–88% as a residual desaturation that is brief and not meeting a sustained hypoxemia threshold, and identify whether those minutes occurred before pressure optimization, during an arousal, or in a leak epoch.
Validity statement. Note that titration sleep time was 3 hours 10 minutes, which is adequate but not generous, and that the pressure was confirmed in supine REM, which is the strongest single validation available. This makes the prescription defensible despite the shorter night.
Follow-up recommendations. Prescribe fixed CPAP at 13 cm H2O with heated humidification and heated tubing, arrange a data download at 2–4 weeks to check residual AHI, 95th-percentile leak, and adherence, and schedule clinical follow-up at 4–6 weeks for symptom reassessment. If adherence is adequate and residual AHI is above 5/hr or the desaturations recur on download, re-titrate rather than empirically adding oxygen. Do not order nocturnal oxygen now; the brief desaturation does not meet sustained criteria and should first be shown to persist on effective PAP.