MSLT Scheduled Five Days After Stopping Venlafaxine in a Shift Worker
A 27-year-old woman is referred for a multiple sleep latency test because of 2 years of daily sleepiness (Epworth 18) and two episodes of knee buckling while laughing. She works rotating 12-hour shifts and reports 5–6 hours in bed on work nights. Night 1 in-laboratory polysomnography shows total sleep time 6 hours 5 minutes, sleep latency 3 minutes, AHI 3/hr, periodic limb movement index 2, one REM period with a REM latency of 62 minutes, and SpO2 nadir 92%. She takes venlafaxine 150 mg daily, which she stopped on her own 5 days ago, and drinks four cups of coffee daily, the last at 16:00. Five naps are scheduled starting at 07:30 after a 06:00 final awakening. The ordering provider documents "rule out narcolepsy" and asks the technologist to proceed so the patient does not lose another day of work.
1. Should the MSLT proceed as scheduled? Evaluate each confounding factor.
Read the worked answer
No. Proceeding would produce an uninterpretable study and a real risk of a wrong diagnosis in either direction.
Insufficient and irregular sleep. Five to six hours in bed on work nights, confirmed by a polysomnographic total sleep time of only 6 hours 5 minutes, means some of her sleepiness is likely chronic sleep restriction. A short mean sleep latency under those conditions is not diagnostic of a central hypersomnia. The test requires documented sufficient sleep for roughly 2 weeks beforehand, typically 7 or more hours nightly for adults, verified by diary and actigraphy.
Antidepressant withdrawal. Venlafaxine has a short half-life; stopping 5 days earlier produces REM rebound and autonomic withdrawal effects that can shorten REM latency, distort nap architecture, and confound the SOREMP count. Standard preparation requires about 2 weeks off most REM-suppressing antidepressants, and 5–6 weeks for fluoxetine. Five days is inadequate.
Circadian timing. A rotating 12-hour shift schedule means 07:30 nap opportunities may fall at an adverse circadian phase, systematically shortening sleep latency regardless of pathology.
Stimulants. Four cups of coffee daily, last at 16:00, can affect sleep onset and nap latency; caffeine should be stabilized and avoided on test day.
Borderline PSG total sleep time. Six hours is at the lower edge of the usual minimum; many protocols prefer 7 hours to protect against a false-positive short latency.
Conclusion. Postpone, document the specific reasons, and speak with the ordering provider the same day rather than simply refusing the study.
2. Describe the preparation that must occur before rescheduling, and what to do if venlafaxine cannot be safely discontinued.
Read the worked answer
Sleep stabilization. Two weeks of a written sleep diary plus actigraphy, with a target of at least 7 hours in bed nightly and a consistent sleep–wake schedule. If she is on rotating shifts, the study should be scheduled after a stable stretch of day or evening sleep rather than at the end of a night rotation, and the schedule should be documented.
Medication. Coordinate with the prescribing clinician, not the patient alone, to taper and stop venlafaxine about 2 weeks before the test if clinically appropriate. Confirm no other REM-suppressing agents (SSRIs, SNRIs, tricyclics, trazodone, some anticonvulsants, stimulants, modafinil, amphetamines, hypnotics) and document the discussion. A urine drug screen on the morning of the MSLT is standard practice to verify compliance.
Substances. Stabilize caffeine, avoid it after about noon, and none on the test day. No alcohol on the day before or the day of testing. Nicotine should be stabilized, not abruptly withdrawn.
Night-before requirements. A polysomnogram documenting adequate total sleep time, generally at least 6 hours and preferably 7, before naps begin. Naps should be offered at 2-hour intervals beginning 1.5–3 hours after the final morning awakening, with a fixed protocol for sleep-onset REM determination.
If venlafaxine cannot be stopped. Proceed only with explicit acknowledgment in the report that REM-suppressing medication was present and that SOREMP results are unreliable; consider alternative approaches such as CSF hypocretin-1 measurement if cataplexy is convincingly present, or clinical diagnosis and treatment with later confirmatory testing. Do not silently run the test and report a normal or abnormal result as if unconfounded.
3. The test is later repeated properly: mean sleep latency 6.5 minutes with one SOREMP. The preceding PSG REM latency was 62 minutes. Interpret the result against diagnostic criteria and outline next steps.
Read the worked answer
Apply the criteria directly. Narcolepsy requires a mean sleep latency of 8 minutes or less plus at least two SOREMPs, one of which may be supplied by a REM latency of 15 minutes or less on the preceding polysomnogram. Here the mean sleep latency of 6.5 minutes is short enough, but there is only one SOREMP; the PSG REM latency of 62 minutes does not count as a second. The narcolepsy criteria are therefore not met on these data.
What the result does support. A mean sleep latency of 8 minutes or less with fewer than two SOREMPs is the pattern used for idiopathic hypersomnia, but that is a diagnosis of exclusion after insufficient sleep, circadian rhythm disorder, obstructive sleep apnea, depression, medication effects, and chronic medical illness have been addressed. Her shift work and prior sleep restriction must be definitively excluded before that label is applied.
Weigh the clinical picture. Knee buckling with laughter is a red flag for cataplexy. Cataplexy with a short mean sleep latency but fewer than two SOREMPs is not automatically narcolepsy, but it justifies escalation rather than reassurance.
Next steps. Review adherence to sleep extension, actigraphy, and medication washout for this specific run. If cataplexy remains convincing, discuss CSF hypocretin-1 testing with the sleep physician. Consider repeating the MSLT if preparation was imperfect, and treat symptomatically while the diagnosis is clarified. Document the discordance between symptoms and test results clearly so it is not lost in follow-up.