Anxiety Treatment Plan Template (with Worked Examples by DSM-5 Subtype)
Outline
Authored by Dr. Hannah Lin, counselling psychologist trained in CBT, ACT, and IFS, with a decade of clinical practice in anxiety and complex trauma.
An anxiety treatment plan is not one document. It is at least four: the plan you write for generalized anxiety looks little like the one for panic disorder, social anxiety, or specific phobia, even though all four share the same template skeleton. The subtype drives the goals, the SMART objectives, the modality choice, and the outcome measure. A plan that treats anxiety as a single homogeneous category tends to read well to the client and poorly to the audit reviewer, because the interventions named never quite match what the diagnosis calls for.
This guide gives you a copy-ready template, walks through the differential between GAD, panic, social anxiety, and specific phobia at the planning stage, fills in worked SMART objectives for each subtype, and notes the modality fit and outcome-measurement anchors that payers expect to see. If you need the broader treatment-planning context first, the treatment plan templates and outcomes tracking hub is the parent piece. For the generic skeleton that every modality bends to fit, see the blank treatment plan template.
Educational content for licensed therapists, not clinical or legal advice. Diagnostic and treatment decisions sit inside formulation; what follows is a planning scaffold, not a protocol.
What an anxiety treatment plan must contain
A defensible plan covers ten sections, in the same order most US and Australian payers expect. The structure mirrors the broader skeleton in the counseling treatment plan template, with one subtype-specific addition: the document should name the anxiety subtype explicitly, not just “anxiety,” because the subtype determines the active ingredient of treatment.
- Client and clinician information. Name or ID, date, clinician name and credential, plan number, review date.
- Presenting concern. One or two sentences in the client’s words. “I worry about everything” reads differently from “I have these waves where I think I’m dying,” and both belong on the page.
- Diagnostic impression. DSM-5-TR or ICD-10 code with the specific anxiety subtype. F41.1 (generalized anxiety), F41.0 (panic disorder), F40.10 (social anxiety), or F40.2x (specific phobia) are the four most common.
- Problem list. Two to four targets drawn from the presenting concern, framed in terms of behaviour, mood, or functional impact rather than DSM language.
- Long-term goals. One to two outcome-level statements in client-recognizable language. Plain English, not “reduce symptomatology.”
- SMART objectives. Two to four per goal, measurable and time-bound. This is the section that varies most by subtype.
- Interventions. Named modality (CBT, exposure-based CBT, ACT, ERP, or a blend) plus the specific techniques you will use. “CBT” by itself does not pass audit; “thought records, behavioural experiments, interoceptive exposure, worry exposure” does.
- Measurable outcomes. A validated measure tied to the subtype: GAD-7 for generalized anxiety, PDSS for panic, LSAS or SPIN for social anxiety, and a subjective units of distress (SUDS) hierarchy for specific phobia.
- Frequency, duration, and review cadence. Weekly is standard. A 12 to 16 session course is typical for first-episode anxiety; longer arcs are defensible if you document why.
- Risk and safety, signatures. Anxiety presentations carry a non-trivial suicide-risk overlap (especially with comorbid depression), so risk gets documented even when low.
A plan that omits any of these sections will struggle on audit. The good news is that an anxiety plan typically fits on one to two pages once you have the subtype clear.
A blank template you can copy
The template below is copy-ready. Square brackets mark the spots you fill in. Drop it into your EHR or a Word document, keep the headings, and trim only what your setting explicitly does not require.
The DSM-5 anxiety differential at the planning stage
The most useful thing you can do before drafting the document is to spend ten minutes confirming which subtype you are planning for. Misallocating the subtype tends to push the plan toward generic “anxiety management” objectives that look fine on paper and fail to deliver the active ingredient the client actually needs. The four common adult anxiety presentations split cleanly at the planning stage on three dimensions: the cue, the feared consequence, and the avoidance pattern.
Generalized anxiety disorder (F41.1) cues to almost anything that could go wrong, the feared consequence is uncertainty itself, and the avoidance pattern is mental rather than behavioural (worry as avoidance of feared outcomes). Plans pivot on worry exposure, intolerance-of-uncertainty work, and behavioural experiments that test catastrophic predictions. GAD-7 is the standard measure.
Panic disorder (F41.0) cues to interoceptive sensations (heart rate, dizziness, shortness of breath), the feared consequence is a catastrophic interpretation of bodily sensation (dying, going crazy, losing control), and the avoidance pattern is both safety behaviours and agoraphobic restriction. Plans pivot on interoceptive exposure, cognitive restructuring of catastrophic misinterpretation, and in-vivo exposure to avoided situations. PDSS or the panic-specific items on the GAD-7 are the standard measure.
Social anxiety disorder (F40.10) cues to scrutiny by others, the feared consequence is humiliation or rejection, and the avoidance pattern is both overt avoidance and covert safety behaviours (over-rehearsing, post-event rumination). Plans pivot on graded behavioural experiments, attention-bias work, and post-event rumination interventions. LSAS or SPIN is the standard measure.
Specific phobia (F40.2x) cues to a specific stimulus, the feared consequence is harm from the stimulus, and the avoidance pattern is direct stimulus avoidance. Plans pivot on a hierarchical in-vivo or imaginal exposure protocol with SUDS-rated steps. A SUDS hierarchy plus session-by-session SUDS ratings is the standard measure.
When the plan you write does not name the subtype and choose interventions accordingly, it tends to default to generic relaxation and psychoeducation, which is the part of the toolkit with the weakest effect-size literature for clinical anxiety. The plan that explicitly names the active ingredient is both more clinically useful and more defensible. For the F-code reference on GAD specifically, see the F41.1 generalized anxiety disorder guide; the diagnostic detail there pairs with the formal DSM-5-TR criteria.
Worked SMART objectives by subtype
The most common failure point in an anxiety treatment plan is the objectives section. “Reduce anxiety” is not a SMART objective; “reduce GAD-7 from 16 to 8 by session 12” is. The four sets below are templates you can adapt to the specific client, with the diagnostic subtype determining the active ingredient. (For the broader SMART-objective library across presentations, the treatment plan goals and objectives examples companion piece carries 30+ worked objectives.)
Generalized anxiety disorder (F41.1)
Panic disorder (F41.0)
Social anxiety disorder (F40.10)
Specific phobia (F40.2x)
The pattern across the four subtypes is the same: name the active ingredient (worry exposure, interoceptive exposure, behavioural experiment, in-vivo hierarchy), tie it to a measure, set a time-bound target, and document the discharge criterion. A plan that follows this structure tends to clear audit on the first pass.
Modality choice: CBT, exposure-based CBT, and ACT
The dominant evidence base for adult anxiety is cognitive-behavioural, with subtype-specific exposure protocols carrying the most consistent effect sizes. The American Psychological Association’s evidence-based treatment summary for GAD and the equivalent panic and social-anxiety pages catalogue the protocols by subtype, and the NICE guideline on generalized anxiety disorder and panic disorder (CG113) is the corresponding UK reference. The practical question at the planning stage is not “is CBT the right modality” (it almost always is, for first-episode adult anxiety) but “which CBT protocol fits the subtype, and is there a reason to deviate.”
CBT-classic with thought records and behavioural activation fits generalized anxiety. Exposure-based CBT with interoceptive plus in-vivo exposure fits panic. Cognitive therapy for social anxiety (Clark and Wells model) fits social anxiety. Hierarchy-based in-vivo exposure fits specific phobia. The active ingredient differs across subtypes even though the umbrella term is the same.
ACT is the most defensible second-line modality for adult anxiety, particularly when the client has done CBT before and not responded, when the anxiety is chronic and tightly tied to avoidance of valued activity, or when the formulation suggests cognitive fusion and experiential avoidance are central. ACT plans replace the symptom-reduction primary objective with a values-consistent action objective, and use the AAQ-II for psychological flexibility alongside the subtype-specific measure. The blank treatment plan template walks through ACT framing in more detail.
Where you should not casually write “CBT” without naming the protocol is when the subtype is panic or specific phobia. Both have specific exposure protocols that are not interchangeable with generic CBT techniques, and “CBT” without “interoceptive exposure” or “in-vivo hierarchy” on the plan reads as under-specified to a reviewer who knows the literature.
Measurement-based care anchors
A defensible plan names a validated measure, sets a baseline, sets a target, and re-administers on a known cadence. The GAD-7 is the most commonly used measure across anxiety subtypes, partly because it is brief (seven items, 90 seconds to complete), free to use, and well-anchored on cut-scores (5 = mild, 10 = moderate, 15 = severe). A useful default is to administer the GAD-7 at every session for the first six sessions and then every four sessions thereafter; this lets you catch non-response early without burdening the client with paperwork.
For panic, the Panic Disorder Severity Scale (PDSS) is the published gold standard but takes longer; many practitioners use GAD-7 plus a brief panic-attack frequency log for the same effect. For social anxiety, the Liebowitz Social Anxiety Scale (LSAS) is the published standard, and the briefer Social Phobia Inventory (SPIN) is a reasonable alternative for weekly use. For specific phobia, a clinician-rated SUDS hierarchy with weekly re-rating substitutes for a published scale.
The measurement-based care guide walks through the broader rationale and a sample workflow for embedding measures in the session rhythm; the short version is that the plan should not just name the measure but show the cadence and the target on the page.
Payer and audit defensibility notes
The plans that survive audit cleanly share three features: the subtype is named in the diagnostic impression, the interventions section names specific techniques rather than just the modality, and the objectives are quantified against a validated measure. The plans that get queried tend to fail on one of those three.
A second pattern worth knowing: when a reviewer asks for “evidence of medical necessity,” they are looking for the specific cause-and-effect chain from diagnostic subtype → identified intervention → measurable target. “CBT for anxiety” does not show that chain; “exposure-based CBT for panic disorder, with interoceptive exposure protocol, targeting reduction in panic-attack frequency from [baseline] per week to ≤ 1 per week by session 12” does. The same logic applies to documentation in your progress notes; the mental health progress note templates and examples hub covers the note-side of the same defensibility discipline.
The third practical note is that a 12-session course is well within the modal coverage envelope for most US commercial payers and for the UK IAPT high-intensity step. If you anticipate going beyond 16 sessions, document the rationale in the plan (treatment-resistant presentation, comorbidity, partial response with continued functional impairment) so the extended course is defensible at re-authorization.
Worked example: GAD plan for an adult client
The following is a complete worked plan for a hypothetical adult client presenting with generalized anxiety. Details are illustrative.
The plan above is short, payer-defensible, and clear about the active ingredients. It would not pass for panic disorder, social anxiety, or specific phobia, which is the point: a good anxiety treatment plan is subtype-specific by design. For depression-side counterparts the depression treatment plan template follows the same structural pattern; for the underlying CBT protocol detail, the cbt treatment plan example carries a full session-by-session pacing walkthrough.
Less time on the template, more time on the work
A treatment plan is only useful if writing it does not eat the session before it. Most experienced clinicians build the plan with the client, on the page, in real time, so the document stays collaborative and the after-hours admin load stays modest. The plan you write with the client is also the plan they actually engage with between sessions, which matters for an exposure-heavy presentation where the between-session work is the active ingredient.
Emosapien’s Planning Agent drafts an anxiety treatment plan from your intake and updates it from your session notes, with the subtype-specific SMART objectives and modality framing already in place. The agent is not a generic AI scribe (it is the opposite of Heidi or DAX, which only document); it actively assists in the plan and progress-note workflow, with the modality-pack framing (CBT, exposure-based CBT, ACT) baked into the draft. The Engagement Agent then carries the between-session exposure work forward, so the client is doing the homework rather than the form. Sign up for free to try it on your next intake; no card required, and you keep editorial control over every plan and every note.