Skip to main content
v2026.11,772 entries · CC-BY 4.0

E/M With a Psychotherapy Add-On: Why Time Cannot Select the Level

When a psychotherapy add-on code (90833, 90836, 90838) is reported with an office visit, the E/M level must be selected on medical decision making alone. Time is barred, prolonged services are barred, and the note must state minutes for each service separately.

Written and maintained by CASRAI Editorial Board

Last updated

Outpatient psychiatry is the one setting where the ordinary rules for selecting an office visit level stop applying. When a psychotherapy add-on code is reported alongside an evaluation and management (E/M) service, the usual choice between medical decision making and total time collapses: time is off the table entirely, and the E/M level must be selected on medical decision making alone. Practices that carry an institutional habit of coding visits by the clock — a habit the 2021 office/outpatient revisions actively encouraged everywhere else in medicine — reproduce that habit in psychiatry and generate a claim the record cannot support.

This page is about that boundary: which add-on codes exist, what the E/M must be built from when one is used, what has to appear in the note, and where the documentation actually fails on review.

Verified against Noridian Healthcare Solutions, Jurisdiction E Part B, “Billing Tips for Psychotherapy with Evaluation & Management (E/M),” last updated 27 November 2024; the CMS Medicare Coverage Database Billing and Coding Article A57480 (Psychiatry and Psychology Services); and the American Psychiatric Association’s coding and reimbursement guidance. CPT codes and descriptions are copyright the American Medical Association. Contractor-specific behaviour is flagged where it applies.

The Three Add-On Codes

Psychotherapy provided on the same day as an E/M service is not reported with a standalone psychotherapy code. It is reported with an add-on code that exists specifically for the with-E/M situation. Per the Noridian billing tips article, the three are:

  • 90833 — Psychotherapy, 30 minutes with patient when performed with E/M
  • 90836 — Psychotherapy, 45 minutes with patient when performed with E/M
  • 90838 — Psychotherapy, 60 minutes with patient when performed with E/M

These are add-on codes in the CPT sense: they cannot stand alone, and they attach to an E/M code rather than to a standalone psychotherapy code. The standalone psychotherapy codes and the with-E/M add-on codes describe the same clinical activity but live in different reporting worlds. Choosing the wrong family is one of the two most common errors in this space; the other is the time problem described below.

Noridian states the condition plainly: “Psychotherapy with an eligible E/M service can be reported using an E/M code plus a psychotherapy add-on code when performed on the same day.”

The Rule That Catches People: Time Cannot Select the E/M Level

This is the core of the page, and it is worth quoting the contractor language directly. Noridian, summarising the CPT codebook instructions for these codes, lists:

  • “Type and level of E/M service selected based on medical decision-making”
  • “Time spent on E/M activities not used to report time for psychotherapy”
  • Time may not be used as basis of E/M code selection
  • “Prolonged services may not be reported when psychotherapy with E/M services are reported”

Four consequences follow, and each one is a live audit exposure.

First, the E/M level is an MDM determination and nothing else. A psychiatrist who spends 55 minutes with a patient, 38 of them in psychotherapy, cannot reach a higher office visit level by pointing at the total. The E/M level is built from the number and complexity of problems addressed, the amount and complexity of data reviewed, and the risk of complications from management — the same three MDM elements that apply elsewhere, assessed on the medical portion of the encounter only.

Second, the two time pools are separate and non-overlapping. Minutes spent on the medical evaluation and management work do not count toward the psychotherapy time that determines whether 90833, 90836 or 90838 is correct. A 45-minute encounter that includes 20 minutes of medication management and 25 minutes of psychotherapy supports 90833, not 90836 — the psychotherapy clock reads 25 minutes, not 45.

Third, prolonged services are unavailable. When psychotherapy with E/M is reported, prolonged service codes are simply not reportable. This is a categorical bar, not a documentation threshold to be met.

Fourth, the two services must each be significant and separately identifiable. Noridian’s requirements for same-day billing state that “Two services must be significant and separately identifiable” and that the “E/M key components — history, examination, and medical decision-making — must be separately identifiable from psychotherapy service.” A medication check folded into a therapy hour is one service, not two.

What the Note Has to Contain

Noridian sets out an explicit documentation requirement that is stricter than what most E/M encounters demand:

“Time spent for each service must be clearly documented in the medical record by number of minutes, or start and stop times, for each service.”

Read that carefully. It is per service, not for the encounter. A note that says “50-minute session” satisfies nothing. A note that says “psychotherapy 30 minutes” and leaves the E/M time unstated is incomplete on its face — even though E/M time cannot select the level, the contractor still wants the separation demonstrated, because the separation is what proves the psychotherapy minutes were not inflated by medical work.

Noridian also flags an additional requirement: “Focus is on patient with their presence required for all or majority of time.” Time spent with family or collateral contacts, when the patient is not present, does not straightforwardly count toward the psychotherapy add-on.

The article notes that the Comprehensive Error Rate Testing (CERT) contractor reviews medical records on samples of processed claims, and that review findings “may include elements necessary to improve documentation and ensure the service for psychotherapy and E/M was supported.” In other words: this code pair is actively sampled, and the sampling looks specifically at whether both halves were supported.

Worked Example One: The Time Trap

An established patient with recurrent major depressive disorder and generalised anxiety presents for a scheduled 50-minute appointment. The psychiatrist spends roughly the first 15 minutes on medication management — reviewing response to a recent SSRI dose increase, screening for activating side effects, reviewing a metabolic panel drawn the previous week, and deciding to hold the current dose. The remaining 35 minutes are spent in structured cognitive behavioural work on avoidance behaviours.

The wrong instinct: total encounter time is 50 minutes, which in a general outpatient setting would support a higher office visit level on time alone. Code the E/M by the clock and add a psychotherapy code.

The correct analysis:

  • The psychotherapy clock reads 35 minutes. That supports 90836 (45 minutes) only if 35 minutes falls within that code’s reporting range; if it does not, 90833 (30 minutes) is the correct add-on. The medication management minutes are excluded from this count entirely.
  • The E/M level is determined by MDM on the medical portion: one chronic illness with exacerbation or side effects under treatment, review of an external laboratory result, and prescription drug management. Those three feed the standard MDM grid. The 50-minute total is irrelevant to the E/M selection.
  • Prolonged services are not reportable regardless of how long the encounter ran.
  • The note must state the minutes for each service separately.

Worked Example Two: When It Is Only One Service

A patient on a stable regimen attends a 45-minute psychotherapy session. Near the end, the patient mentions mild dry mouth. The psychiatrist acknowledges it, confirms the patient is tolerating the medication, and makes no change.

There is no separately identifiable E/M service here. A brief, unstructured acknowledgement of a tolerated side effect with no assessment and no management decision does not constitute the history, examination and medical decision making of an E/M service. The correct report is a standalone psychotherapy code for the session — not an E/M plus an add-on.

This is the mirror image of the modifier 25 problem in procedural specialties, and the reasoning is identical. Our guide to modifier 25 and the significant, separately identifiable test works through what “significant and separately identifiable” actually requires and what fails it; the analytical structure transfers directly, even though modifier 25 itself is not the mechanism used to link an E/M to a psychotherapy add-on.

Worked Example Three: The Initial Evaluation

A new patient is seen for a diagnostic evaluation. The psychiatrist takes a full psychiatric history, performs a mental status examination, reviews records from a previous treating clinician, establishes a working diagnosis, and initiates a medication. No psychotherapy is delivered.

This is an E/M service alone, or a psychiatric diagnostic evaluation code, depending on the practice’s coding convention and payer instruction. No add-on applies, because no psychotherapy occurred. Reporting 90833 because the visit “felt therapeutic” is exactly the misapplication the CERT samples are designed to catch — the add-on codes describe psychotherapy as a distinct service with its own time, not the supportive quality of a good clinical conversation.

Telehealth

Psychiatry delivers a large share of its outpatient volume by telehealth, and the modifier and place-of-service decisions ride on top of everything above rather than replacing any of it. The E/M-level-by-MDM rule and the separate-time-documentation rule do not relax because the encounter was conducted by video. For the modifier side of that decision, see our guides to modifier 95 versus modifier 93 and the place-of-service choice and to what remains of the older GT and G0 telehealth modifiers.

Who May Bill What

The add-on codes attach to an E/M service, which means the billing practitioner must be someone who can report an E/M service. That excludes practitioners whose scope does not include E/M — a psychologist providing psychotherapy reports the standalone psychotherapy codes, not an E/M plus an add-on. Where a non-physician practitioner participates in the encounter alongside a physician, the office-setting question is governed by incident-to rules rather than by anything in the psychotherapy code family; see our guide to incident-to billing and its conditional test, setting by setting.

Frequently Asked Questions

Can I select the E/M level by total time if the psychotherapy was brief?

No. The bar is categorical whenever a psychotherapy add-on is reported alongside the E/M. Noridian states it without qualification: “Time may not be used as basis of E/M code selection.” If the psychotherapy is too brief to support any add-on and you report no add-on, then you are reporting an ordinary office visit and the normal time-or-MDM choice returns.

Does the psychotherapy time include time spent documenting the note?

The add-on descriptors specify time “with patient,” and Noridian adds that the “focus is on patient with their presence required for all or majority of time.” Documentation performed after the patient has left does not fit that description.

Can I report prolonged services if the encounter genuinely ran long?

No. “Prolonged services may not be reported when psychotherapy with E/M services are reported.” This is one of the clearest bright lines in the whole area.

Do the two services need different diagnoses?

Nothing in the contractor guidance makes a second diagnosis a qualifying criterion. The test is whether two distinct services were performed and documented, not whether two diagnosis codes were attached. The same principle governs the analogous modifier 25 decision, where the NCCI Policy Manual is explicit that different diagnoses are neither required nor sufficient.

What is the single most common documentation failure?

A single undifferentiated time statement for the whole encounter. The requirement is minutes, or start and stop times, for each service. Without that split, a reviewer cannot confirm that the psychotherapy add-on was supported by psychotherapy minutes rather than by medical management minutes, and the add-on line is the one that falls.

Does a Medicare Advantage or commercial payer follow the same rules?

The CPT instructions on E/M level selection and time separation come from the codebook and apply generally. Coverage, documentation sampling and any additional edits are payer-specific. Where a local coverage article or payer policy exists — Noridian’s Jurisdiction E guidance is one example, and other MACs publish their own — that document governs the contractor-dependent details.

Related Reading

Follow CASRAI

Research-administration guidance, standards updates and independent tool reviews.

Ask CASRAI · included with Regulatory Radar

Ask about E/M With a Psychotherapy Add-On: Why Time Cannot Select the Level

Ask CASRAI answers research-administration questions and cites the passages behind every claim — and says so when the corpus does not cover something, instead of guessing. It comes with a Regulatory Radar subscription at $29 a month, alongside the daily digest of regulatory changes and the dashboard of what changed.

150 questions a day, on this site, over the API, or inside your own tools through the CASRAI MCP server.

Everything CASRAI publishes — this page, the dictionary, the guides and the news — stays free to read, with no account and no card.

Referenced across the research world

University of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logoUniversity of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logo
  • University of Cambridge logo
  • Columbia University logo
  • Crossref logo
  • University of Edinburgh logo
  • Harvard University logo
  • University of Oxford logo
  • Princeton University logo
  • Stanford School of Medicine logo
  • University College London logo
  • ORCID logo

View CASRAI adoption →

Regulatory Radar

Stop finding out after the fact

$29/month, cancel anytime. Daily digest updates from our analysis, a dashboard holding the same items, and a cited assistant for everything they raise.

  • Federal Register, Federal Register+, Grants.gov, Regulations.gov, NSF News, UKRI, plus CASRAI’s own published content.
  • 72,264 indexed passages, and every answer cites the ones it drew on.