Generalised anxiety disorder and panic disorder in adults according to NICE: the treatment decision tree, step by step
What the British guideline recommends for generalised anxiety disorder and panic disorder in adults. One document, two stepped-care pathways: choose the pathway below, time runs up the trunk, choices branch off it, and the canopy carries what applies at every step. Hover a node: the exact wording appears on the right, and the tree stays whole.
Intended reader. Health professionals. This tool replaces neither clinical assessment, nor the guidelines in force, nor the summary of product characteristics. It contains no dosages.
Read this before the tree. You are about to see two different things.
The strength of the recommendation answers the question: should it be done, and how firmly. That is what colours the tree by default. It is read in the wording of the recommendation.
The certainty of the evidence answers a quite different question: how well do we know what this act produces. It is read in the studies, not in the recommendation.
On this guideline, the second view is almost entirely grey, and that is the message. CG113 publishes no level of evidence in the body of its recommendations: no GRADE, no rating, no reference to a study at the node considered.
The same document carries two pathways. Switch to compare.
This tree reads from the bottom up. The entry point sits at the very bottom, the outcomes and what to do at the top, and the step numbers climb with you.
Colour, the verb used
- Recommended. The text says “offer”.
- To consider. The text says “consider”.
- Advised against when the prohibition is outright, not routinely when the text attaches an exception.
- Situation. A moment in the pathway, not a recommendation.
Shape, the nature of the act
- Drug or formulation.
- Psychological intervention or self-help.
- Assessment or review.
- Timing or monitoring over time.
- Diamond. Clinical situation or fork.
In the tree, the icon takes the colour of its node. The nature of the act never governs the colour.
The line, what the source does not say
- Dashed. A deduced step, which no recommendation formulates. We draw it this way rather than pretend.
- Amber trunk. Time, from the entry of the pathway upwards.
- Twigs. They rise to the canopy, which depends on no step.
Entry of the pathway
Your pathway
- Click a card in the tree: the path leading to it from the entry point is listed here.
Nothing leaves your browser. What you type here is neither sent nor stored, on this site or anywhere else, and it is cleared when you close or reload the page. The printed sheet fits on one A4 page and keeps only the essentials of each step. Identify the patient by initials, never by name. Keeping the sheet in the medical record is your responsibility and falls under medical confidentiality.
The tree in full
What follows is not a copy of the drawing, it is its source. Each node is written once, here, with its coordinates, its parent, its colour and its badge carried in attributes. The script reads this text to draw the tree; it does not build it. There is therefore no second copy that could diverge.
Generalised anxiety
Repeated reassurance seeking, somatic symptoms, worry that cannot be controlled
Entry point. The guideline does not describe a patient complaining of anxiety, it describes one who consults for something else.
Entry point.
Next: step 2, Step 1 · Education about the disorder, and active monitoring.
What the recommendation says
“ consider with the person whether some of their symptoms may be due to GAD ”CG113 1.2.5 [2011]
What it does not say
No threshold, no duration, no number of consultations after which the question should arise. The trigger rests entirely on the clinician’s attention.
In France
The only French document in force on this subject is the ALD 23 list of acts and services for severe anxiety disorders, updated January 2025, verified in force on 28 August 2026. It states that it is not a good practice guideline and is not a decision aid. Its scope covers six disorders, including generalised anxiety and panic disorder, defined by reference to ICD-10 and DSM-IV. Not established: the status of the 2001 ANAES guideline on generalised anxiety disorder in adults, which PEB could not verify, the HAS search engine being inaccessible. Until that is settled, PEB does not write that no French guideline exists.
What it rests on
Nobody has graded it. No source grades this step.
The meta-analysis of 165 studies in the PEB corpus concerns mortality and suicide in anxiety disorders, not identification. It bears on why the step matters, not on how well it performs. The guideline itself publishes no level of evidence in the body of its text.
To see for yourself
Step 1 · Education about the disorder, and active monitoring
The step most often skipped in practice.
Comes after step 1, Repeated reassurance seeking, somatic symptoms, worry that cannot be controlled.
Next: step 3, Have the symptoms improved?
What the recommendation says
“ Following assessment and diagnosis of GAD: provide education about the nature of GAD and the options for treatment ”CG113 1.2.10 [2011]
What this means
Active monitoring. The text asks that symptoms and functioning be monitored, without saying how often, with what tool, or for how long. That is all the source means by the term.What it does not say
The duration of this step is never set. How long to monitor before concluding that education has not been enough is nowhere stated.
What it rests on
Nobody has graded it. No source grades this step.
No analysis in the PEB corpus addresses this step. And the guideline itself publishes no level of evidence in the body of its recommendations: no GRADE, no rating, no reference to a study at the node considered. Nobody has graded it.
Have the symptoms improved?
A deduced step. The guideline never asks this question, it presupposes it.
Comes after step 2, Step 1 · Education about the disorder, and active monitoring.
Next: step 4, Step 2 · Non-facilitated self-help, guided self-help, or psychoeducational group.
What the recommendation says
“ For people with GAD whose symptoms have not improved after education and active monitoring in step 1 ”CG113 1.2.12 [2011], subordonnée d’ouverture
What it does not say
No criterion of improvement. No scale is named to measure it, although GAD-2 is named for identification. The guideline can screen, not follow up.
What it rests on
Nobody has graded it. No source grades this step.
No analysis in the PEB corpus addresses this step. And the guideline itself publishes no level of evidence in the body of its recommendations: no GRADE, no rating, no reference to a study at the node considered. Nobody has graded it.
Step 2 · Non-facilitated self-help, guided self-help, or psychoeducational group
Three options of equal rank, and preference decides.
Comes after step 3, Have the symptoms improved?
Next: step 5, Inadequate response to step 2, or marked functional impairment.
What the recommendation says
“ offer 1 or more of the following as a first-line intervention, guided by the person’s preference ”CG113 1.2.12 [2011]
What this means
What the three options mean. Non-facilitated self-help is written or electronic material based on CBT principles, worked through alone and systematically over at least six weeks, with minimal therapist contact, for example an occasional call of no more than five minutes. Guided self-help uses the same material, supported by a trained practitioner who reviews progress, usually five to seven weekly or fortnightly sessions of twenty to thirty minutes, face to face or by telephone. A psychoeducational group is interactive, based on CBT principles, run by trained practitioners, one therapist to about twelve participants, usually six weekly sessions of two hours.What it does not say
Nothing ranks the three options against each other. No duration is set, and no criterion says when one has failed.
In France
The French list of reimbursed acts names no psychotherapy. The January 2025 update removed psychotherapy and the psychologist from the list, and a scripted count finds no occurrence of “TCC” or “thérapie cognitive” in the document. HAS gives the formal reason: such a list only covers acts reimbursed per act by the health insurance. The fact remains that the first-line intervention of the British guideline has no existence in the French document.
What it rests on
Nobody has graded it. No source grades this step.
No analysis in the PEB corpus addresses this step. And the guideline itself publishes no level of evidence in the body of its recommendations: no GRADE, no rating, no reference to a study at the node considered. Nobody has graded it.
Inadequate response to step 2, or marked functional impairment
The most important node of the tree, and the one the source leaves most open.
Comes after step 4, Step 2 · Non-facilitated self-help, guided self-help, or psychoeducational group.
Next: step 6, Step 3 · CBT, or applied relaxation; step 6a, Step 3 · An SSRI, sertraline first.
What the recommendation says
“ Base the choice of treatment on the person’s preference as there is no evidence that either mode of treatment is better ”CG113 1.2.17 [2011]
What it does not say
“Marked functional impairment” is defined nowhere in the document. Yet it is the criterion that moves a person straight from step 1 to step 3.
What it rests on
Nobody has graded it. No source grades this step.
The guideline itself writes that no evidence establishes the superiority of one mode of treatment over the other. It is the only place in the pathway where the text explicitly qualifies the state of the evidence, and it does so to say that it is lacking.
To see for yourself
Step 3 · CBT, or applied relaxation
Both are offered at the same rank.
Comes after step 5, Inadequate response to step 2, or marked functional impairment.
What the recommendation says
“ If a person with GAD chooses a high-intensity psychological intervention, offer either CBT or applied relaxation ”CG113 1.2.18 [2011]
What this means
What the two options mean. Both have exactly the same format in the text: twelve to fifteen weekly one-hour sessions, fewer if the person recovers sooner, more if clinically required, delivered by trained practitioners and based on the manuals used in the trials. Applied relaxation teaches progressive muscle relaxation, to be triggered in everyday anxiety-provoking situations. It is little used in France.What it does not say
No clinical feature guides the choice between the two. Applied relaxation, little used in France, is described only by reference to a manual.
In France
The French list of reimbursed acts names no psychotherapy. The January 2025 update removed psychotherapy and the psychologist from the list, and a scripted count finds no occurrence of “TCC” or “thérapie cognitive” in the document. HAS gives the formal reason: such a list only covers acts reimbursed per act by the health insurance. The fact remains that the first-line intervention of the British guideline has no existence in the French document.
What it rests on
Nobody has graded it. No source grades this step.
No analysis in the PEB corpus addresses this step. And the guideline itself publishes no level of evidence in the body of its recommendations: no GRADE, no rating, no reference to a study at the node considered. Nobody has graded it.
Step 3 · An SSRI, sertraline first
A first line chosen on an economic argument, and off-label.
Comes after step 5, Inadequate response to step 2, or marked functional impairment.
Next: step 7, Review, and continue for at least a year if the drug works; step 7a, Cannot tolerate SSRIs or SNRIs: consider pregabalin; step 7b, No benzodiazepine, except as a short-term measure during a crisis; step 7c, No antipsychotic in primary care.
What the recommendation says
“ Consider offering sertraline first because it is the most cost-effective drug ”CG113 1.2.23 [2011, amendé 2020]
What this means
The drugs the text names. Sertraline first. If it is ineffective, another SSRI or an SNRI, taking into account the withdrawal syndrome, more marked with paroxetine and venlafaxine, the side-effect profile, interactions, suicide risk and toxicity in overdose, greater with venlafaxine. No other drug is named for this step.What it does not say
The text itself adds that at publication, in January 2011, sertraline had no UK marketing authorisation for this indication, and flags off-label use for some SSRIs. No dose, no titration schedule. French authorisations do not follow from the British text and are not established in this tree.
In France
The French document names six selective serotonin reuptake inhibitors, plus venlafaxine, duloxetine and clomipramine, among twenty-three drugs in all. It says nothing about indications, treatment lines, or the order in which to try them: that is not its purpose. French marketing authorisations follow neither from this document nor from the British text, and PEB has not verified them drug by drug.
What it rests on
Nobody has graded it. No source grades this step.
No analysis in the PEB corpus addresses this step. And the guideline itself publishes no level of evidence in the body of its recommendations: no GRADE, no rating, no reference to a study at the node considered. Nobody has graded it.
Review, and continue for at least a year if the drug works
Comes after step 6a, Step 3 · An SSRI, sertraline first.
Next: step 8, No or partial response at step 3, and when to refer to step 4.
What the recommendation says
“ every 2 to 4 weeks during the first 3 months of treatment and every 3 months thereafter ”CG113 1.2.32 [2011]
“ advise the person to continue taking it for at least a year as the likelihood of relapse is high ”CG113 1.2.33 [2011]
What it does not say
Nothing is said about stopping: not how, not over what period, not what to watch for. The only duration set is a floor.
What it rests on
Nobody has graded it. No source grades this step.
No analysis in the PEB corpus addresses this step. And the guideline itself publishes no level of evidence in the body of its recommendations: no GRADE, no rating, no reference to a study at the node considered. Nobody has graded it.
Cannot tolerate SSRIs or SNRIs: consider pregabalin
Third line, and a controlled substance in the UK since 2019.
Comes after step 6a, Step 3 · An SSRI, sertraline first.
What the recommendation says
“ If the person cannot tolerate SSRIs or SNRIs, consider offering pregabalin ”CG113 1.2.25 [2011, amendé 2022]
What this means
Pregabalin is an antiepileptic also licensed in generalised anxiety. In the UK it has been a controlled substance since April 2019, which the text recalls by asking that a history of abuse be evaluated before prescribing.What it does not say
The text points to the UK agency’s alerts on misuse and on pregnancy, without saying how to assess a history of abuse before prescribing.
What it rests on
Nobody has graded it. No source grades this step.
No analysis in the PEB corpus addresses this step. And the guideline itself publishes no level of evidence in the body of its recommendations: no GRADE, no rating, no reference to a study at the node considered. Nobody has graded it.
No benzodiazepine, except as a short-term measure during a crisis
A prohibition with a named exception. Panic disorder, in the same document, has no such exception.
Comes after step 6a, Step 3 · An SSRI, sertraline first.
What the recommendation says
“ Do not offer a benzodiazepine for the treatment of GAD in primary or secondary care except as a short-term measure during crises ”CG113 1.2.26 [2011]
What it does not say
Neither the length of that short-term measure nor what counts as a crisis is defined. The text refers to the British national formulary.
In France
This is where the gap is sharpest. The French document names ten benzodiazepines among the drugs covered for severe anxiety disorders, where the British guideline sets them aside in both pathways. Again, a reimbursement list is not a recommendation to prescribe, and HAS says so. But the French practitioner looking for what the system covers finds ten benzodiazepines and no psychotherapy.
What it rests on
Nobody has graded it. No source grades this step.
No analysis in the PEB corpus addresses this step. And the guideline itself publishes no level of evidence in the body of its recommendations: no GRADE, no rating, no reference to a study at the node considered. Nobody has graded it.
No antipsychotic in primary care
An outright prohibition, but limited to one care setting.
Comes after step 6a, Step 3 · An SSRI, sertraline first.
What the recommendation says
“ Do not offer an antipsychotic for the treatment of GAD in primary care ”CG113 1.2.27 [2011, amendé 2018]
What it does not say
Nothing is said about secondary care, where the prohibition therefore does not apply. Nor does the text say what would justify an antipsychotic elsewhere.
In France
The French document names two antipsychotics, cyamemazine and sulpiride, among the drugs covered.
What it rests on
Nobody has graded it. No source grades this step.
No analysis in the PEB corpus addresses this step. And the guideline itself publishes no level of evidence in the body of its recommendations: no GRADE, no rating, no reference to a study at the node considered. Nobody has graded it.
No or partial response at step 3, and when to refer to step 4
Three distinct situations, and three different courses of action.
Comes after step 7, Review, and continue for at least a year if the drug works.
Next: step 9, Step 4 · Specialist assessment of needs and risks.
What the recommendation says
“ If a person’s GAD has not responded to drug treatment, offer either a high-intensity psychological intervention or an alternative drug treatment ”CG113 1.2.35 [2011]
“ Consider referral to step 4 if the person with GAD has severe anxiety with marked functional impairment in conjunction with: ”CG113 1.2.37 [2011]
What it does not say
PEB summarises the other two cases: if psychological treatment fails, move to a drug; if the drug response is partial, consider adding psychological treatment. No definition of partial response is given. PEB summarises the four accompanying criteria: risk of self-harm or suicide, significant comorbidity, self-neglect, or inadequate response to step 3. Severity itself remains undefined.
What it rests on
Nobody has graded it. No source grades this step.
No analysis in the PEB corpus addresses this step. And the guideline itself publishes no level of evidence in the body of its recommendations: no GRADE, no rating, no reference to a study at the node considered. Nobody has graded it.
To see for yourself
Step 4 · Specialist assessment of needs and risks
Comes after step 8, No or partial response at step 3, and when to refer to step 4.
Next: step 10, Combinations, knowing the evidence is lacking.
What the recommendation says
“ Offer the person with GAD a specialist assessment of needs and risks ”CG113 1.2.38 [2011]
What it does not say
No tool, no framework, no waiting-time standard. Step 4 usually means community mental health teams, which the text notes in brackets without making it a recommendation.
What it rests on
Nobody has graded it. No source grades this step.
No analysis in the PEB corpus addresses this step. And the guideline itself publishes no level of evidence in the body of its recommendations: no GRADE, no rating, no reference to a study at the node considered. Nobody has graded it.
Combinations, knowing the evidence is lacking
A recommendation that itself states what it does not rest on.
Comes after step 9, Step 4 · Specialist assessment of needs and risks.
What the recommendation says
“ evidence for the effectiveness of combination treatments is lacking ”CG113 1.2.41 [2011]
What it does not say
No combination is named, no order is proposed, and the text adds that side effects and interactions are more likely when antidepressants are combined or augmented.
What it rests on
Nobody has graded it. No source grades this step.
The guideline writes that evidence for the effectiveness of combination treatments is lacking, and still proposes to consider them. Strength of recommendation and certainty of evidence openly diverge here, which is rare and honest.
What applies at every step
Recommendations that do not depend on the position in the pathway. Canopy entries are summaries of the recommendation, not word-for-word quotations.
GAD-2
Identify with GAD-2
PEB summary
Ask about anxiety and about the ability to stop or control worrying, using the two-item scale. The text says to think of it particularly with a past anxiety disorder, somatic symptoms, or after a recent traumatic event.
Where it is written
CG113 1.2.2 [2011]
What this means
GAD-2 is these two questions. Over the last two weeks, how often have you been bothered by the following problems: “feeling nervous, anxious or on edge”, then “not being able to stop or control worrying”. Each item is scored 0 to 3, not at all, several days, more than half the days, nearly every day. The score therefore runs from 0 to 6. These are the first two items of the GAD-7.Scale developed by Robert L. Spitzer, Kurt Kroenke, Janet B. W. Williams and Bernd Löwe. The notice carried on the questionnaire reads: “Copyright Pfizer Inc. No permission required to reproduce, translate, display, or distribute.”
What it does not say
NICE says to consider using GAD-2, and gives no threshold. Nothing in the guideline says at what score assessment should be taken further. GAD-7 does not appear once in the whole document.
To see for yourself
Disability and cognitive impairment
Learning disability or cognitive impairment
PEB summary
Mild: the same interventions as for everyone else, adjusting delivery or duration if needed. Moderate to severe: consider consulting a relevant specialist before assessing or offering an intervention.
Where it is written
CG113 1.1.1, 1.1.2 [2011]
Over-the-counter
Discuss over-the-counter use
PEB summary
Raise products taken without prescription, explain possible interactions with other treatments, and say that nothing establishes their safe use.
Where it is written
CG113 1.2.11 [2011]
Under 30
Under 30 on an SSRI or SNRI
PEB summary
Warn of the increased risk of suicidal thinking and self-harm in a minority of people under 30, see the person within a week of first prescribing, then monitor that risk weekly for the first month.
Where it is written
CG113 1.2.30 [2011]
Bleeding risk
Bleeding on SSRIs
PEB summary
Take into account the increased bleeding risk on SSRIs, particularly in older people or when combined with a non-steroidal anti-inflammatory drug or aspirin. Consider a gastroprotective drug in that case.
Where it is written
CG113 1.2.29 [2011]
Language
In the person’s language
PEB summary
Consider providing all interventions in the person’s preferred language where possible.
Where it is written
CG113 1.2.22 [2011]
Shared decision
Information and shared decision making
PEB summary
Have evidence-based information about treatments available, and discuss the possible options, so that shared decision making is possible.
Where it is written
CG113 1.2.44 [2004]
Panic disorder
Step 1 · Panic attack in the emergency department
Entry point, and the most counter-intuitive recommendation in the document.
Entry point.
Next: step 2, Step 1 · Diagnosis is made in the interview, without an instrument.
What the recommendation says
“ not usually be admitted to a medical or psychiatric bed ”CG113 1.3.6 [2004]
What it does not say
PEB summarises the rest: ask whether treatment is already under way, do only the investigations needed to exclude an acute physical problem, refer back to primary care with written information. No investigation is named, no referral delay is set.
In France
The only French document in force on this subject is the ALD 23 list of acts and services for severe anxiety disorders, updated January 2025, verified in force on 28 August 2026. It states that it is not a good practice guideline and is not a decision aid. Its scope covers six disorders, including generalised anxiety and panic disorder, defined by reference to ICD-10 and DSM-IV. Not established: the status of the 2001 ANAES guideline on generalised anxiety disorder in adults, which PEB could not verify, the HAS search engine being inaccessible. Until that is settled, PEB does not write that no French guideline exists.
What it rests on
Nobody has graded it. No source grades this step.
The meta-analysis of 165 studies in the PEB corpus concerns mortality and suicide in anxiety disorders, not identification. It bears on why the step matters, not on how well it performs. The guideline itself publishes no level of evidence in the body of its text.
To see for yourself
Step 1 · Diagnosis is made in the interview, without an instrument
To be compared with the GAD pathway, which names GAD-2.
Comes after step 1, Step 1 · Panic attack in the emergency department.
Next: step 3, From step 2 to step 3: severity decides.
What the recommendation says
“ There is insufficient evidence on which to recommend a well-validated, self-reporting screening instrument to use in the diagnostic process ”CG113 1.3.3 [2004]
What this means
What this means in the consultation. No questionnaire is recommended to make the diagnosis. The text asks that consultation skills be relied upon to elicit the necessary information.What it does not say
The same document names GAD-2 for identifying GAD. Here it concludes that evidence is insufficient. The two recommendations are not the same age, 2011 against 2004, and the text never brings the two positions together.
What it rests on
Nobody has graded it. No source grades this step.
No analysis in the PEB corpus addresses this step. And the guideline itself publishes no level of evidence in the body of its recommendations: no GRADE, no rating, no reference to a study at the node considered. Nobody has graded it.
From step 2 to step 3: severity decides
Moving from one step to the next hangs on a single word, and that word is defined nowhere.
Comes after step 2, Step 1 · Diagnosis is made in the interview, without an instrument.
Next: step 4, Step 2 · Mild to moderate: self-help, facilitated or not; step 4c, Step 3 · Moderate to severe: CBT, or an antidepressant; step 4a, Step 3 · No benzodiazepine; step 4b, Step 3 · No sedating antihistamine, no antipsychotic.
What the recommendation says
“ For people with moderate to severe panic disorder (with or without agoraphobia), consider referral for: ”CG113 1.3.12 [2011]
What it does not say
No criterion separates mild to moderate from moderate to severe. Not attack frequency, not impairment, not a scale. Yet it is the only fork in the pathway.
What it rests on
Nobody has graded it. No source grades this step.
No analysis in the PEB corpus addresses this step. And the guideline itself publishes no level of evidence in the body of its recommendations: no GRADE, no rating, no reference to a study at the node considered. Nobody has graded it.
To see for yourself
- Fear disorders and generalised anxiety, genetic overlap
- The step number, in a pill at the corner of each card. It gives the rank along the path. Cards of the same rank on another branch are marked a, b, c.
- Pale blue glow. On hover or click, the steps that may follow light up in pale blue, and the path already travelled down to the entry lights up in amber.
Step 2 · Mild to moderate: self-help, facilitated or not
Comes after step 3, From step 2 to step 3: severity decides.
What the recommendation says
“ offer or refer for 1 of the following low-intensity interventions ”CG113 1.3.9 [2011]
What this means
And here the source defines nothing. The panic disorder section names facilitated and non-facilitated self-help without ever saying what they are. The only definitions in the document sit in the GAD section, and they do not use the same word: guided on one side, facilitated on the other. For want of better, PEB carries over the definitions from the other section: CBT-based material, worked through alone with minimal contact, or supported by a trained practitioner over five to seven sessions of twenty to thirty minutes.What it does not say
No duration, no failure criterion, and nothing on what to do when these interventions are not enough, apart from the node on referral after two interventions.
In France
The French list of reimbursed acts names no psychotherapy. The January 2025 update removed psychotherapy and the psychologist from the list, and a scripted count finds no occurrence of “TCC” or “thérapie cognitive” in the document. HAS gives the formal reason: such a list only covers acts reimbursed per act by the health insurance. The fact remains that the first-line intervention of the British guideline has no existence in the French document.
What it rests on
Nobody has graded it. No source grades this step.
No analysis in the PEB corpus addresses this step. And the guideline itself publishes no level of evidence in the body of its recommendations: no GRADE, no rating, no reference to a study at the node considered. Nobody has graded it.
Step 3 · Moderate to severe: CBT, or an antidepressant
The antidepressant does not sit at the same rank as CBT: three conditions trigger it.
Comes after step 3, From step 2 to step 3: severity decides.
Next: step 5, Step 3 · CBT, optimal duration 7 to 14 hours; step 5a, Step 3 · An SSRI licensed for panic disorder.
What the recommendation says
“ CBT or an antidepressant if the disorder is long-standing or the person has not benefitted from or has declined psychological intervention ”CG113 1.3.12 [2011]
What this means
Which antidepressant. The text points to SSRIs licensed for panic disorder, set out in the next node. It names no drug at this point.What it does not say
“Long-standing” is not quantified. Nor does the text say how long a psychological intervention must have run before concluding it has not helped.
What it rests on
Nobody has graded it. No source grades this step.
No analysis in the PEB corpus addresses this step. And the guideline itself publishes no level of evidence in the body of its recommendations: no GRADE, no rating, no reference to a study at the node considered. Nobody has graded it.
Step 3 · CBT, optimal duration 7 to 14 hours
One of the few numbers in the document.
Comes after step 4c, Step 3 · Moderate to severe: CBT, or an antidepressant.
What the recommendation says
“ CBT in the optimal range of duration (7 to 14 hours in total) should be offered. ”CG113 1.3.15 [2004]
What this means
What this CBT covers. Usually weekly sessions of one to two hours, completed within four months. A briefer form exists, around seven hours, provided it is supported by structured self-help materials.What it does not say
PEB summarises what follows: usually weekly sessions of one to two hours, completed within four months. No study is cited in support of that range in the body of the recommendations.
In France
The French list of reimbursed acts names no psychotherapy. The January 2025 update removed psychotherapy and the psychologist from the list, and a scripted count finds no occurrence of “TCC” or “thérapie cognitive” in the document. HAS gives the formal reason: such a list only covers acts reimbursed per act by the health insurance. The fact remains that the first-line intervention of the British guideline has no existence in the French document.
What it rests on
Nobody has graded it. No source grades this step.
No analysis in the PEB corpus addresses this step. And the guideline itself publishes no level of evidence in the body of its recommendations: no GRADE, no rating, no reference to a study at the node considered. Nobody has graded it.
Step 3 · An SSRI licensed for panic disorder
Unlike the GAD pathway, no molecule is named as first line.
Comes after step 4c, Step 3 · Moderate to severe: CBT, or an antidepressant.
Next: step 6, Step 3 · No improvement at 12 weeks: imipramine or clomipramine; step 6a, Follow-up · Assessment at twelve weeks: continue, or change.
What the recommendation says
“ Unless otherwise indicated, an SSRI licensed for panic disorder should be offered. ”CG113 1.3.25 [2004]
What this means
The drugs the text names. A June 2020 amendment lists the antidepressants licensed in the UK for panic disorder: escitalopram, sertraline, citalopram, paroxetine and venlafaxine. The text adds that the classes with an evidence base are SSRIs, SNRIs and tricyclics, and that antidepressants are the only pharmacological intervention in longer-term management.What it does not say
That list is of UK marketing authorisations at the date of the amendment. French indications do not follow from it and are not established in this tree. No dose, no titration schedule, no order of preference among the five.
In France
The French document names six selective serotonin reuptake inhibitors, plus venlafaxine, duloxetine and clomipramine, among twenty-three drugs in all. It says nothing about indications, treatment lines, or the order in which to try them: that is not its purpose. French marketing authorisations follow neither from this document nor from the British text, and PEB has not verified them drug by drug.
What it rests on
Nobody has graded it. No source grades this step.
No analysis in the PEB corpus addresses this step. And the guideline itself publishes no level of evidence in the body of its recommendations: no GRADE, no rating, no reference to a study at the node considered. Nobody has graded it.
Step 3 · No improvement at 12 weeks: imipramine or clomipramine
Off-label, and the text flags it.
Comes after step 5a, Step 3 · An SSRI licensed for panic disorder.
What the recommendation says
“ If an SSRI is not suitable or there is no improvement after a 12-week course and if a further medication is appropriate, imipramine or clomipramine may be considered. ”CG113 1.3.26 [2004, amendé 2020]
What it does not say
No dose, no titration schedule, and no specific cardiac precaution is recalled at this node, although these are tricyclics.
What it rests on
Nobody has graded it. No source grades this step.
No analysis in the PEB corpus addresses this step. And the guideline itself publishes no level of evidence in the body of its recommendations: no GRADE, no rating, no reference to a study at the node considered. Nobody has graded it.
Follow-up · Assessment at twelve weeks: continue, or change
A follow-up recommendation, filed in the document under the step 5 heading.
Comes after step 5a, Step 3 · An SSRI licensed for panic disorder.
Next: step 7, Step 4 · After two interventions, refer to specialist services.
What the recommendation says
“ At the end of 12 weeks, an assessment of the effectiveness of the treatment should be made, and a decision made as to whether to continue or consider an alternative intervention. ”CG113 1.3.42 [2004]
What it does not say
PEB summarises what follows: beyond twelve weeks, review the person every eight to twelve weeks. No criterion of effectiveness is given, and no scale is named to establish it. Note: this recommendation belongs to a monitoring and follow-up subsection that the document files under the step 5 heading, although it concerns every step.
What it rests on
Nobody has graded it. No source grades this step.
No analysis in the PEB corpus addresses this step. And the guideline itself publishes no level of evidence in the body of its recommendations: no GRADE, no rating, no reference to a study at the node considered. Nobody has graded it.
Step 3 · No benzodiazepine
A prohibition with no exception, justified by long-term outcome. The GAD pathway, by contrast, allows one.
Comes after step 3, From step 2 to step 3: severity decides.
What the recommendation says
“ Benzodiazepines are associated with a less good outcome in the long term and should not be prescribed for the treatment of individuals with panic disorder. ”CG113 1.3.20 [2004]
What it does not say
No figure accompanies the claim of a less good outcome. The text points to no study at this place, and provides nothing for a person already taking a benzodiazepine.
In France
This is where the gap is sharpest. The French document names ten benzodiazepines among the drugs covered for severe anxiety disorders, where the British guideline sets them aside in both pathways. Again, a reimbursement list is not a recommendation to prescribe, and HAS says so. But the French practitioner looking for what the system covers finds ten benzodiazepines and no psychotherapy.
What it rests on
Nobody has graded it. No source grades this step.
The claim of a less good long-term outcome on benzodiazepines comes with no figure and no reference to a study at that point in the text. It carries the 2004 label. No analysis in the PEB corpus addresses this question.
Step 3 · No sedating antihistamine, no antipsychotic
Comes after step 3, From step 2 to step 3: severity decides.
What the recommendation says
“ Sedating antihistamines or antipsychotics should not be prescribed for the treatment of panic disorder. ”CG113 1.3.21 [2004]
What it does not say
No restriction by care setting, unlike the antipsychotic prohibition in the GAD pathway, which covers primary care only.
In France
The French document names two antipsychotics, cyamemazine and sulpiride, and lists hydroxyzine among the other anxiolytics covered. Hydroxyzine is a sedating antihistamine, precisely the class the British guideline sets aside in panic disorder.
What it rests on
Nobody has graded it. No source grades this step.
No analysis in the PEB corpus addresses this step. And the guideline itself publishes no level of evidence in the body of its recommendations: no GRADE, no rating, no reference to a study at the node considered. Nobody has graded it.
Step 4 · After two interventions, refer to specialist services
The only referral criterion in the pathway, and it is a count.
Comes after step 6a, Follow-up · Assessment at twelve weeks: continue, or change.
Next: step 8, Step 5 · Thorough reassessment in specialist services.
What the recommendation says
“ In most instances, if there have been 2 interventions provided and the person still has significant symptoms, then referral to specialist mental health services should be offered. ”CG113 1.3.35 [2004]
What it does not say
The text counts interventions without saying what makes an intervention count. No minimum duration, no adherence, no dose reached. Two poorly conducted trials therefore count as much as two complete ones.
What it rests on
Nobody has graded it. No source grades this step.
No analysis in the PEB corpus addresses this step. And the guideline itself publishes no level of evidence in the body of its recommendations: no GRADE, no rating, no reference to a study at the node considered. Nobody has graded it.
Step 5 · Thorough reassessment in specialist services
Comes after step 7, Step 4 · After two interventions, refer to specialist services.
What the recommendation says
“ Specialist mental health services should conduct a thorough, holistic reassessment of the individual, their environment and social circumstances. ”CG113 1.3.36 [2004, amendé 2020]
What it does not say
PEB summarises what the reassessment should cover: previous treatments and adherence, substance use including nicotine, alcohol and caffeine, comorbidities, day-to-day functioning, social networks, chronic stressors, the role of agoraphobic avoidance. No tool is named.
What it rests on
Nobody has graded it. No source grades this step.
No analysis in the PEB corpus addresses this step. And the guideline itself publishes no level of evidence in the body of its recommendations: no GRADE, no rating, no reference to a study at the node considered. Nobody has graded it.
What applies at every step
Recommendations that do not depend on the position in the pathway. Canopy entries are summaries of the recommendation, not word-for-word quotations.
Support groups
Offer information on support groups
PEB summary
Offer information about support groups where they exist. The text notes they may take the form of meetings, telephone groups sometimes based on CBT principles, or additional information.
Where it is written
CG113 1.3.10 [2004]
Physical activity
Discuss exercise
PEB summary
Discuss the benefits of exercise as part of good general health with all people with panic disorder, as appropriate.
Where it is written
CG113 1.3.11 [2004]
Following up self-help
Self-help is not left without contact
PEB summary
Offer people receiving self-help interventions contact with primary healthcare professionals, so progress can be monitored and an alternative intervention considered if needed.
Where it is written
CG113 1.3.44 [2004]
Measuring outcome
Short questionnaires to follow up
PEB summary
Use short self-completed questionnaires wherever possible, for example the panic subscale of the agoraphobic mobility inventory. Note: no instrument is recommended for diagnosis, yet one is named for follow-up.
Where it is written
CG113 1.3.45 [2004]
Progress under CBT
Assessing progress under CBT
PEB summary
Each practice should have a process to assess the progress of a person undergoing CBT. The nature of that process is left to a case-by-case basis, which the text states plainly.
Where it is written
CG113 1.3.40 [2004]
Stopping antidepressants
Do not stop abruptly
PEB summary
Stopping an antidepressant abruptly can cause discontinuation or withdrawal symptoms. To limit that risk, the dose should be reduced gradually over an extended period.
Where it is written
CG113 1.3.30 [2004]
What this tree does not cover. CG113 addresses only generalised anxiety disorder and panic disorder in adults. Social anxiety, specific phobia, obsessive-compulsive disorder and post-traumatic stress disorder fall under other British guidelines, which are not in this tree. The French ALD 23 document covers all six of those disorders at once: the two scopes do not overlap.
Our quotation regime
This tree quotes one sentence per node, the one that decides, with its recommendation number and date label. It reproduces no bullet list from the source text: those are summarised and announced as summaries. Twenty-nine short quotations cover forty of the ninety-five recommendations in the document. A reader who wants to check word for word follows the link to NICE. This choice follows from the fact that the NICE open content licence covers the United Kingdom only.
How this tree was built
The 29 quotations come from the NICE text itself, taken from the source and never from memory, with their recommendation number and date label. The figures on this page were recounted one by one rather than copied.
Two points could not be settled, the status of the 2001 French guideline on generalised anxiety disorder and the French marketing authorisations drug by drug. We therefore do not settle them here. The detail of these checks is set out on the page Comment PEB vérifie ce qu’elle publie.
Sources
The full critical appraisal of this guideline, with its AGREE II methodological assessment, is published separately: PEB appraisal of NICE CG113.
- National Institute for Health and Care Excellence. Generalised anxiety disorder and panic disorder in adults: management. NICE Clinical Guideline No. 113. Published 26 January 2011, last updated 15 June 2020, checked by NICE in March 2020. Recommendations chapter consulted on 28 August 2026. It replaces guideline CG22.
- Haute Autorité de santé. ALD no. 23, Severe anxiety disorders. Acts and services for long-term conditions, January 2025 update. Adopted by the Board on 23 January 2025, published online on 31 January 2025. HAS page consulted on 28 August 2026.
- Spitzer RL, Kroenke K, Williams JBW, Löwe B. GAD-7 scale, whose first two items constitute GAD-2. The questionnaire carries the notice “Copyright Pfizer Inc. No permission required to reproduce, translate, display, or distribute”.
And on the other side of the Channel
This tree sets out what the British guideline says. France, for its part, has no established good practice guideline on this subject, but a coverage document listing twenty-three reimbursed drugs under the long-term condition scheme, including ten benzodiazepines, and no psychotherapy. The two are read one after the other, and the gap between them is the subject.
Open the ALD 23 page, anxiety disordersPsychiatry Evidence Base, evidence-based psychiatry, explained with rigour.