Differential Diagnosis Template for Therapists (Write-Up Format & Worked Example)

What a Differential Diagnosis Write-Up Is

A differential diagnosis write-up is the structured, candidate-by-candidate section of a diagnostic document where you compare the plausible explanations for a client's presentation and commit — in writing — to what the evidence supports, what it argues against, and what remains unresolved. Each candidate diagnosis gets its own block: evidence for, evidence against, a status label (confirmed, provisional, ruled out, or deferred), and a note on what would change the picture. It is the reasoning table of a diagnostic evaluation.

It is worth distinguishing this from a diagnostic formulation, because the two terms are often used interchangeably. The formulation is the full narrative document: it integrates history, mental status findings, testing, and collateral data into a prose justification of your diagnostic conclusions and links them to treatment. The differential write-up is the structured comparison that sits at the core of that narrative. Think of the differential as the table and the formulation as the story wrapped around it — most complete evaluations contain both, and this page gives you a reusable template for the table itself.

A documented differential is also your best protection against the most common diagnostic errors — premature closure, anchoring on the referral diagnosis, and confirmation bias — because the format forces you to write an evidence against line for the diagnosis you already favor. Diagnostic report guidelines, including the ETS guidelines for diagnostic reports, expect exactly this: explicit rule-out of mimicking conditions with stated rationale, not a bare conclusion.

Scope note: rendering a mental health diagnosis is restricted to licensed professionals acting within their scope of practice and jurisdictional rules. This template supports the documentation of diagnostic reasoning; it is not a diagnostic instrument and does not substitute for clinical judgment, supervision, or access to the DSM-5-TR itself.

When You Need a Documented Differential

• Intake and diagnostic assessments where more than one condition could plausibly account for the presentation
• Presentations with heavily overlapping symptom pictures — mood instability, concentration problems, sleep disturbance, and irritability each appear in a half-dozen diagnostic categories
• When a client arrives with prior diagnoses that conflict with each other or with your clinical impression
• Insurance and audit defensibility — a chart that shows why a diagnosis was assigned (and what was considered and excluded) withstands review; a bare code does not
• Referrals for testing, psychiatry, or higher levels of care, where the receiving provider needs your reasoning, not just your conclusion
• Supervision and training, where the write-up makes diagnostic reasoning visible and correctable

The Six-Step Screen Behind Every Differential

Before the candidate-by-candidate comparison, run the systematic screen from Michael First's DSM-5-TR Handbook of Differential Diagnosis — the standard framework for psychiatric differential reasoning. The six steps, briefly:

1. Rule out malingering and factitious presentation. Consider whether external incentives (legal, financial, disability) or a drive to occupy the sick role could be shaping the symptom report. This requires documented consideration, not accusation.
2. Rule out substance etiology. Could the symptoms be attributable to intoxication, withdrawal, or the physiological effects of a medication or substance? Temporal sequencing between use and symptoms does the diagnostic work here.
3. Rule out an etiological medical condition. Coordinate with the client's medical provider on plausible medical mimics (thyroid disease and depressive symptoms being the classic example).
4. Determine the primary disorder(s). This is where the candidate-by-candidate comparison below lives.
5. Differentiate adjustment disorder from other specified/unspecified conditions. When symptoms follow a stressor but full criteria for a specific disorder are not met, decide between adjustment disorder and the other specified/unspecified categories.
6. Establish the boundary with no mental disorder. Confirm the presentation crosses the threshold of clinically significant distress or impairment rather than an expectable reaction to circumstances.

The Handbook's decision trees are worth owning and consulting directly — describe your reasoning through the steps in your write-up, but send readers (and trainees) to the Handbook itself for the full trees. The template below builds the six steps in as a compact screening block so none of them are silently skipped.

The Differential Diagnosis Template

The blocks below are the reusable structure. Repeat the candidate block once per diagnosis under serious consideration.

Try this template in My Clinical WriterDownload Differential Diagnosis Template (.docx)ShareFeedback

Feedback on this template

differential-diagnosis-template

What’s wrong or missing?Email, only if you’d like a reply (optional)

Please don’t include any client or patient details.

CancelSend

Rule-In and Rule-Out Language That Holds Up

The status labels are where differential write-ups most often go wrong, so it pays to use them precisely.

Provisional is the DSM-endorsed label for diagnostic uncertainty. Append "provisional" when you believe the client will likely meet full criteria but you cannot yet confirm it — because history is incomplete, collateral is pending, or the decisive observation (a fully characterized hypomanic episode, for instance) has not yet been made. A provisional diagnosis is a real diagnosis with a stated confidence level, and it should always travel with the missing information named and a reassessment date attached.

"Rule out" is informal shorthand, not a DSM convention. "R/O Bipolar II" communicates "actively under investigation" between clinicians, and there is nothing wrong with using it that way — but it is not an official specifier, it is not a diagnosis, and it carries a specific documentation risk: rule-outs left unresolved in the chart, or worse, migrating onto claims as though they were established diagnoses. If a payer is billed under a code your own chart labels "R/O," your documentation contradicts your billing. Every R/O should either graduate to a status label (confirmed, provisional, ruled out) or be explicitly closed.

Document the transition when a provisional diagnosis resolves. When mood charting, collateral history, or longitudinal observation settles the question, write it down: "Bipolar II Disorder, previously provisional (see evaluation of [date]), is now confirmed based on prospectively charted four-day hypomanic episode corroborated by partner report; provisional qualifier removed." The dated transition note is what makes the chart internally coherent from intake through treatment.

Ruled out means you can say why. "Ruled out" without a stated reason is an assertion, not reasoning. One or two sentences naming the disconfirming evidence — or the hierarchy rule that excludes the diagnosis — is enough.

Deferred means the question cannot be responsibly answered yet. Personality disorder diagnoses at a first session are the classic case: the defining feature is a pervasive, stable pattern, which a single cross-sectional interview often cannot establish. Deferring with a plan is more defensible than either premature assignment or premature exclusion.

Worked Example: Bipolar II vs. Borderline Personality Disorder

This is arguably the hardest common differential in outpatient practice: both presentations involve mood instability, impulsivity, irritability, and self-harm risk, and the treatment implications of getting it wrong run in opposite directions (mood stabilization and relapse prevention vs. a structured psychotherapy such as DBT, with different medication expectations). The research literature — most systematically the Bayes and Parker line of work in Current Psychiatry Reports — offers a crucial guide for the evidence for / evidence against lines: the features that actually differentiate the two conditions are not the ones clinicians most often reach for.

More differentiating: family history of bipolar disorder; onset and course (episodic and autonomous vs. pervasive and interpersonally reactive); the phenomenology of the elevated periods (euphoric, energized, sleep-curtailed episodes lasting days vs. hours-long mood shifts triggered by interpersonal events); identity disturbance and abandonment sensitivity; and stability of self-concept.

Less differentiating (despite intuition): impulsivity, self-harm history, childhood trauma history, overall comorbidity rates, and gender distribution — these occur at meaningful rates in both conditions and cannot carry a differential on their own.

Two honesty notes before the example. First, the bipolar II/BPD boundary is contested territory in the literature, and confounded or genuinely comorbid presentations occur in roughly 20% of cases — the differential is not always resolvable at intake, and a write-up that pretends otherwise is less defensible, not more. Second, the DSM duration anchor for a hypomanic episode is at least four days of a distinct, persistent change in mood and activity; sub-threshold elevated periods shorter than that are precisely the zone where this differential is hardest.

The following example is a fictional composite created for teaching purposes; it does not describe a real client.

Notice what the example models: the favored diagnosis still gets an evidence against line, the personality disorder question is deferred rather than forced, comorbidity is named as a live possibility rather than a failure, and every unresolved label has a plan and a date attached.

How to Write It Step by Step

Step 1: Inventory symptoms before touching diagnostic categories. Onset, duration, frequency, severity, context, functional impact, and the client's own words. Premature categorization at this stage is where anchoring bias enters.

Step 2: List candidates generously. Include the mimics of your leading candidate even if you doubt them — the point of the document is to show they were considered.

Step 3: Run the six-step screen. Malingering/factitious, substance, medical, adjustment-vs-other-specified, boundary with normality. Two sentences each is often enough; silence on a step reads as an omission.

Step 4: Write the evidence-for and evidence-against lines for every candidate — especially the one you favor. The evidence-against line on your leading diagnosis is the single highest-value sentence in the document. Weight the genuinely differentiating features (course, phenomenology, family history, identity stability) over the intuitive but shared ones (impulsivity, self-harm, trauma history).

Step 5: Assign a status label to every candidate. Confirmed, provisional, ruled out, or deferred — no candidate leaves the document unlabeled, and no "R/O" survives without a resolution path.

Step 6: Attribute symptoms in the final impression. When more than one diagnosis stands, say which symptoms belong to which. This prevents double-counting and makes the comorbidity clinically usable.

Step 7: Date the plan. Every provisional and deferred label gets the specific missing data named and a reassessment date. Document the transition when it resolves.

Common Mistakes

1. Premature closure. Assigning the first plausible diagnosis and never writing the differential. If the presentation could be two things, the chart should show both were weighed.

2. Candidates without evidence-against. A differential where every candidate has only supporting evidence is a list, not a comparison — and it signals confirmation bias to any reviewer.

3. Criteria dumping. Reproducing DSM-5-TR criteria sets verbatim in the write-up adds length, not reasoning (and the criteria are APA-copyrighted). Paraphrase at the symptom-domain level, cite criterion letters at most, and keep the manual at your desk.

4. Unresolved "R/O" fossilizing in the chart. A rule-out written at intake and never revisited is the most common audit finding in diagnostic documentation. Every R/O needs a resolution path and an end date.

5. Treating comorbidity as failure. Some pairs genuinely co-occur — forcing an either/or where the evidence supports both produces a tidy chart and a wrong one. Equally, avoid the opposite error of listing both because the data would not resolve; that is what provisional and deferred are for.

6. Leaning on non-differentiating features. Impulsivity, self-harm, and trauma history feel diagnostic but are shared across many of the hardest differentials. Build your evidence lines from course, phenomenology, family history, and longitudinal pattern.