Documenting sleep assessment in a psychiatric visit: what to capture, MDM support, and coding implications
Sleep complaints are the most common symptom brought to a psychiatric visit that is not the presenting problem itself. In a recent large outpatient sample, 50 to 80 percent of patients with major depressive disorder report clinically significant insomnia, and the rates are comparable in generalized anxiety disorder, PTSD, and bipolar disorder. For many patients the insomnia is what finally brings them in, even when the underlying mood or anxiety condition has been present for months or years.
Despite this, sleep documentation in psychiatric notes is often thin: a single line in the MSE or a checkbox in a template. That is a missed opportunity. A thorough documented sleep assessment strengthens your medical decision making (MDM), provides objective data for severity arguments, and opens the door to either a comorbid sleep disorder diagnosis or a strong rationale for why the sleep complaint is a symptom of the primary psychiatric condition rather than a separate medical problem. Both routes carry coding implications worth understanding.
This guide covers what to capture, how to document it, and what the documentation means for MDM level and code selection.
What to capture: the structured sleep history
A sleep assessment in a psychiatric visit is not a sleep medicine consult. You are not expected to differentiate narcolepsy from idiopathic hypersomnia or to interpret a full polysomnogram. But you are expected to document enough detail to tell a reviewer whether the sleep complaint is a primary symptom, a side effect, an independent condition, or normal variation. Five domains cover almost every psychiatric presenting complaint:
1. Sleep onset
How long does it take the patient to fall asleep once they lie down? A subjective estimate (thirty minutes, two hours) is standard and sufficient. Capture whether the delay is caused by racing thoughts, rumination, physical restlessness, or no identifiable reason. This distinction often maps directly onto the underlying diagnosis: anxious rumination in GAD, pressured thought flow in mania, a quiet inability to quiet the mind in depression, or physical akathisia secondary to an antipsychotic.
2. Sleep maintenance
Does the patient wake during the night? How many times, for how long, and what happens during the waking? The two most common psychiatric maintenance patterns are mid-sleep awakening with racing thoughts (anxiety) and early morning awakening with inability to return to sleep (depression). Neither is pathognomonic, but the pattern should be documented because it shapes both the differential and the treatment target. A patient who wakes three times a night to use the bathroom and returns to sleep quickly has a different problem from one who wakes at 3 a.m. and lies awake for two hours.
3. Early morning awakening (EMA)
Wake time and the ability to fall back asleep. Early morning awakening is the classic melancholic feature in major depressive disorder, but it also appears in primary insomnia, substance withdrawal, and as a side effect of SSRIs and SNRIs. Documenting the wake time, the associated mood state on waking, and whether it represents a change from the patient's baseline strengthens the link between the sleep symptom and the primary diagnosis.
4. Perceived sleep quality and daytime consequences
The patient's subjective rating of sleep quality ("refreshed on waking?"), total sleep time, and the daytime burden: fatigue, sleepiness, cognitive fog, irritability, and functional impairment. Daytime consequences are where the sleep complaint becomes load-bearing for MDM, because they tie the sleep deficit to functional decline, which is part of severity and medical necessity arguments for every service level.
5. Sleep hygiene, substances, and behavioral factors
Caffeine, alcohol, nicotine, cannabis, screen use before bed, irregular bedtimes, shift work, and caregiving demands. Each of these is a modifiable factor, and documenting an attempt to address them before moving to pharmacotherapy is documentation best practice. A note that says "advised sleep hygiene" with no detail is a checkbox. A note that says "patient drinks 2 beers nightly as a sleep aid and uses a phone in bed for 90 minutes" shows clinical reasoning and a candidate intervention.
Documenting sleep on the mental status exam
The MSE captures what you observe, not what the patient reports. For sleep, that means the observable consequences: behavior and demeanor during the interview. Drowsiness, yawning, inattention, irritability, and slowed responses can all follow from sleep deprivation, and documenting them in the MSE provides an objective anchor for the patient's subjective complaint. A patient who reports severe insomnia but appears alert, euthymic, and engaged may still be truthful, but the discrepancy is worth noting because it helps distinguish acute from chronic states.
In the MSE, the relevant domain is behavior and appearance or general presentation, depending on your note structure. A typical documented entry:
Example MSE entry
Appearance: Casually dressed, well groomed, drowsy appearing with frequent yawning during the interview. Slouched posture, eyes heavy. Cooperative throughout.
Cognition: Attentive but required two repetitions of mildly complex instructions. Reported subjective concentration at "3 out of 10" today, attributing it to only three hours of sleep last night.
How sleep assessment supports the MDM level
Medical decision making under the 2021 and 2023 E/M revisions is built on three pillars: the number and complexity of problems addressed, the amount and complexity of data reviewed, and the level of risk from management options. Sleep findings can contribute to all three.
Problems addressed
Insomnia, hypersomnia, or a circadian rhythm disturbance documented as part of the presenting problem adds to the complexity of problems addressed. The key is whether you treat the sleep complaint as a separate problem or as a symptom of the primary condition.
- Symptom of the primary condition (e.g., "insomnia due to GAD"): the sleep complaint adds depth to the severity of the primary problem. A chronic, partially controlled psychiatric illness with a moderately severe sleep symptom that is not adequately responsive is closer to the "one or more chronic illnesses with exacerbation, progression, or side effects of treatment" descriptor for moderate MDM.
- Comorbid diagnosis (e.g., F51.01 Primary Insomnia alongside F41.1 GAD): two problems, each of which may independently meet the threshold for treatment. This is a stronger MDM argument because it separates the domains of clinical attention, but it places a higher documentation burden on the differential reasoning.
Data reviewed
When you order sleep studies (polysomnogram, actigraphy, home sleep apnea test) or review sleep logs, actigraphy reports, or prior sleep studies from the patient's PCP or a sleep specialist, those enter the data-pillar of MDM. Ordering a sleep study for a psychiatric patient with suspected sleep apnea or periodic limb movement disorder is a straightforward data contribution that can move MDM from straightforward to low or from low to moderate.
Document the decision to order or to defer. A note that says "discussed sleep study; patient prefers to trial sleep hygiene first" is a documented medical decision, and that counts under risk because you are selecting an initial management option with a defined endpoint.
Risk of management options
Sleep medications carry risk: sedation, falls, cognitive impairment, dependence, and drug interactions with psychiatric medications. Documenting that you considered those risks, discussed them with the patient, and selected a specific agent with a rationale strengthens the risk pillar. A typical moderate-risk scenario is prescribing trazodone or a low-dose sedating antidepressant to a patient already on another serotonergic agent; the serotonin syndrome risk, though low, warrants acknowledgment in the note.
The risk column is also where non-pharmacologic management earns credit. Referral for cognitive behavioral therapy for insomnia (CBT-I) is a management option with defined risks, and documenting the referral supports the MDM level, especially at lower thresholds where risk is the only pillar reaching moderate.
| MDM level | How sleep findings contribute | Example documentation trigger |
|---|---|---|
| Straightforward | Sleep complaint is mild, transient, or clearly situational. No separate intervention needed beyond primary treatment. | "Reports 1-2 nights of poor sleep before a stressful event. No daytime impairment." |
| Low | Sleep complaint is present but follows the expected pattern of the primary diagnosis. Sleep hygiene advice given, no medication started. One problem, mild severity. | "Insomnia consistent with depressive episode. Reviewed sleep hygiene. Patient will increase morning light exposure." |
| Moderate | Sleep complaint is moderately severe with daytime consequences. Medication started or adjusted. Two conditions managed (e.g., GAD plus primary insomnia). Sleep study ordered. | "Insomnia persists despite improvement in mood at week 6 of escitalopram. Starting trazodone 50 mg at bedtime after reviewing fall risk and interaction profile." |
| High | Sleep complaint presents with serious risk: sleep deprivation driving suicidal ideation, sleep-related hallucinations, uncontrolled sleep apnea complicating bipolar disorder, or sleep disturbance during pregnancy with teratogenicity concerns around options. | "Patient with bipolar I disorder in depressive episode reports 2 hours of sleep per night for 5 nights with racing thoughts. High risk of conversion to mania." |
Coding implications: when sleep becomes a separate diagnosis
The most consequential coding decision around sleep documentation is whether you report the sleep disturbance as a symptom of the primary psychiatric condition or as a separate diagnosis. The choice affects the claim, the payer response, and the long-term record structure.
Sleep as a symptom of the primary psychiatric condition
This is the default and the most common approach in outpatient psychiatry. The primary diagnosis (e.g., F33.1 Major Depressive Disorder, recurrent, moderate) is the only coded condition. The sleep complaint is documented in the HPI as a symptom of the depression. Advantages: no separate medical-necessity argument for the sleep diagnosis, no risk that a payer questions whether a sleep condition should be managed by the PCP, and no modifier requirement to distinguish overlapping services.
Sleep as a comorbid diagnosis
When the sleep complaint meets full diagnostic criteria for a sleep disorder (F51.01 Primary Insomnia, F51.11 Primary Hypersomnia, G47.33 Obstructive Sleep Apnea, F51.5 Nightmare Disorder), you have the option to code it separately. The advantages for MDM are substantial: you are now managing two conditions, which pushes the problem count higher and justifies more follow-up visits and a higher E/M level when the sleep complaint requires its own attention.
The documentation burden is also higher. To code a comorbid sleep disorder, the note must show that you ruled out other causes, that the sleep complaint exceeds what is expected from the primary psychiatric condition, and that the sleep problem requires independent management. A single sentence like "also has insomnia" does not support a separate diagnosis.
Coding sleep medication management
When you prescribe a sleep medication at a psychiatric visit, the code selection depends on whether the sleep problem is coded separately or is a symptom of the primary condition. If the sleep medication is part of the treatment plan for the primary psychiatric diagnosis (e.g., prescribing trazodone for insomnia due to MDD), it supports the E/M level of the same visit. If you are managing a separately coded sleep disorder and that is the sole or primary reason for the visit, the visit structure changes: the note needs a chief complaint, assessment, and plan specific to the sleep disorder, and the medical necessity argument shifts accordingly.
A complete sleep assessment note example
The patient below is fictional. The ALL-CAPS headings are the structure OneStep Scribe generates for psychiatric evaluations and progress notes. The example shows a follow-up visit where sleep assessment is a major element of the encounter.
Psychiatric follow-up note (99214 + 90833)
CHIEF COMPLAINT
"I still can't sleep. The medicine helps my mood during the day but my sleep is worse than before."
HISTORY OF PRESENT ILLNESS
Maya Chen, a 34-year-old woman with F32.1 Major Depressive Disorder, single episode, moderate, returns for week 6 follow-up after starting sertraline 50 mg daily. She reports clinically significant mood improvement: mood rated 4/10 (from 7/10 at intake), anhedonia improved, energy and appetite normalized. However, her sleep has worsened.
Onset insomnia is the primary complaint. She lies awake 90 to 120 minutes after bed, with rumination about work performance, a pattern she links directly to her supervisor's recent critical feedback. She previously required 20 to 30 minutes to fall asleep when depressed. Sleep maintenance is also affected: she wakes once or twice per night and takes 30 to 45 minutes to return to sleep. Total sleep time estimated at 4 to 4.5 hours nightly, down from 5.5 hours at baseline before sertraline. Wake time has shifted earlier by about one hour (now 5:00 a.m. vs. 6:00 a.m.), without difficulty returning to sleep on first awakening but worsening later in the night. Daytime consequences include fatigue rated 6/10, trouble concentrating in afternoon meetings, and a subjective "fog" that affects her work performance. She denied drowsy driving and microsleep episodes.
Relevant factors: caffeine intake increased from 2 cups to 3-4 cups daily to compensate for fatigue. She drinks one glass of wine before bed most nights, which she reports helps with onset but notes that she wakes more during the night. Bedtime is irregular (10:30 p.m. to midnight) due to variable work schedule. No recent travel, shift work, or caregiving changes. No prior sleep studies. No snoring or witnessed apneas reported by partner.
MEDICATION REVIEW
Sertraline 50 mg daily, started 6 weeks ago. Tolerated well except for worsened sleep and mild GI upset in week 1 (resolved). No other medications. NKDA.
MENTAL STATUS EXAMINATION
Alert, casually dressed, drowsy appearing with slowed blink rate and one yawn during interview. Mood "frustrated," affect somewhat irritable but full range and reactive. Speech normal rate and tone. Thought process linear, no psychosis. Insight and judgment intact. Denies SI, HI, and self-harm. Subjective attention "off today" but able to track interview without redirection.
RISK ASSESSMENT
Current ideation: Denies SI, HI, and self-harm. Risk considerations: chronic sleep deprivation of moderate severity, but no current ideation, no access to lethal means during waking hours, strong protective factors (supportive partner, treatment engagement). Overall risk level: Low.
ASSESSMENT
F32.1 Major Depressive Disorder, single episode, moderate. The depressive syndrome has responded partially to sertraline 50 mg daily at week 6: mood, anhedonia, and neurovegetative symptoms have improved, but sleep has worsened, consistent with a known activating side effect of sertraline in some patients. The sleep complaint is predominantly onset insomnia with a secondary maintenance component, driven by a combination of medication side effect, situational anxiety about work performance, and behavioral factors (evening alcohol use, irregular bedtime, increased caffeine). The differential for worsening sleep on an SSRI includes medication activation (the leading possibility here given the temporal relationship), unmasked anxiety, primary insomnia independent of the depression, and the natural course of untreated GAD (though the patient does not meet full GAD criteria).
The patient does not currently meet criteria for a separate primary insomnia disorder: the sleep symptom is temporally linked to a life stressor and a medication change, and it follows the pattern of the depressive episode rather than existing independently. I will give the sertraline several more weeks before treating the sleep complaint as a separate clinical problem. I reviewed the risks of adding a sedating agent at this point (drug interaction, daytime sedation, potential to mask a developing hypomanic switch) and the patient agrees to a behavioral-first approach with reassessment.
PLAN
1. Continue sertraline 50 mg daily. Shift dose to morning (currently taking at dinner) to reduce sleep interference.
2. Sleep hygiene interventions discussed in detail and agreed upon: eliminate evening alcohol as a sleep aid, set a consistent bedtime of 10:30 p.m., no screens in bed, and reduce caffeine to 2 cups before noon. Patient verbalized understanding.
3. CBT-I referral initiated. Patient given handouts; will provide list of local CBT-I providers by next visit. She will also use a simple sleep diary this week, to be reviewed at next visit.
4. Deferred sleep medication. Reviewed trazodone and low-dose mirtazapine as options if behavioral interventions fail. Patient prefers to try behavioral changes first.
5. Discussed activating SSRI side effect: patient now recognizes that her early sleeplessness may be medication-related and understands that shifting the dose to morning is the first step. She knows to call sooner if sleep worsens or if she has fewer than 3 hours of sleep for two consecutive nights.
6. Follow up in 4 weeks, or sooner if sleep worsens acutely. She verbalized understanding and agreement.
Five documentation pitfalls specific to sleep
- Sleep documented only as a checkbox or template line. "Sleep: fair" or "Insomnia: yes" adds no clinical weight. A reviewer sees a template entry. If the sleep complaint is part of the presenting problem, it needs HPI narrative.
- Sleep symptom in the MSE but absent from the HPI. Drowsiness observed during the interview is MSE material, but if that is the only mention of sleep in the note, there is no evidence you asked about sleep or took a history. The HPI must carry the patient's reported experience.
- Coding a comorbid sleep disorder without documenting the differential. Adding F51.01 alongside F32.1 without a differential paragraph invites a downcode. A separate diagnosis needs separate reasoning, temporal relationship, and treatment plan.
- Prescribing a sleep medication without documenting risk discussion. Starting zolpidem or trazodone without a word about fall risk, morning sedation, or dependence potential is a documentation gap that a plaintiff's expert will note.
- Ignoring sleep apnea risk in patients on sedating medications. A patient on clonazepam or quetiapine with undiagnosed sleep apnea and daytime somnolence represents a liability. A note documenting that you considered OSA (snoring, witnessed apneas, daytime sleepiness, elevated BMI) and either ruled it out clinically or deferred to a sleep study shows medical decision making; silence suggests you did not consider it.
The sleep assessment in intake versus follow-up visits
At intake, the sleep assessment is part of the HPI for the presenting problem. The structured history described above (onset, maintenance, EMA, quality, daytime consequences, behavioral factors) is the standard. Document total sleep time, pattern, duration of the complaint, and relation to the presenting symptoms.
At follow-up visits, sleep documentation should track the trend. Is sleep improving, stable, or worsening relative to the last visit? A trendline ("sleep improved from 4 hours to 6 hours on average since starting medication") is stronger than a static statement ("sleep is okay"). If you use a sleep diary or a sleep item on a symptom measure (the PHQ-9 sleep item, for example), reference the score and its trajectory. The mechanics of documenting serial scale scores are covered in the PHQ-9 and GAD-7 documentation guide.
Frequently asked questions
Do I need a separate sleep diagnosis to bill for sleep medication management?
No. Prescribing a sleep medication as part of the treatment plan for a primary psychiatric condition is standard practice and does not require a separate sleep disorder diagnosis. The medical necessity flows from the primary condition. If the sleep problem becomes the predominant focus of successive visits and the primary condition is stable, however, consider whether a separate sleep diagnosis better reflects the clinical picture and supports the service level.
Can I bill a separate E/M for sleep management on the same day as a psychiatric visit?
Generally not by the same clinician addressing the same patient on the same day. Medicare and most commercial payers bundle same-day E/M services by the same clinician under the highest level code. If a different clinician (e.g., a PCP) addresses the sleep problem on the same day, modifier rules apply. For psychiatric visits, the sleep assessment is part of the psychiatric E/M, not a separate service.
Does documenting a sleep study change the code?
Ordering or reviewing a sleep study adds to the data pillar of MDM but does not itself change the E/M level unless it contributes to a higher complexity threshold. The value is in the clinical reasoning it captures, not in the code. Document why you ordered it, what you expect to learn, and how the result will change management.
Should I code insomnia that is clearly caused by an SSRI?
Not as a separate diagnosis. The insomnia is an adverse effect of medication, not a primary or comorbid sleep disorder. Document it as a side effect of the medication and include it in the assessment as part of the treatment response narrative. If the side effect becomes the primary treatment target (e.g., you switch or adjust the SSRI on the basis of the insomnia), that supports the MDM level without a separate sleep code.
What about shift workers or patients with irregular sleep schedules?
Document the sleep schedule, the duration of the pattern, and whether the patient meets criteria for Circadian Rhythm Sleep-Wake Disorder (G47.2x). Shift work sleep problems are commonly underdiagnosed in psychiatric patients and can masquerade as or exacerbate mood and anxiety conditions. If you add a separate circadian disorder diagnosis, the same documentation standard applies: the disorder must be temporally and causally distinct from the primary psychiatric condition.
OneStep Scribe is an AI scribe for mental health: psychiatrists, PMHNPs, therapists, psychologists, and other mental health providers. It listens to the visit and drafts the evaluation or progress note in a structured format, including detailed sleep history, MSE, risk assessment, and a plan that ties the sleep findings to your clinical reasoning. Every account is NPI-verified.
Start a 14-day free trialThis article is educational and reflects one clinician's documentation approach. It is not legal, billing, or payer-specific advice, and it does not guarantee reimbursement. CPT is a registered trademark of the American Medical Association. The patient described is fictional. Always verify requirements with your payers and compliance advisors. Sleep medicine billing rules and medical necessity criteria vary significantly by payer; consult your coding team for payer-specific guidance.