Maria writes SOAP because her graduate program taught SOAP, and she has never once asked whether SOAP is the right container for a psychodynamic session, a Medicaid community-mental-health case, or a Thursday-night substance-use group billed under 90853. Most clinicians inherit a note format the way they inherit a filing cabinet: it was there when they arrived. But format is not cosmetic. DAP, BIRP, and GIRP each foreground a different thing (data, behavior, goals), each maps onto a different payer expectation, and each defends a different kind of claim. This lesson teaches you to pick the format deliberately, run AI conversion across all three from the same session material, and defend your choice out loud to a clinical supervisor. By the end you will hold a format decision tree you can apply to every client on your caseload, and the conviction to stop writing every note in the same shape regardless of what the chart needs, which is the difference between a BIRP vs DAP note chosen on purpose and one chosen by habit.

Three Formats, One Question Each

Strip away the acronyms and each format is a different question put to the session. DAP (Data, Assessment, Plan) asks: what happened, what does it mean, what next? It compresses SOAP's Subjective and Objective into a single Data section, which makes it the natural home for sessions where the line between what the client reported and what you observed is less load-bearing than the clinical meaning you draw from the whole. BIRP (Behavior, Intervention, Response, Plan) asks: what did the client do, what did you do, what happened when you did it, what next? Its center of gravity is the intervention-response pair, which is exactly what a Medicaid reviewer wants to see: a skilled service delivered, and a measurable client response to that service, every single session. GIRP (Goal, Intervention, Response, Plan) asks the same chain but anchors it to a treatment plan goal: which goal did this session serve, what intervention advanced it, how did the client respond, what next? GIRP is the format that makes goal linkage impossible to omit, which is why it shines wherever the payer or the program demands that every session visibly serve the plan.

Think of formats as camera angles on the same session. The session is the session; the format decides what is in the foreground of the frame. DAP foregrounds clinical reasoning, BIRP foregrounds the service transaction, GIRP foregrounds the treatment plan. A camera angle is not a lie and not a truth; it is an editorial choice about what the future reader, an auditor, a reviewer, a successor clinician, a board investigator, needs to see first. Choosing the angle is clinical judgment. The AI can reframe the footage into any angle you pick; it cannot pick the angle, because the angle depends on who will read the note and why, and the model does not know your payer mix, your program's contract, or your supervisor's audit history.

One more orientation point before the worked examples: format choice is constrained from three directions. The payer (some state Medicaid manuals prefer BIRP for outpatient behavioral health, and a community mental health center's contract may require it), the setting (group programs and SUD treatment lean toward goal- and behavior-anchored formats), and the modality (insight-oriented work fits awkwardly into a behavior-transaction frame). Where the payer speaks, the payer wins; where the payer is silent, modality and setting decide.

DAP for the Psychodynamic Session

Take a psychodynamic session: a 90834 with a client working on a repeating relational pattern, where the session's clinical substance was the client's transference reaction to the therapist's vacation announcement and the exploration that followed. Try to force this into BIRP and you feel the misfit immediately. What was the "Behavior"? Sitting and talking. What was the discrete "Intervention"? Interpretation, clarification, holding the frame, none of which decompose into the technique-response transactions BIRP wants. The note comes out either falsified (inventing a tidy intervention-response pair) or vacant ("therapist provided supportive listening; client responded well"), and vacant is what payer reviewers call insufficient.

DAP lets the session be what it was. Data: "Client arrived five minutes late, which she connected, unprompted, to feelings about the upcoming treatment break. She reported a recurrence of the abandonment-themed dream and stated, 'You'll forget about me by Tuesday.' Affect was irritable early in session, softening to sadness by the midpoint." Assessment: "Client's reaction to the announced break is consistent with the core relational theme identified in the treatment plan (anticipatory abandonment leading to preemptive withdrawal). Her ability to name the connection in session, rather than act on it by canceling, represents measurable progress in affect tolerance and reflective functioning. Symptoms of low mood persist and continue to strain her marriage; continued weekly psychodynamic psychotherapy remains medically necessary to consolidate this emerging capacity." Plan: "Continue weekly 45-minute sessions; therapist will return to the break material next session; client will log withdrawal urges between sessions." Notice the Assessment still carries the medical-necessity chain (symptom, functional impairment, progress, ongoing need); DAP does not exempt you from necessity, it just gives clinical reasoning more room than a transaction log would.

The AI conversion prompt for DAP mirrors the SOAP prompt from the previous lesson with one changed clause: "Format: three labeled sections, Data, Assessment, Plan. In Data, integrate client report and clinician observation into one narrative, keeping client quotes marked and clinician observations attributed. In Assessment, connect the session material to the treatment plan theme and state the medical-necessity reasoning." The constraint block (use only supplied facts, placeholder anything missing) never changes across formats. The facts are yours in every format; only the frame moves.

BIRP for the Community Mental Health Medicaid Case

Now the Medicaid case: a community-mental-health client with schizoaffective disorder seen for skills-focused individual therapy, in a state where the Medicaid manual prefers BIRP for outpatient behavioral health documentation. Here the preference is not aesthetic. Medicaid programs audit at volume through Recovery Audit Contractors, and BIRP's structure produces, in four labeled boxes, exactly the audit elements a reviewer must find: a documented clinical presentation (Behavior), proof a skilled service occurred (Intervention), evidence the service did something (Response), and continuity of care (Plan). When the state manual says BIRP, writing DAP is not wrong in some abstract sense; it is friction at audit time, because the reviewer must hunt for elements your format scattered, and a hunting reviewer is an unhappy reviewer.

The worked example. Behavior: "Client presented with disorganized speech that became more linear as session progressed. He reported hearing voices 'two or three times' this week, down from daily at intake, and reported attending two of three scheduled day-program days. He stated, 'The bus thing still wrecks me,' referring to panic when riding public transit." Intervention: "Clinician delivered cognitive-behavioral coping-skills training targeting auditory hallucinations: reviewed and rehearsed the client's grounding sequence, then conducted graduated exposure planning for one bus ride with a support person this week." Response: "Client demonstrated the grounding sequence independently with one prompt, an improvement from requiring step-by-step coaching two weeks ago. He agreed to the bus-ride plan and identified his cousin as the support person." Plan: "Continue weekly individual sessions per treatment plan; client to attempt one supported bus ride; clinician to coordinate with day-program staff regarding attendance; next session to review exposure outcome." Every box is doing payer work. The Response section in particular is the one clinicians underwrite and Medicaid reviewers most need: it is the only place the note proves the service was not just delivered but received.

What the AI cannot supply here is everything that makes this note pass: the voices count and its delta from intake, the day-program attendance ratio, the quote, the observed speech change, and the demonstration-with-one-prompt detail. Those are your shorthand items. The model's contribution is the discipline of the four boxes and the professional register, which at 9:54 PM is worth real minutes, but the audit value lives entirely in the clinician-supplied specifics.

The format decides what stands in the foreground of the note; the payer, the setting, and the modality decide the format; and the clinician decides all of it, because the AI can reframe the footage but can never choose the camera angle.

GIRP for the 90853 Substance-Use Group

Group therapy billed under CPT 90853 has a documentation problem all its own: one session, eight clients, eight individualized notes, and the temptation to write one note eight times. Auditors know the temptation, and near-identical group notes across members are a classic recoupment trigger. GIRP is the structural antidote, because its first box forces individualization at the root: which of THIS client's treatment plan goals did THIS group session serve? Eight clients in the same relapse-prevention group are there under eight treatment plans, and the Goal section cannot be copy-pasted without the fraud becoming visible on its face.

The worked example, for one member of a Thursday relapse-prevention group with a substance-use focus. Goal: "Session addressed treatment plan Goal 2: client will identify personal relapse triggers and demonstrate two coping strategies, in support of sustained remission from alcohol use disorder." Intervention: "Clinician facilitated a structured relapse-prevention group (90853) using an MI-informed trigger-mapping exercise; clinician elicited this client's high-risk situation (payday Fridays with former drinking coworkers) and guided group feedback on refusal strategies." Response: "Client identified payday Fridays as his primary trigger, role-played one refusal script with another member, and rated his confidence at 6 of 10, up from 4 at the last group. He remained engaged throughout and gave constructive feedback to two peers." Plan: "Client will use the refusal script this Friday and report outcome at next group; continue weekly group per treatment plan; individual session scheduled for [date] to review craving log." Each of the other seven notes shares the Intervention's group-level description but diverges everywhere else, because each client's goal, response, and plan are their own.

The AI workflow for groups is a batch operation with a hard rule: you supply per-client shorthand (goal number, the client's specific contribution, the client's specific response, the client's specific plan item), and the prompt explicitly forbids reusing any client-specific sentence across notes. The group-level intervention description may repeat; everything else must trace to that client's shorthand line. This is the one place in the chapter where AI's speed is most seductive and most dangerous: it can produce eight plausible notes from one shorthand line, and seven of them would be fabricated. Eight shorthand lines in, eight notes out, eight separate verification passes before eight signatures.

The Format Decision Tree

Here is the decision logic, in the order the constraints actually bind. Question one: does the payer or program specify a format? Check the provider manual, the Medicaid behavioral health documentation manual for your state, and your agency's contract. If the answer is yes, that format wins, full stop; a beautifully reasoned DAP note in a BIRP-mandated program is a self-inflicted audit finding. Question two, if the payer is silent: is this service goal-driven or group-based? Group therapy (90853), intensive outpatient, partial hospitalization, ABA-adjacent skills programs, and any setting where the treatment plan is the unit of accountability point to GIRP, because goal linkage is the thing those settings are audited on. Question three: is the clinical work transaction-shaped or meaning-shaped? Skills training, exposure work, behavioral activation, structured CBT and DBT protocols decompose naturally into intervention-response pairs: BIRP. Insight-oriented, psychodynamic, existential, and relational work where the session's value is the understanding reached: DAP, with the medical-necessity chain carried in the Assessment. Question four, the override: what does your supervisor or clinical director require for consistency across the practice? A group practice that audits its own charts needs internal uniformity per program, and Jordan's 25 clinicians writing in five formats across three counties is a compliance officer's migraine.

Two cautions on the tree. First, SOAP remains a legitimate default for general outpatient work with commercial payers; this lesson adds formats to your repertoire, it does not banish the one you know. Second, format-switching mid-episode for the same client should be deliberate and rare, ideally at a treatment-plan review point, because a chart that changes shape every few sessions reads as disorganized in board complaint discovery, where the chart is read end to end as a story about your care.

Running One Session Through All Three Formats

The fastest way to internalize the differences is to take one set of session shorthand and have AI render it in all three formats, side by side. The prompt: "Below is session shorthand. Render it three times: once as DAP, once as BIRP, once as GIRP (assume treatment plan Goal 1: reduce depressive symptoms and restore occupational functioning). Use only facts in the shorthand; insert [CLINICIAN: VERIFY OR ADD] for any required element the shorthand does not contain. Past tense, third person." Run it on Maria's F33.1 shorthand from the previous lesson and study where the placeholders land, because the placeholders are the diagnostic readout: the GIRP version will demand a goal linkage her shorthand never stated, the BIRP version will want a crisper response-to-intervention than "brightened at end," and the DAP version will absorb everything comfortably. Where a format generates placeholders, it is telling you what that format requires you to capture in shorthand going forward.

This exercise also exposes the deepest point of the lesson: the facts do not change across formats, but the gaps do. A format is a completeness specification. BIRP makes a missing response observation visible; GIRP makes a missing goal linkage visible; DAP makes thin clinical reasoning visible, because an Assessment section with nothing in it has nowhere to hide. Choosing a format is choosing which omissions you will be forced to confront at writing time rather than at audit time. That is the supervisor's framing when you defend your choice: "I chose BIRP for this Medicaid case because the state manual prefers it and because the skills-training work produces intervention-response pairs the format showcases; here is the response evidence in every note." A defense names the payer rule, the modality fit, and the audit element the format foregrounds. "It's what I've always used" is not a defense; it is the absence of one.

The Applied Problem: Your Format Decision Tree

Your artifact is a one-page Format Decision Tree for your actual caseload, built in three steps. Step one: write the four questions in order (payer/program mandate? goal-driven or group setting? transaction-shaped or meaning-shaped work? practice-wide consistency rule?) with your real answers per program. Pull your top two payers' provider manuals and your state's Medicaid behavioral health documentation manual and record, with the manual section cited, whether any format is specified or preferred; if you work CMH or take Medicaid, this citation is the keystone of the tree, because where the payer speaks the payer wins.

Step two: run the three-format conversion exercise on one real (de-identified) or fabricated session per major service line you bill: an individual outpatient session, and if applicable a group 90853 and a skills-focused case. Paste the side-by-side prompt from this lesson, mark where each format generated placeholders, and write one sentence per service line stating which format you choose and why, in the supervisor-defense form: payer rule, modality fit, audit element foregrounded. These sentences go on the tree verbatim; they are what you say when a supervisor, an auditor, or your own future self asks why the chart looks the way it does.

Step three: update your shorthand capture template from the previous lesson to feed your chosen formats. If you chose BIRP for a service line, add "client response to intervention, observed" as a mandatory capture item; if GIRP, add "goal number served this session." Then test the whole loop once end to end: shorthand in, formatted draft out, line-by-line verification, every placeholder resolved by memory or honest omission. Done looks like: a one-page tree naming a format per service line with a cited payer source where one exists, a written one-sentence defense per choice, and a shorthand template updated so the format's required elements are captured at session time. Tape it next to the conversion workflow. The two documents together are your note-writing system.

Key Takeaways

• DAP, BIRP, and GIRP are camera angles on the same session: DAP foregrounds clinical reasoning, BIRP foregrounds the intervention-response transaction, GIRP foregrounds treatment-plan goal linkage. The AI can reframe the footage into any format, but choosing the angle is clinical judgment that depends on payer, setting, and modality.
• The decision tree binds in order: payer or program mandate first (some state Medicaid manuals prefer BIRP for outpatient behavioral health, and where the payer speaks the payer wins), then goal-driven or group settings toward GIRP, then transaction-shaped work toward BIRP and meaning-shaped work toward DAP, with practice-wide consistency as the override.
• DAP fits psychodynamic and insight-oriented sessions because forcing them into BIRP produces either falsified intervention-response pairs or vacant ones, and the medical-necessity chain still lives in the DAP Assessment: symptom, functional impairment, progress, ongoing need.
• BIRP earns its Medicaid preference because its four boxes deliver exactly what a Recovery Audit Contractor must find: documented presentation, proof of skilled service, evidence of client response, and continuity. The Response section is the most underwritten and the most audit-critical, and its specifics (the voices-per-week delta, the demonstration with one prompt) can only come from the clinician.
• GIRP is the structural antidote to the 90853 group-note problem: the Goal section forces individualization at the root, the group-level intervention description may repeat, and every client-specific sentence must trace to that client's own shorthand line. Eight clients means eight shorthand lines, eight verification passes, and eight separate signatures.
• A format is a completeness specification: run one session's shorthand through all three formats and the placeholders show you what each format demands you capture. Update your shorthand template to feed your chosen format, so the gaps surface at writing time instead of audit time.
• Defending a format choice to a supervisor takes one sentence naming the payer rule, the modality fit, and the audit element foregrounded. "It's what I've always used" is not a defense, and the clinician who signs the note owns the format choice along with every word inside it.