Sign Up
Login
Try for Free
Login
An eating disorder assessment is a structured intake evaluation that combines a validated screener such as the SCOFF or EDE-Q, DSM-5-TR criterion mapping, eating and weight history, and a medical-risk appraisal with a level-of-care rationale. Therapists, psychologists, and counselors complete it at intake, coordinating with medical and nutrition colleagues. Most run 400 to 1,000 words.
Therapists, psychologists, and counselors, in coordination with medical and nutrition teams
Treating clinician, medical and nutrition team, supervisors, payers, auditors
400 to 1,000 words · 30 to 60 minutes by hand (clinical team estimate)
Structured intake assessment (compare: biopsychosocial assessment, intake note, substance use assessment)
At intake when disordered eating presents or a screener flags it; measures repeated at level-of-care decisions
APA guideline and payer expectations shape content in the US and Canada; Australia's MBS items legislate it
An eating disorder assessment is a composite clinical document: a diagnostic evaluation focused on eating pathology that combines a brief screener, a dimensional measure, DSM-5-TR criterion mapping, and a medical-risk appraisal with a level-of-care rationale. No single author created the composite, but its components have verifiable lineages: the SCOFF questionnaire (Morgan, Reid, and Lacey, BMJ 1999), the Eating Disorder Examination and its self-report EDE-Q (Cooper and Fairburn in the 1980s; Fairburn and Beglin 1994), and the EAT-26 (Garner and colleagues, 1982), all now read against DSM-5-TR (2022) and the APA practice guideline for eating disorders, 4th edition (February 2023). Clinicians and payers also call it an ED evaluation, an eating disorder intake, or, in US billing terms, a psychiatric diagnostic evaluation (90791) when disordered eating is the presenting problem. In Australia it feeds the Eating Disorder Treatment and Management Plan (EDTMP).
The distinction that carries this page: almost nothing about this document is required by law in the United States or Canada. The near-universal beliefs that a clinician must weigh the client, must order labs, and must obtain medical clearance before outpatient therapy are APA guideline recommendations (rated 1C) and payer expectations, not statutes. Australia is the outlier: the MBS eating disorder items, in force since 1 November 2019, write assessment content directly into national payment law, down to a global EDE-Q score of 3 or higher as an eligibility threshold for non-anorexia diagnoses. Everywhere else, the eating disorder assessment is a focused specialization of the biopsychosocial assessment, one level deeper than a general intake note: the same frame, plus eating-specific history, behavior domains, a medical-risk appraisal, and a level-of-care argument.
Psychologists, professional counselors, clinical social workers, and marriage and family therapists complete it at intake in outpatient practices, college counseling centers, and eating disorder specialty programs; psychiatrists and primary care teams run the parallel medical workup, and dietitians write the separate nutrition assessment. Reach for it when disordered eating is the presenting concern, or when a screener flags it partway into treatment. At later decision points, re-administer the measures rather than redoing the whole document (an outcome measure note handles those re-administrations). When substances share the picture, a substance use assessment runs alongside rather than instead: the two documents argue different risks.
Referral context and presenting concerns. Who sent the client and why, and the concern in the client's own words, quoted briefly. Then onset, course, and prior treatment episodes with response. Pitfall: opening with your formulation instead of the client's words; on a 90791 review, the chief complaint in the client's own language is the first element a payer reviewer looks for.
Screening and standardized measures. Name the instrument, the date, who administered it, and where the scored form is filed. A SCOFF at the front desk plus a self-report measure in the waiting room is a common pairing; the interview-based EDE belongs to specialty settings with 45 to 75 minutes to spend on it. Pitfall: a score with no instrument, date, or administrator attached. Australia's Professional Services Review has cited outcome tools that were "incorrectly scored" in mental health services; an unattributed number reads the same way.
Eating, weight, and behavior history. Trajectory and pattern at the domain level: restriction, binge episodes, compensatory behaviors, driven exercise, food-repertoire narrowing, and the share of the day spent preoccupied with food, weight, and shape. Record which domains are endorsed and which are denied; the quantified detail lives in the standardized measure and the medical record, referenced by location. Pitfall: reproducing weight numbers and behavior counts in the narrative. The assessment travels further than the measure does (releases, record requests, the client's own portal), so domain-level narrative plus a pointer keeps it clinically complete and safe to share.
Medical risk and the handoff. The section that separates a defensible assessment from a liability. Document the medical-risk domains reviewed (vital signs, orthostatic symptoms, labs, ECG when indicated), who owns each one, the referral made, the release signed, and when results are due back. The therapist's duty is the handoff and its follow-through, not the values. Pitfall: transcribing lab values and vitals into the therapy record. "CBC and metabolic panel ordered by Dr. R., results to this office by 08/20" shows the duty discharged; copied numbers add nothing and go stale.
Mental status, co-occurring conditions, and suicide risk. A full mental status exam, screening for depression, anxiety, trauma, and substance use, and a suicide-risk assessment with stated reasoning. Eating disorders carry elevated suicide risk, so silence here is the worst omission available. Pitfall: a checked risk box with no reasoning; billing analysts place checkbox risk assessment among the leading documentation-based 90791 denials.
DSM-5-TR diagnosis and severity. Map the endorsed findings to criteria, state the diagnosis and specifier, and mark it provisional if medical findings or collateral records are pending. Pitfall: a diagnostic label with no criterion trail. When documented findings cannot be matched to the billed diagnosis, the denial arrives as an information or medical-necessity error (the CARC 16 and 50 family).
Level-of-care rationale. The recommended level (outpatient, intensive outpatient, partial hospitalization, residential, inpatient), the criteria considered, and the facts supporting the choice. This is the paragraph payers actually read. Pitfall: arguing level of care from a DSM severity band. Payer criteria weight physiologic instability and rate of change over any static category, so anchor the rationale in current status, trajectory, and the medical team's findings.
Collaborative plan and communication. Therapy modality and frequency, the medical monitoring schedule, nutrition referral, releases on file, who communicates with whom, and the review date. Pitfall: "referred to PCP" with no recipient, date, or release. For auditors and licensing boards alike, multidisciplinary communication that is not documented did not happen.
EATING DISORDER ASSESSMENT Client: [initials] DOB: ______ Date: ______ Clinician: ________ Referral source: ______________ Reason: ___________________________ PRESENTING CONCERNS (client's words): HISTORY (onset, course, prior treatment and response): MEASURES (instrument | date | administered by | filed where) _________________ | ______ | _______________ | _____________________ EATING / WEIGHT / BEHAVIOR HISTORY (domain level; detail in measure) Domains endorsed: ____________________ Denied: _____________________ Weight status and trajectory: assessed by _________ (medical record) MEDICAL RISK AND HANDOFF Domains reviewed (vitals / orthostatics / labs / ECG): ______________ Referral to: ____________ Date: ______ Release signed: [ ] ______ Results due: ______ Priority symptoms flagged: ____________________ MSE AND CO-OCCURRING CONDITIONS: SUICIDE RISK (findings and reasoning): DSM-5-TR DIAGNOSIS (criteria mapped; provisional if pending): LEVEL-OF-CARE RATIONALE (status, trajectory, criteria used): COLLABORATIVE PLAN (therapy / medical / nutrition / review date): Clinician signature/credentials: Date signed:
Free to use and share, no signup. The PDF includes a one-page cheat sheet with section-by-section pitfalls and a pre-sign checklist; the DOCX is the blank template, ready to adapt.
Scenario: adult outpatient, self-referred after a partner raised concern, evaluated by a psychologist in private practice. Screening instruments are named without scores, and medical values stay with the medical team, deliberately. All details are fictional.
Client: R.L., 29 · Date: 08/11/2026 · Clinician: D. Marsh, PsyD · Service: Diagnostic evaluation (90791), 74 min, in office
Referral and presenting concerns: Self-referred after her partner raised concern about restrictive eating and withdrawal from shared meals. Client: "I plan my whole day around not eating in front of people." Reports rigid food rules beginning in her mid-20s, a gradually narrowing food repertoire, and growing distress around work lunches. One prior course of general therapy; eating was not addressed. No prior eating disorder treatment.
Measures: SCOFF administered verbally at the start of session; screen positive. EDE-QS completed in the waiting room; global score in the clinical range. Scored forms filed in the measures section of the chart and shared with the medical team under the release below; scores are not repeated in this narrative.
Eating, weight, and behavior history: Restriction and food-repertoire narrowing endorsed. Episodes of loss-of-control eating endorsed. Compensatory behavior endorsed at intake (a driven exercise pattern); frequencies are documented in the EDE-QS and the medical record, not here. Client describes preoccupation with food and shape occupying most of the day. Weight status and trajectory will be assessed by the client's primary care physician; client declined in-office weighing today, which was accepted and documented, and a records request went out 08/11/2026.
Medical risk and handoff: Review of systems notable by client report for lightheadedness on standing and cold intolerance; no fainting. This office does not collect vitals. Same-week appointment confirmed with A. Osei, MD (primary care) for vital signs, orthostatic check, CBC, and comprehensive metabolic panel, with ECG at Dr. Osei's discretion per APA guideline indications. Release of information signed 08/11/2026 for two-way communication. Lightheadedness flagged to Dr. Osei's office as the priority item; results due to this office by 08/20/2026.
Mental status, co-occurring conditions, and suicide risk: On time, casually dressed, cooperative, guarded when food is discussed directly. Mood "worn down," affect congruent and constricted. Thought process linear; no psychotic symptoms. Depression and anxiety screens completed and filed with the measures. Alcohol and substance use screened; denies problematic use. Suicide risk: denies current ideation, plan, and intent; no history of attempts; connected to partner and sister and engaged in seeking care. Risk judged low today, with reasoning noted and monitoring planned at each session given the elevated suicide risk that travels with eating disorders.
DSM-5-TR diagnosis: Endorsed pattern of restriction, loss-of-control eating, and compensatory behavior mapped against DSM-5-TR feeding and eating disorder criteria in interview. Provisional diagnosis: other specified feeding or eating disorder (OSFED), pending medical findings and prior records; differential includes bulimia nervosa and anorexia nervosa, to be revisited when weight-status data arrives from Dr. Osei.
Level-of-care rationale: Outpatient level recommended at this time: stable function at work, no reported fainting or rapid change in weight status, motivation for treatment, and daily support at home. Criteria for a higher level of care (physiologic instability, rapid trajectory change, inability to interrupt behaviors at outpatient intensity) reviewed and not met on today's information. This rationale will be re-argued, not assumed, when the medical results return.
Collaborative plan: Weekly individual therapy, CBT-E informed, starting 08/18/2026. Primary care evaluation this week as above; case call with Dr. Osei after results. Referral offered to a registered dietitian experienced in eating disorders; client will decide by next session. Re-administer the EDE-QS at week 8. Partner session offered for meal-support psychoeducation; client considering. Diagnosis and level of care to be reviewed 09/08/2026, or sooner if medical findings require. Signed: D. Marsh, PsyD, 08/11/2026.
This sample is fictional and for educational purposes. It does not describe a real patient.
Writing these after every session? BastionGPT drafts complete notes from bullets, dictation, or a transcript.
Generate a note from bulletsAn eating disorder assessment is part of the designated record set: the client can request it, the medical and nutrition team will rely on it, and payers can audit it. That reach should shape what goes in it. Keep behavior detail at domain level with a pointer to the measure, keep weight data with the medical team, and the document stays accurate, useful, and safe to share with a client who may be acutely distressed by numbers. Authorship is a scope question in every jurisdiction: in the US, psychologists, counselors, clinical social workers, and marriage and family therapists complete the evaluation and assign the diagnosis within state scope of practice; in several Canadian provinces, formally communicating a diagnosis is a controlled act under provincial health-professions law; in Australia, only a GP, consultant psychiatrist, or consultant paediatrician may make the EDTMP eligibility decision, whatever assessment data allied health contributes. A dietitian's nutrition assessment is a companion document, not a substitute. Retention is law, and it is local: commonly 6 to 10 years for adults under US state rules (longer for minors), at least 10 years from the last interaction (or 10 years past the client's 18th birthday) in Ontario, and generally 7 years for adults with minors' records kept to age 25 in Australia.
For US insurance work, the billing container is the psychiatric diagnostic evaluation (90791), and coverage turns on medical necessity under section 1862(a)(1)(A) of the Social Security Act. Medicare contractor fact sheets expect a complete history, a mental status exam, a diagnosis, and an initial treatment plan; once-per-episode and once-per-year limits are payer rules, not CPT rules; and managed behavioral health organizations make prior authorization the leading initial-evaluation denial (CARC 197, precertification absent). None of that mandates a form: in the US and Canada, the format is a convention and the content is the requirement. Australia inverts this. MBS Explanatory Note AN.36.2 legally enumerates what the plan must document (assessment results, patient needs, goals and actions, referrals, and a review date), eligibility for non-anorexia diagnoses is written into the item rules (a global EDE-Q score of 3 or higher plus behavioral and clinical indicators), and reviews around sessions 10, 20, and 30 gate access to the full 40 psychological and 20 dietetic sessions. Build the assessment so its measures re-administer cleanly (the outcome measure note covers that cadence) and the medical-necessity argument reads in one place.
No OIG or CERT report isolates the diagnostic evaluation codes, and none gives an eating-disorder-specific error rate, so the enforcement record here is analogue and worth reading that way. The nearest US signal is the HHS-OIG psychotherapy audit (A-09-21-03021, 2023), which estimated $580 million in improper Medicare psychotherapy payments in a single year, with missing treatment plans and undocumented time among the causes. The nearest Australian signal is the Professional Services Review's run of mental-health-plan findings: in the April 2023 Director's Update, a practitioner repaid $896,000 where plans "did not include the level of information required" by the item descriptor, and in May 2023 another repaid $330,000 where plans were "template based" with generic goals. The BastionGPT Clinical Advisory Board sees the same errors most often in eating disorder assessment reviews:
BastionGPT is specifically trained, tuned, and clinically tested on eating disorder assessments.
See how clinicians use it day to day on the AI therapy notes page.
Many BastionGPT users report saving more than 90 minutes per day on documentation.
HIPAA-compliant with a signed BAA on every plan. Your data is never used to train models. BastionGPT drafts, you review and sign.
The written document usually runs 400 to 1,000 words and takes 30 to 60 minutes to write by hand. Administration time depends on the instruments: the SCOFF takes under a minute, the EDE-Q about 15 to 20 minutes as self-report, and the full EDE interview 45 to 75 minutes, which is why the interview stays in specialty settings. Most outpatient clinicians pair a brief screener with a self-report measure inside a standard intake hour.
No law in the United States or Canada requires a therapist to weigh a client, order labs, or obtain medical clearance before outpatient eating disorder treatment. Those expectations come from the APA practice guideline (rated 1C recommendations) and from payer level-of-care criteria: real reasons to arrange them, but a different kind of reason than a statute. If weighing is deferred or declined, document the clinical reasoning and route weight monitoring to the medical team. The Academy for Eating Disorders' medical care standards are the reference most teams hand their medical partners.
In the US, yes, within state scope of practice: psychologists, counselors, clinical social workers, and marriage and family therapists complete diagnostic evaluations and assign DSM-5-TR diagnoses, and the counselor and MFT professions gained permanent Medicare billing in 2024. The medical-risk half of the picture still requires medical involvement, which is why the handoff section matters. In several Canadian provinces, formally communicating a diagnosis is a controlled act. In Australia, allied health can contribute assessment data, but only a GP, consultant psychiatrist, or consultant paediatrician can make the eating disorder plan eligibility decision.
90791, the psychiatric diagnostic evaluation (90792 when a prescriber adds medical services); the Medicare national non-facility rate for 90791 is $173.35 for 2026. Medical necessity is carried by the content: a chief complaint in the client's words, full history, complete mental status exam, risk assessment with reasoning, a criterion-mapped diagnosis, and an initial plan. Frequency limits such as once per provider per year are payer rules, not CPT rules, and prior authorization is the top managed-care denial for initial evaluations, so check both before the appointment.
The generic intake note and biopsychosocial assessment survey every domain at even depth. The eating disorder assessment keeps that frame and goes deeper in one place: eating-specific behavioral history, weight-history coordination with the medical team, a medical-risk appraisal, and a level-of-care rationale. When substance use shares the presentation, run a substance use assessment alongside it: withdrawal risk and refeeding risk are different arguments, and payers read them separately.
No US or Canadian regulation names one, so the choice is clinical. The SCOFF is fastest, with pooled sensitivity of 0.86 and specificity of 0.83 across 25 validation studies, and the same review (Kutz et al. 2020) concluded the evidence is insufficient for the full range of DSM-5 eating disorders, so treat a positive as a flag, not a finding. The EDE-Q adds subscale and global scores and is the instrument Australia wrote into its MBS eligibility rules, which makes it the default wherever Australian portability matters. Whichever you pick, document instrument, date, administrator, and filing location.
The items (90250 to 90257 for GP plans, with psychiatrist, paediatrician, and telehealth equivalents) took effect 1 November 2019 and legislate the content: Explanatory Note AN.36.2 requires the plan to document assessment results, patient needs, goals and actions, referrals, and a review date. Anorexia nervosa qualifies on clinical diagnosis; other eligible diagnoses need a global EDE-Q score of 3 or higher plus behavioral and clinical indicators from the item rules. Only a GP, consultant psychiatrist, or consultant paediatrician can author the plan, reviews fall around sessions 10, 20, and 30, and the scheme caps at 40 psychological plus 20 dietetic sessions in 12 months, one plan per patient per year.
Retention is set by state, provincial, and professional rules, not by the document type. US states commonly require 6 to 10 years for adult records and longer for minors. Ontario colleges expect at least 10 years from the last interaction, or 10 years past the client's 18th birthday. Australian boards generally expect 7 years from the last entry for adults, with minors' records kept until age 25. The assessment anchors the diagnosis and the level-of-care decision, so it should survive as long as anything in the chart.
Yes. Give it an intake transcript, a dictation, or bullets, and it drafts the full assessment: domains sectioned, client quotes preserved, the medical handoff documented, and the level-of-care rationale argued from status and trajectory. It also flags stray numbers that belong in the measure rather than the narrative. BastionGPT is HIPAA-compliant with a signed BAA on every plan, and your data is never used to train models.
Educational content, not legal or billing advice. Sample notes are fictional. Follow your organization's policies and your board, payer, and jurisdiction requirements.