In-Lab Sleep Study (Polysomnography)

What an in-lab sleep study is

An in-lab sleep study, also called attended polysomnography (PSG) or a Type I sleep study, is a comprehensive overnight diagnostic test conducted in a sleep laboratory under continuous observation by a registered sleep technologist. Unlike a home sleep apnea test — which records four to seven respiratory channels and can diagnose obstructive sleep apnea in selected adult patients — a PSG records a full montage of physiological signals that together capture not just breathing during sleep, but sleep itself.

The American Academy of Sleep Medicine (AASM) classifies sleep monitoring devices into four levels:

• Type I — full attended in-lab polysomnography. The diagnostic standard. Records at minimum: EEG (brain activity, multiple channels), EOG (eye movement), EMG (muscle tone, chin and limbs), ECG (heart rhythm), airflow (thermistor and nasal pressure), respiratory effort (chest and abdomen belts), oxygen saturation, and snore/body-position channels. Always staffed by a sleep technologist who can intervene if needed.
• Type II — full unattended portable PSG with the same channel set as Type I but used in the home without a technologist present. Used in some research and limited clinical settings; not common in routine US clinical practice.
• Type III — limited-channel portable monitors (the standard adult home sleep apnea test). Typically four to seven respiratory channels; no EEG, no EOG, no chin EMG.
• Type IV — single- or dual-channel devices, usually pulse-oximetry-only or airflow-only. Limited diagnostic utility; not generally recommended as a primary diagnostic tool.

This page covers Type I in-lab PSG specifically. The presence of a sleep technologist is what most clearly separates PSG from home testing: the technologist watches the live signals through the night, can replace a fallen-off lead before data is lost, can apply CPAP mid-study during a split-night protocol, and can document behaviors a passive recording can't catch. The signal set, the observer, and the controlled environment together define the gold-standard study against which all other sleep diagnostics are compared.

When in-lab PSG is the right test

The 2017 AASM clinical practice guideline for diagnostic testing in adult obstructive sleep apnea (Kapur et al.) recommends home sleep apnea testing as the first-line diagnostic approach for adults with a high pretest probability of moderate-to-severe OSA and no significant comorbidities. The same guideline is explicit that in-lab PSG is the right starting point — not HST — in several specific clinical situations:

• Significant cardiopulmonary disease. Heart failure, COPD, neuromuscular disease, or recent cardiovascular events make HST less reliable and a missed diagnosis riskier. The full PSG channel set, the technologist's real-time observation, and the controlled lab environment together provide diagnostic safety that home testing can't match.
• Suspected central sleep apnea. HST devices generally can't reliably distinguish central from obstructive events; the chest/abdomen belts can miss central effort patterns that the EEG-supplemented PSG captures clearly. Treatment-emergent central apnea (a complication of CPAP) similarly requires PSG to confirm.
• Suspected non-apnea sleep disorder. Parasomnias (sleepwalking, REM sleep behavior disorder, night terrors), narcolepsy, periodic limb movement disorder, and insomnia with paradoxical features all require EEG sleep staging and limb-EMG recording — channels HST simply doesn't have. A patient whose primary symptom is dream enactment, daytime sleepiness despite apparently adequate sleep, or restless-leg sensations needs a PSG, not an HST.
• Pediatric sleep evaluation. The AASM does not endorse HST for diagnosis of pediatric OSA in most clinical contexts. Children's sleep architecture, smaller airway anatomy, and behavioral cooperation challenges call for in-lab evaluation.
• Negative HST in a high-suspicion patient. Per the AASM CPG, a negative or technically inadequate HST in a patient with persistent clinical suspicion of OSA should be followed by an in-lab PSG. HST under-diagnoses OSA in roughly 10–20% of high-suspicion cases, often because the patient slept poorly during the home recording or the device missed events.
• CPAP titration. When the diagnosis is established and the next step is determining the optimal CPAP pressure (or testing alternative modes — BiPAP, ASV), a titration study is performed in-lab. The technologist adjusts pressure throughout the night based on observed events and arousal patterns. Auto-adjusting CPAP machines used at home accomplish much of this for routine cases, but complex cases (treatment-emergent central apnea, severe REM-related disease, comorbid hypoventilation) still benefit from in-lab titration.
• Multiple sleep latency test or maintenance of wakefulness test. MSLT (used to diagnose narcolepsy) and MWT (used to evaluate residual daytime sleepiness despite treatment) are daytime studies that follow an overnight in-lab PSG; the overnight PSG establishes that the patient slept adequately the prior night, which is required for the daytime study to be interpretable.

Many patients pass through HST first and only proceed to in-lab PSG if the home study is inconclusive. Others go directly to PSG when one of the conditions above is present from the start. The decision belongs to the ordering clinician, who weighs pretest probability of OSA against the likelihood of a non-apnea diagnosis or a complicating factor.

What in-lab PSG measures that HST cannot

The single most important capability difference is sleep staging. HST records breathing during what is presumed to be sleep, but it does not actually measure whether the patient is asleep at any given moment. PSG, with its EEG and EOG channels, classifies every 30-second epoch of the night into wake, REM sleep, or one of three non-REM stages. That distinction has a chain of clinical implications:

• Apnea-hypopnea index (AHI) calculation. AHI is events per hour of sleep. HST estimates the denominator from total recording time (how long the device was on), which is always longer than actual sleep time. PSG measures actual sleep time directly. The same patient can have an AHI of 12 by HST and 18 by PSG simply because PSG correctly excludes the wake periods — which moves the patient from "mild" to "moderate" OSA and changes treatment guidance.
• REM-related sleep apnea. Some patients have apnea events concentrated almost entirely in REM sleep. An HST that records a low overall AHI can mask a high REM AHI in a patient who happened to spend less time in REM during the home night. PSG captures REM percentage and stage-specific event rates, which informs both diagnosis and treatment intensity.
• Sleep architecture itself. Total sleep time, sleep efficiency (sleep time divided by time in bed), wake after sleep onset, REM latency, and the proportions of N1/N2/N3/REM are diagnostically useful in their own right. Reduced REM with depression and PTSD; suppressed N3 with chronic insomnia and certain medications; abnormal REM-onset patterns in narcolepsy. None of this is observable from a home recording.
• Parasomnia detection. Behaviors during sleep — sleepwalking, dream enactment in REM sleep behavior disorder, night terrors, sleep-related eating — are observed by the technologist, time-locked to the EEG record, and frequently video-recorded as part of the study. The combination of stage-tagged EEG, EMG, and video provides the evidence needed for diagnoses that HST simply cannot make.
• Limb movements during sleep. Periodic limb movements of sleep are scored from limb EMG channels not present on HST. PLMS is a diagnostic criterion for periodic limb movement disorder and a useful adjunct in evaluating restless legs syndrome severity.
• Cardiac arrhythmia capture. The full ECG channel on PSG can document sleep-related arrhythmias — particularly atrial fibrillation, which has a strong bidirectional relationship with sleep apnea. HST oximetry-derived heart rate is too coarse for this.

The hstVsPsgChannels diagram below summarizes the channel-level difference that drives all of the above:

For a patient whose only concern is "do I have moderate-to-severe obstructive sleep apnea, yes or no," HST is often sufficient and the additional PSG channels add little. For everyone else, the channels above are the difference between a diagnosis and a missed diagnosis.

What the night actually involves

Most patients arrive at the sleep lab between 7:30 and 9:00 PM, depending on the patient's typical bedtime. The lab sets the schedule based on your usual routine when possible — being asked to fall asleep at 8 PM if you normally sleep at midnight makes for a poor study, and good labs adjust accordingly.

After check-in, a sleep technologist walks you to a private bedroom — most modern accredited labs use single-occupancy rooms with a bed, private bathroom, and ambient-control lighting. You change into your own sleepwear (comfort matters; most labs encourage what you'd wear at home). The technologist then begins electrode placement, which takes 30 to 60 minutes. Lead set-up typically includes EEG (8–12 scalp electrodes attached with paste), EOG (around each eye for eye-movement detection), EMG (chin and each leg, the latter for limb-movement detection), ECG (chest), respiratory effort belts (chest and abdomen), a nasal cannula and small thermistor under the nose for airflow, a pulse oximeter on a finger, an optional snore microphone, and integrated position/movement sensors.

The full setup looks more dramatic than it feels. The wires are bundled together so you can move and turn over without snagging. Most patients describe being able to find a comfortable sleep position within five to ten minutes once the setup is complete.

The technologist runs a brief calibration — they'll ask you to look up, down, left, right, blink, hold your breath briefly, and move each leg. This confirms every channel is recording cleanly. Lights go out, and you sleep.

Through the night, the technologist watches your live signals from a control room. They are looking for events worth scoring, equipment problems worth fixing, and behaviors worth noting. If a lead falls off, they'll quietly enter the room to reattach it. If you need the bathroom, an intercom or call button summons them to disconnect you briefly. Modern labs use infrared video cameras so the tech can see you in the dark without disturbing you.

Most patients sleep less well in the lab than at home — the "first-night effect" is well-documented and accounted for in how studies are interpreted. Even so, four to five hours of recorded sleep is usually enough to make a diagnosis. If the study is a split-night protocol (diagnostic data in the first half, CPAP titration in the second half), the technologist will fit a mask once enough diagnostic data has been collected.

Wake time is typically 5:30 to 7:00 AM. The technologist returns, removes the leads (paste residue may need a shower to fully remove), and you complete a brief morning questionnaire. Most patients are out the door by 7:30 AM and can drive home — the study isn't sedating, and the night usually doesn't impair daytime function the next day.

Who staffs the lab, and who reads the study

An in-lab sleep study is touched by several distinct professional roles before, during, and after the night:

• The ordering clinician. Usually a primary care physician, sleep medicine specialist, pulmonologist, neurologist, ENT, or psychiatrist. They evaluate symptoms, decide that PSG (rather than HST) is appropriate, and write the order specifying the type of study (diagnostic, split-night, MSLT, titration).
• The sleep technologist. The night-of role. Most US labs employ technologists credentialed as Registered Polysomnographic Technologists (RPSGT) or Registered Sleep Technologists (RST). The technologist applies leads, monitors live signals, intervenes during titration, documents observed events, and writes a tech summary that travels with the recording to the interpreting physician.
• The interpreting sleep physician. A board-certified sleep medicine physician who reads the study, scores it (or reviews technologist scoring), and writes the formal diagnostic interpretation. AASM-accredited labs are required to have a sleep-medicine-trained physician as medical director.
• The respiratory therapist. Often involved in CPAP titration studies and patient education when treatment is initiated. Some sleep labs are integrated with a DME (durable medical equipment) supplier; others coordinate handoff to a separate DME company that provides the CPAP machine and supplies.
• The dental sleep specialist. Not on staff at most sleep labs but a separate referral when oral appliance therapy is being considered as an alternative or adjunct to CPAP.

The ordering clinician and the interpreting sleep physician may be the same person (in a sleep medicine practice) or different people (when the order comes from primary care or another specialty). Either way, the report makes its way back to the ordering clinician, who discusses results and decides next steps.

Lab accreditation is worth checking. The AASM accredits sleep facilities against published standards for staffing, equipment, scoring, and quality control. An accredited facility provides a meaningful floor that an unaccredited facility doesn't.

How the study is scored, and what the report says

After the night, the recording is scored — manually reviewed by the technologist and physician, with software assistance, against the AASM Manual for the Scoring of Sleep and Associated Events (most recent published version). Scoring resolves every 30-second epoch into a sleep stage, marks every respiratory event (apnea, hypopnea, RERA), counts arousals, identifies leg movements, and notes any cardiac arrhythmias or unusual behaviors.

The report you'll receive — typically routed to your ordering clinician within five to ten business days — contains the following information at minimum:

• Sleep architecture summary. Total sleep time. Sleep efficiency (percent of time in bed actually asleep). Sleep latency (time to fall asleep). REM latency. Percentage of time in each sleep stage (N1, N2, N3, REM). Number of awakenings.
• Respiratory events. Apnea-hypopnea index (AHI) overall and by sleep stage and body position. Respiratory disturbance index (RDI). Lowest oxygen saturation (nadir SpO2). Time spent below 88% saturation. Oxygen desaturation index (ODI).
• Cardiac findings. Average and range of heart rate. Any arrhythmias observed.
• Limb movements. Periodic limb movement index (PLMI) and arousals from PLM events.
• Notable behaviors. Any parasomnia events observed by the technologist, time-stamped to the recording.
• Interpretation and impression. A narrative paragraph from the sleep physician summarizing findings and the clinical impression, often with a recommended next step (treatment with CPAP, referral for oral appliance, additional testing, no diagnosis identified).

The AASM publishes severity thresholds for the most common findings. AHI under 5 is normal; 5–14 is mild OSA; 15–29 is moderate; 30 or above is severe. AASM scoring rules also define what counts as a hypopnea (a partial reduction in airflow) — there are two definitions in current use, and the choice between them can shift a borderline patient's AHI by several points. Most US labs use the recommended AASM rule (≥30% airflow reduction with ≥3% desaturation OR an arousal); some labs and Medicare-historical contexts use the older ≥4% desaturation rule. If you're tracking your own AHI across studies over time, ask which scoring rule was used.

Common variations on the basic study

Not all in-lab sleep studies are the same. Several common variations adjust the protocol for specific clinical questions:

• Diagnostic study. The default. One full night of recording without treatment intervention, used to establish a diagnosis.
• Titration study. CPAP (or BiPAP, ASV, or other PAP) is fitted at the start of the night and the technologist adjusts pressure throughout the night to find the setting that resolves events. Used after a separate diagnostic study has confirmed the disorder.
• Split-night study. The first half of the night runs as a diagnostic study; if the patient meets criteria for OSA early enough, the technologist transitions to CPAP titration in the second half. Saves a study but requires both halves to capture enough sleep to be interpretable.
• MSLT (multiple sleep latency test). A daytime study following an overnight PSG. The patient lies down for five 20-minute nap opportunities at two-hour intervals. Sleep latency and the presence of sleep-onset REM episodes are scored. Used primarily to evaluate suspected narcolepsy.
• MWT (maintenance of wakefulness test). The opposite of MSLT — the patient is asked to stay awake in a quiet, dimly lit room for four 40-minute trials. Used to evaluate residual sleepiness despite treatment, often for occupational or safety determinations.
• Pediatric PSG. Same instrumentation, different scoring rules and stage-percentage norms. Typically performed in dedicated pediatric sleep labs or in adult labs with pediatric-specific protocols. A parent usually stays in the room overnight.
• Type II home PSG. Full PSG montage but unattended in the home. Increasingly available with newer equipment; appropriate in selected adult patients without significant comorbidity.

Your ordering clinician will specify the protocol on the order. If you arrive at the lab uncertain what kind of study is planned, the technologist can confirm and explain.

What happens after the study

The recording is scored and interpreted in the days following the study. Your ordering clinician usually receives the report within five to ten business days, sometimes longer. They'll then schedule a follow-up to review results — most patients see results within two to four weeks of the study itself, although urgent findings (severe OSA with significant desaturation, complex arrhythmias) may move faster.

The follow-up conversation depends on what the study showed:

• OSA diagnosed, no titration done. Treatment options are discussed: CPAP (the first-line treatment for moderate-severe OSA), oral appliance therapy (often appropriate for mild-to-moderate OSA or as an alternative when CPAP isn't tolerated), positional therapy, or — in selected severe cases — surgical evaluation including upper airway stimulation. See CPAP therapy.
• OSA diagnosed, titration done. The optimal pressure has been identified during the study. The next step is acquiring the CPAP machine and supplies through a DME provider, with follow-up to assess adherence and effectiveness — usually a 30-day and 90-day check-in.
• Non-apnea diagnosis. Periodic limb movement disorder, REM sleep behavior disorder, narcolepsy, parasomnia — each has its own treatment pathway. Some involve medication; some involve safety counseling; some involve daytime testing (MSLT for narcolepsy) before treatment is initiated.
• No specific diagnosis identified. Sometimes the study is normal or shows only borderline findings. The clinician may recommend behavioral interventions (sleep hygiene, CBT-I), repeat testing in different conditions, or evaluation for non-sleep-disorder causes of the symptoms (depression, thyroid, medication effects).

If the study reveals OSA but you weren't aware of having any of its risk factors, the implications extend beyond sleep itself — untreated OSA is associated with hypertension, atrial fibrillation, stroke, type 2 diabetes, and accelerated cognitive decline. Treatment isn't only about feeling rested; it's about cardiovascular and metabolic risk over decades. See sleep and heart disease for the cardiovascular evidence and why sleep matters for the broader context.

If you're reading this page before a scheduled study, the most useful preparation is to keep your normal pre-bed routine the day of the study, avoid caffeine after noon, avoid alcohol entirely that day, bring your own pillow if you have a strong preference, and bring something to read or a phone for the half-hour or so between arrival and lights-out. The study itself is straightforward; the rest is a clinician's interpretation, and that's outside your control.