Clinicians

How to Read an ADHD Evaluation Report: A Guide for Physicians and Schools

September 11, 2026 · 6 min read

Last updated September 11, 2026

Physicians are asked to prescribe on the strength of them. School teams and university disability offices are asked to grant accommodations on the strength of them. Yet most people receiving ADHD evaluation reports were never taught how to read one, and the quality of what arrives varies from a comprehensive twelve-page document to a single paragraph from a telehealth visit. This is a short guide to telling the difference.

One: the referral question is stated

A competent report opens by saying who asked what. ‘Evaluate for ADHD and rule out anxiety and learning disorder’ is a referral question. ‘Patient reports difficulty focusing’ is not. Without a stated question you cannot judge whether the evaluation answered it, and the absence usually means the evaluator did not consider alternatives.

Two: developmental history with a source

Look for a childhood history and for where it came from: a parent interview, school records, report cards, or the client alone. Adult self-report of childhood symptoms is the weakest source, and DSM-5-TR requires onset before age 12. If the report says symptoms ‘have been present since childhood’ with no named source, treat that criterion as unverified rather than met.

Three: more than one informant and more than one setting

The diagnosis requires impairment in two or more settings. A report built on a single self-report scale cannot establish that. For children expect parent and teacher ratings; for adults expect the client plus a partner, parent, or colleague where possible. Discrepancies between raters are informative and a good report discusses them rather than averaging them away.

Four: the differential is visible on the page

The report should name the conditions that were considered and say how each was ruled in or out: anxiety, depression, trauma, sleep disorder, substance use, learning disorder, and, in adults, medical causes such as thyroid disease. A screening measure for each, or a documented interview section, is the evidence. A report that mentions no alternative explanation is a screen, not an evaluation, regardless of length.

Five: standardized measures with scores, not adjectives

Instruments should be named with the version, and results reported as standardized scores or percentiles, not as ‘elevated’ alone. For rating scales that means T-scores or symptom counts against the cutoff. For a continuous performance test such as the TOVA it means the attention comparison score and the individual variables. Scores let the reader judge the strength of the finding; adjectives do not.

Six: a diagnostic formulation that connects evidence to conclusion

The best two paragraphs in any report explain how the history, ratings, and test results fit together, and what does not fit. If the client has high self-reported symptoms and a normal continuous performance test, the report should say so and explain why the diagnosis is still, or is not, warranted. Then the recommendations should follow from the formulation: specific accommodations, a treatment sequence, referrals. Generic recommendations copied between reports are a red flag.

Red flags worth acting on

A single visit under thirty minutes. No collateral informant. No differential section. A diagnosis of ADHD ‘by history’ alone in an adult with no childhood corroboration. A report from a platform whose business model depends on the prescription. Any of these justify asking for a fuller evaluation before prescribing controlled substances or granting testing accommodations that universities and licensing boards will later scrutinize.

What you can reasonably ask for

Prescribers and schools are entitled to ask the evaluator for raw scores, the list of instruments, and the basis for the childhood-onset criterion. In Michigan and most states, licensed counselors, psychologists, and physicians can all produce a defensible report; the credential matters less than the content. If you refer patients for evaluation, the referral page describes what my reports include and how quickly they are returned.

The bottom line

Accept an ADHD report when it states its question, documents childhood onset from a real source, uses more than one informant, shows its differential, reports standardized scores, and explains its reasoning. Decline to act on one that does none of those, however confident its conclusion sounds.