Cognitive therapy for panic

1. Introduction

Version followed: the main protocol is Clark and Salkovskis's Cognitive Therapy for Panic Disorder manual (Oxford, IAPT edition, 2009), with the brief five-session variant (Clark et al., 1999) noted where it differs. Panic control treatment (PCT) is given as a second variant. That variant follows the session plan of Mastery of Your Anxiety and Panic (MAP), 5th edition (Barlow & Craske, 2022), and the inhibitory-learning exposure rules of Craske et al. (2014).

What it is. Cognitive therapy (CT) for panic disorder is a structured, time-limited individual therapy built on Clark's (1986) cognitive model. According to the model, panic attacks come from catastrophic misinterpretation of body sensations. Palpitations become "I'm having a heart attack", dizziness becomes "I'm going to faint", and unreality becomes "I'm going mad". The Oxford group summarises the core feature as "a tendency to misinterpret harmless body sensations" as a sign of disaster [source: Oxford Neuroscience page].

How change happens. Three processes keep the misinterpretations going:

• Safety behaviours. Sitting down, holding on, tensing the legs, controlling breathing and escaping all stop the person from learning that the catastrophe would not have happened anyway.
• Avoidance.
• Heightened attention to the body. This makes the person notice more sensations.

Therapy changes the beliefs directly. It reproduces the sensations to show their harmless causes, and it runs behavioural experiments with safety behaviours dropped so the predicted catastrophe can be tested [source: Clark & Salkovskis 2009 manual; Salkovskis et al. 1999].

PCT (Barlow, Craske and colleagues) has the same target, fear of body sensations. It puts more weight on repeated interoceptive and situational exposure, together with cognitive skills. The 5th-edition MAP is built around inhibitory learning. The aim is to build tolerance of discomfort and to violate feared expectancies, not to reduce fear within a session [source: Psychology Tools MAP 5th ed. page; Craske et al. 2014].

2. Framework

Cognitive therapy (Clark & Salkovskis)

Format. CT is individual and outpatient, with weekly sessions. The manual gives 5 to 15 sessions as typical. The 1999 trial describes standard CT as 12 to 15 one-hour sessions [source: CT manual 2009; Clark et al. 1999 abstract]. Each session has the same structure:

1. Review the questionnaires and the panic diary.
2. Agree an agenda.
3. Review homework.
4. Work on one or two main topics, asking for feedback often.
5. Set new homework.

Patients record the sessions and listen back as homework [source: CT manual 2009]. Near the end, sessions are spaced further apart to build self-reliance [source: CT manual 2009].

Brief CT. Brief CT kept outcomes equivalent to full CT. It used 6.5 hours of therapist time in total, including booster sessions, and relied heavily on self-study modules between sessions [source: Clark et al. 1999 abstract]. The exact split of sessions and boosters, and the content of the modules, are [uncertain]. The adolescent brief-CT manual on OxCADAT is not drawn on here.

NICE CG113 (UK benchmark). CBT for panic should total 7 to 14 hours, usually in weekly sessions of 1 to 2 hours, finished within 4 months. Briefer CBT (about 7 hours) should be paired with structured self-help materials. Intensive CBT over a short period suits a few people. Practitioners should be trained, supervised and protocol-faithful [source: NICE CG113 recs 1.3.12–1.3.19].

Phases. The manual sets out a sequence of tasks, not numbered sessions. The phase labels below are my own grouping of that sequence [synthesis].

Panic control treatment / MAP (Barlow & Craske, 5th ed.)

Format. MAP is a 12-session structured format that can be adjusted to progress and severity. A brief six-session version for primary care is included as chapter 16 [source: Psychology Tools MAP page].

Content order. Chapter order, which is roughly the session order:

1. Programme introduction
2. Recording panic and anxiety
3. Negative cycles of panic and agoraphobia
4. "Panic attacks are not harmful"
5. Building an agoraphobia hierarchy
6. Breathing skills
7. Thinking skills
8. Facing physical symptoms (interoceptive exposure)
9. Facing agoraphobia situations
10. Involving others
11. Medications
12. Accomplishments, maintenance and relapse prevention

The exact session in which each chapter is delivered was not visible [uncertain].

Group example. A published group protocol taught education, cognitive reappraisal, interoceptive exposure, in vivo exposure and breathing control over 12 weekly sessions with groups of 5 to 7. Interoceptive assessment came at session 5, in-session interoceptive exposure the next week, and home practice in the weeks after that [source: Schmidt & Trakowski 2004].

Moving between phases. In the older habituation-based rule, an interoceptive exercise is finished when the first trial of a session produces no fear [source: Schmidt & Trakowski 2004]. Under the 5th-edition inhibitory-learning approach, a trial ends when the feared expectancy has been tested, and fear is allowed to stay high [source: Craske et al. 2014]. Use one rule consistently.

3. Guiding principles

From the CT manual [source: CT manual 2009 unless marked]:

• Collaborative empiricism. Treat the patient's catastrophic thoughts as hypotheses to test together, not as errors to correct by lecture. In practice: ask "how could we find out?" instead of telling.
• Use the patient's own model. Build the vicious circle from their words and their latest attack. In practice: write what they actually felt and thought, not textbook terms.
• Change the meaning, not the sensation. If relief depends on controlling sensations, panic returns once the control strategy is dropped. In practice: do not teach relaxation or breathing as a way to stop attacks. Breathing retraining was dropped from later versions because patients used it as a safety behaviour.
• Find and drop safety behaviours. They protect the catastrophic belief from disconfirmation [also: Salkovskis et al. 1999]. In practice: for every feared outcome, ask what the person does to prevent it, then plan to drop it.
• Make predictions before every experiment. In practice: write down exactly what the patient expects, then compare it with what happened.
• Record outcomes, not anxiety. The experiment record sheet deliberately has no anxiety rating. In practice: ask "what did you learn?", not "how anxious were you?".
• Rate belief 0–100 before and after each intervention, and during attacks. In practice: use the ratings to steer the session and to show change.
• Tailor education to the specific belief. In practice: give one or two targeted facts that answer this person's fear. Do not give a general lecture.
• Use old contradictory evidence. In practice: ask about times the catastrophe should have happened and did not, for example when panic stopped on distraction.
• Practise early experiments with the therapist present. In practice: do the first overbreathing or exercise test in the room, then assign it as homework.
• Keep the session structured with an agenda, homework review, feedback and new homework.
• Build towards self-reliance. In practice: use audiotapes and a written blueprint, and space out the final sessions.
• Medication stays steady until the patient is confident without it, then is withdrawn gradually. In practice: if medication blocks all attacks, the patient may need to withdraw it to learn that panic is survivable. Coordinate any change with the prescriber.
• Check physical health first. Some experiments need changes for cardiac disease, pregnancy, epilepsy or asthma. In practice: get medical clearance before exercise or overbreathing tasks.

From PCT/MAP and inhibitory learning [source: Psychology Tools MAP page; Craske et al. 2014]:

• Tolerate, don't reduce, fear. In practice: success means the client stayed with the sensation and the prediction failed, not that the client calmed down.
• Violate expectancies. In practice: before each trial, have the client name the feared outcome (the harm, not "feeling anxious") and rate how likely it is. End the trial once it has been tested.
• Hold cognitive reappraisal until after exposure. In practice: don't talk the client out of the fear before the trial, so the trial can disconfirm it.
• Remove safety signals. These include companions, carried medication, phones and water bottles. In practice: phase them out gradually if needed, all at once if the client agrees.
• Vary the practice. In practice: change the order, intensity, length and setting of exposures. Do not climb the hierarchy strictly in order.
• Practise in many contexts. In practice: include alone, unfamiliar places and different times of day.
• Distinguish sensation from anxiety. In practice: rate intensity and fear separately [source: Schmidt & Trakowski 2004].

Exposure consent and safety [synthesis, drawing on the manual's physical-illness caution and Schmidt & Trakowski's spotter and adaptation advice]:

• Explain the rationale.
• Explain that discomfort will rise for a short time and that sensations may linger after the exercise.
• Get explicit agreement to each experiment.
• Agree that the client can stop.
• Use a spotter for spinning.
• Adapt or omit tasks for medical conditions.

4. Intervention inventory

Part A covers the cognitive therapy interventions (Clark & Salkovskis). Unless marked otherwise, the source is the CT manual (2009) [source: CT manual 2009]. All example language is mine and labelled (illustrative).

A1. Assessment of a recent severe attack

• Purpose: Get the raw material for the model: trigger, sensations, catastrophic thoughts and images, and safety behaviours. Use it in session 1 and again whenever a new type of attack appears.
• Steps:
  1. Confirm panic disorder: repeated attacks, some of them unexpected, plus fear of the attacks.
  2. Rule out specific phobia, social anxiety, substance-induced anxiety, OCD and PTSD as the main problem.
  3. Pick one recent, severe attack and go through it slowly:
     • the situation;
     • the first thing noticed;
     • the sensations at the worst point;
     • the thoughts and any mental images at the worst point;
     • what the patient did to stop the catastrophe or to cope.
  4. Ask what the worst thing was that could have happened. If the patient left the situation, ask what would have happened if they had stayed.
  5. Briefly check other attacks for different sensations or fears.
  6. Map triggers, avoided situations and activities, and body checking between attacks.
  7. Ask how significant others understand the attacks and how they respond.
  8. Ask about medication, alcohol and drugs, earlier treatment, onset and course, and comorbidity.
  9. If depression is present, ask a "magic wand" question: if the panic vanished, would the low mood remain? This helps decide whether the depression is secondary.
• Example language (illustrative): "Let's slow the last bad one right down. At the very worst moment, what did you think was going to happen? Did a picture come into your mind?"
• Watch for:
  • The patient says "I just felt awful." Keep probing for the feared outcome, such as dying, collapsing, losing control or embarrassment.
  • Images, for example of being carried off on a stretcher. They often carry the fear.

A2. Standard measures and weekly monitoring

• Purpose: Track symptoms and beliefs across therapy.
• Steps:
  1. Before treatment, give:
     • the Mobility Inventory (avoidance);
     • the Agoraphobic Cognitions Questionnaire, with a belief rating added for each thought;
     • the Safety Seeking Behaviours Questionnaire.
  2. Each week, use a panic weekly summary scale plus the ACQ, BAI and BDI.
  3. Repeat the Mobility Inventory and the SSBQ mid-treatment and at the end.
  4. Review the scores at the start of each session.
• Example language (illustrative): "Your belief that you'll faint has gone from 80 to 40 on the form. What do you think brought that down?"
• Watch for: Scores that don't move, and items that stay high. A belief that stays high marks the next experiment.

(Do not reproduce the instruments themselves. They are published separately.)

A3. Panic diary

• Purpose: Catch attacks as they happen and build the skill of answering panic thoughts.
• Steps:
  1. At first, the patient records each attack's trigger, sensations and thoughts.
  2. Once they can answer catastrophic thoughts in session, add a column for their response to the panic thought and a belief rating.
  3. Review the diary at the top of each session.
• Example language (illustrative): "Write it down as close to the attack as you can: what you noticed first, what you thought it meant, and how much you believed it."
• Watch for:
  • Diaries filled in at the end of the week, which lose the thoughts.
  • Diaries used as a form of body monitoring. If that happens, shorten the diary.

A4. Deriving the individual vicious circle

• Purpose: Give the patient a shared explanation that predicts which interventions will help.
• Steps:
  1. On a whiteboard or paper, start with the trigger. This may be external, or internal such as a sensation or thought.
  2. Draw the arrows round the circle: perceived threat, then apprehension, then the patient's own sensations, then their own catastrophic interpretation, then back to more apprehension.
  3. Add the safety behaviours, with an arrow back to the catastrophic belief: "stops me finding out".
  4. Use only the patient's words.
  5. Ask whether it fits, and change it until it does.
• Example language (illustrative): "So your heart speeds up, you think 'this is a heart attack', you get more frightened, and that speeds your heart up more. Does that loop fit what happens?"
• Watch for: A patient who agrees too easily. Check that they can explain the circle back in their own words.

A5. Getting started: why not control the sensations

• Purpose: Agree the target of treatment, which is the beliefs and not the symptoms.
• Steps:
  1. Ask where the circle could be broken.
  2. Discuss what happens if the plan is to control sensations: when the technique is dropped, the beliefs are unchanged and panic returns.
  3. Agree to work on the meaning of the sensations.
• Example language (illustrative): "If we taught you a trick to stop the sensations, what would you believe about them the day the trick didn't work?"
• Watch for: A patient who wants a coping technique. Acknowledge the wish and explain the trade-off.

A6. Belief ratings (0–100)

• Purpose: Measure conviction and show the effect of each intervention.
• Steps:
  1. Ask for a rating of the catastrophic belief:
     • now;
     • during an attack;
     • before and after each discussion or experiment.
  2. Rate the alternative explanation as well.
• Example language (illustrative): "Right now, how much do you believe the dizziness means you'll faint, from 0 to 100? And at the worst point last Tuesday?"
• Watch for: A low rating in the room and a high one during an attack. The gap tells you that in-situation experiments are still needed.

A7. Examining the evidence for beliefs

• Purpose: Find out what observations hold up the catastrophic belief and re-examine them.
• Steps:
  1. Ask what makes the patient think the catastrophe could happen.
  2. List each piece of evidence.
  3. For each one, explore other explanations together. The manual's example: apparent "deaths from fear" in a family history turned out to be explained by carbon monoxide poisoning.
  4. Re-rate the belief.
• Example language (illustrative): "What have you seen or heard that makes a heart attack seem possible? ... Is there any other way to understand that?"
• Watch for: Evidence that rests on misremembered or misheard medical information. Check it carefully, and with the GP if needed.

A8. Targeted education

• Purpose: Correct the specific misunderstanding behind each fear. Give only what that fear needs.
• Steps:
  1. Pick the patient's belief.
  2. Give the relevant facts. Manual examples:
     • Fainting: ask whether they have ever actually fainted in a panic. Explain that fainting needs a drop in blood pressure, whereas anxiety raises it. The faint feeling comes from blood being redistributed.
     • Heart attack: cardiac chest pain is usually central, not left-sided. A body-outline diagram of chest-pain locations is used. Left-sided pain in panic links to attention and belief.
     • Going mad: panic has no link to psychosis.
     • The fight/flight response and why its sensations are harmless.
     • Heart rate: it rises only modestly during panic.
  3. Re-rate the belief.
• Example language (illustrative): "Fainting happens when blood pressure drops. When you're anxious, it goes up, which is the opposite. So what does that suggest about the dizzy feeling?"
• Watch for: Information taken as reassurance and asked for again and again. Turn it into an experiment instead (B-series).

A9. Using old contradictory information

• Purpose: Use the patient's own history against the catastrophic belief.
• Steps:
  1. Look for events that don't fit the catastrophe. Examples:
     • panic stopped when the patient was distracted (a heart attack wouldn't);
     • attacks only happen when alone, or only when help is unavailable;
     • the fear started after a life event, such as a relative's death.
  2. Ask what each event means for the belief.
• Example language (illustrative): "You said the attack faded when your phone rang. If it had been a heart attack, would a phone call have stopped it?"
• Watch for: The patient discounting the evidence ("I was lucky"). That signals a safety behaviour or a superstitious belief to test next.

A10. Paired associates

• Purpose: Show that thinking about catastrophes can by itself produce the sensations. This supports the circle.
• Steps:
  1. Present a sheet of word pairs, each pairing a sensation with a catastrophe (for example breathless–suffocate, palpitations–dying). Write your own pairs.
  2. Introduce it as a diagnostic task. Do not warn in advance that it may bring on panic.
  3. Have the patient read the pairs aloud and dwell on what they mean.
  4. Ask what they notice in the body and what they are thinking.
  5. Link the sensations they report to the reading.
• Example language (illustrative): "Read each pair aloud and really think about what it means. Then tell me what you notice in your body."
• Watch for:
  • Panic or strong sensations. Use them straight away as evidence.
  • No response. That is fine; move to overbreathing.

A11. Discussion-induced panic

• Purpose: Use an attack that happens in session as live evidence.
• Steps:
  1. If talking about the fears triggers sensations, stop.
  2. Ask what the patient feels and thinks.
  3. Point out that talking alone produced the sensations.
  4. Relate this to the circle.
• Example language (illustrative): "We were only talking about your heart, and now it's racing. What does that tell us about where these feelings can come from?"
• Watch for: The therapist backing off or soothing. Stay with the evidence.

A12. Voluntary hyperventilation

• Purpose: Show that harmless overbreathing reproduces many panic sensations.
• Steps:
  1. After medical screening, have the patient breathe fast and deep, the way they breathe when alarmed, for a short period.
  2. Ask them to list the sensations and compare them with their attacks.
  3. Discuss the harmless cause.
  4. Re-rate the belief.
  5. Note that this is a demonstration. It is not training in breathing control.
• Example language (illustrative): "Breathe like you do when you're frightened, fast and deep, and keep going till I say. Now, how close is this to an attack?"
• Watch for:
  • Partial similarity is still useful.
  • Patients who start "controlled breathing" afterwards as a safety behaviour.

A13. Focus-of-attention experiment

• Purpose: Show that attending to the body increases awareness of sensations.
• Steps:
  1. Ask the patient to focus on a body area (heartbeat, throat, chest) for a few minutes and report what they notice.
  2. Then shift their attention to an external task and compare.
  3. Link this to their body checking between attacks.
• Example language (illustrative): "Just notice your throat for two minutes. ... Now describe everything you can see out of the window. What happened to the feeling in your throat?"
• Watch for: Patients who conclude they should distract themselves. Frame the point as information about the cause, not as a coping tool.

A14. Chest-pain induction

• Purpose: For cardiac fears, show that shallow upper-chest breathing and posture can cause chest discomfort.
• Steps:
  1. Have the patient breathe shallowly into the upper chest for a while, with or without a slumped posture.
  2. Ask where they feel discomfort.
  3. Compare it with their attack pain and with the body diagram from A8.
• Example language (illustrative): "Breathe high and shallow, just in your upper chest, and sit hunched. Tell me if any tightness comes."
• Watch for: Any medical doubt about chest pain. Check with the GP first.

A15. Showing paradoxical effects of safety behaviours

• Purpose: Show that the patient's coping makes the symptoms worse.
• Steps: Run small experiments matched to the patient's own safety behaviours:
  • staring fixedly, to show it produces unreality;
  • walking stiffly or locking the legs, to show it increases unsteadiness;
  • trying to suppress a thought, to show it increases intrusions;
  • deliberate deep breathing, to show it increases breathlessness.
• Example language (illustrative): "You stare at one point to keep things real. Let's try staring at this spot for a minute and see what happens to how real the room feels."
• Watch for: The moment the patient sees that their "protection" is part of the circle. Write that insight down.

A16. Identifying safety behaviours

• Purpose: List every action that stops the patient from finding out the catastrophe won't happen.
• Steps:
  1. For each feared outcome, ask what the patient does to prevent it or to control the sensations. Typical answers:
     • sitting down, holding on, tensing the legs;
     • slowing the breathing, carrying water or pills;
     • staying near exits, keeping a companion close;
     • distraction, escape.
  2. Use the Safety Seeking Behaviours Questionnaire as a prompt.
  3. Ask how much the catastrophe would have happened without the behaviour (0–100). A high rating shows that the behaviour is keeping the belief alive [source: CT manual; Salkovskis et al. 1999].
• Example language (illustrative): "When you feel you might faint, what do you do to stop yourself going down? ... And if you hadn't done that, how sure are you that you would have fainted?"
• Watch for: Subtle mental safety behaviours, such as checking the pulse, scanning for a chair, or praying. Ask about them directly.

A17. Behavioural experiments testing danger (in session first)

• Purpose: Disconfirm the specific catastrophe directly. Use them after the sensations have been reattributed in discussion.
• Steps (all experiments):
  1. Agree the experiment in advance.
  2. Write down the specific prediction and rate belief in it.
  3. Agree which safety behaviours will be dropped.
  4. Do the experiment, with the therapist present for early ones.
  5. Record what actually happened and what was learnt.
  6. Re-rate the belief.
  7. Set the same experiment as homework.
• Manual examples:
  • Cardiac fear: run on the spot or up stairs to push the heart rate up. Medical clearance first.
  • Fainting fear: strenuous overbreathing, then standing up without holding on and without tensing the legs. Check for hidden leg-tensing.
  • Fear of going mad or losing control: deliberately invite the frightening thoughts and the unreality without trying to control them. The therapist can model this first.
  • Fear of suffocating: hold the breath to show that breathing restarts by itself. Breathing is partly reflexive.
• Example language (illustrative): "You predict that if you overbreathe and then stand without holding on, you'll faint, 70%. Let's do it and see. No gripping the chair, no tensing your legs."
• Watch for:
  • Predictions that are vague, such as "something bad". Sharpen them.
  • "It only worked because you were here." Repeat the experiment alone as homework.

A18. Exposure as a behavioural experiment, with safety behaviours dropped

• Purpose: Test the catastrophe in the places where attacks happen. Use for agoraphobic avoidance and for beliefs that are reduced in the room but remain high outside.
• Steps:
  1. Choose an avoided situation.
  2. Write the prediction.
  3. Enter the situation. When symptoms start, deliberately drop the listed safety behaviours.
  4. Optionally, deliberately exaggerate the sensations. The manual's "bricklayer's apprentice" analogy is used here: an apprentice tests a wall by pushing on it hard, not by tiptoeing past.
  5. Record the outcome.
• Rationale: The manual and Salkovskis et al. (1999) report better results when exposure is framed as dropping safety behaviours to test beliefs than when it is framed as habituation.
• Example language (illustrative): "In the supermarket, when the dizziness starts, let go of the trolley and walk to the far aisle. Make the feeling stronger if you can. Let's see whether you fall."
• Watch for:
  • Patients who stay in the situation but keep safety behaviours going. Their belief won't change.
  • Escape at the peak. Plan in advance how to stay until the prediction has been tested.

A19. Agoraphobia and social-catastrophe experiments

• Purpose: Test feared outcomes about collapse and embarrassment in public.
• Steps (manual examples):
  • Simulated fainting: the therapist first pretends to faint or collapse in a public place, then the patient does it. They observe how other people actually react.
  • Simulated loss of bladder control, done as an experiment in public to test the social consequences. The details of how the manual stages this were not visible to me [uncertain].
  • Walking alone further from home than feels safe.
• Example language (illustrative): "You're sure people would stare and laugh if you collapsed. I'll do it first over by the bench. Watch what people actually do."
• Watch for:
  • Get the patient's clear consent.
  • Check the location is safe and won't cause a real emergency response.
  • Debrief afterwards on what was learned about other people.

A20. Idiosyncratic one-off experiments

• Purpose: Test an unusual personal belief with a single, well-designed test.
• Steps:
  1. Translate the belief into an observable prediction.
  2. Design a test the patient agrees is fair. The manual describes an "air-freshener" test for one patient's specific fear.
  3. Run it, record the result, and re-rate the belief.
• Example language (illustrative): "What would we have to see to know whether that's true? Let's set that up."
• Watch for: Tests that are designed to fail or that can't disconfirm anything. Agree beforehand what result would count against the belief.

A21. Behavioural-experiment record sheet (homework)

• Purpose: Structure experiments between sessions.
• Steps:
  1. The sheet has five columns.
  2. Before the experiment, fill in the first three: situation, prediction (with belief rating), and the experiment including which safety behaviours will be dropped.
  3. Afterwards, fill in the last two: outcome, and what I have learnt.
  4. There is deliberately no anxiety column.
  5. Review it next session.
• Example language (illustrative): "Fill in the first three columns now, before you go. Then after you've done it, just two things: what happened, and what that tells you."
• Watch for: Learning written as "I coped" rather than "the catastrophe didn't happen." Re-word it with the patient.

A22. Image modification

• Purpose: Change vivid, recurring catastrophic images, for example of dying or being taken away, when talking alone doesn't shift them.
• Steps:
  1. Discuss simple images verbally first.
  2. For vivid, repeated images, "finish out" the image: run it on past the worst moment to a realistic ending.
  3. Or transform it in stages into a non-threatening version that the patient chooses. The manual's example turns threatening white-coated men into poodles.
  4. The patient practises the new image as homework.
• Example language (illustrative): "Let the picture keep running past the moment you collapse. What happens next? ... And after that?"
• Watch for:
  • The image returning unchanged. Check it was the image driving the fear.
  • Rate belief in the image before and after.

A23. Session audiotapes

• Purpose: Consolidate what was learned in session.
• Steps:
  1. Record each session.
  2. The patient listens again before the next session and notes key points.
• Example language (illustrative): "When you listen back, jot down the one or two things that changed how you see the dizziness."
• Watch for: Listening used as reassurance during an attack. Keep it to review times.

A24. Relapse prevention

• Purpose: Make the gains last after therapy.
• Steps:
  1. Write a "blueprint" together. It summarises:
     • the model;
     • the old beliefs;
     • the evidence and experiments that changed them;
     • what to do if sensations or attacks return.
  2. Anticipate future stressors that might bring sensations back.
  3. Reassess any remaining belief in each catastrophe and run more experiments if any is still above zero.
  4. Do point/counterpoint role-play: the therapist voices the catastrophic belief and the patient answers it, then they swap.
  5. Space out the final sessions.
• Example language (illustrative): "I'll be the part of you that says 'this time it really is your heart'. You answer me with what you now know."
• Watch for: Hesitant answers in the role-play. They show residual belief that needs another experiment.

A25. Medication planning

• Purpose: Make sure patients learn that panic is survivable without medication protecting them.
• Steps:
  1. Keep prescribed medication at a steady dose during CT.
  2. Once the patient is confident in psychological coping, plan a gradual withdrawal with the prescriber.
  3. If the medication suppresses attacks completely, discuss withdrawal earlier so experiments become possible.
  4. Treat "carrying pills" as a safety behaviour.
• Example language (illustrative): "When you feel ready, we could talk to your doctor about reducing slowly. If symptoms come up, they're a chance to test what you've learned."
• Watch for: Abrupt self-discontinuation. Refer to the prescriber; no dosing advice here.

A26. Involving significant others

• Purpose: Address the beliefs and responses of others that keep the problem going.
• Steps:
  1. At assessment, ask what partners or relatives believe about the attacks and how they respond, for example by accompanying, rescuing or reassuring.
  2. With consent, share the model with them.
  3. Agree how they can stop acting as a safety signal.
• Sources: The CT manual covers this at assessment. MAP gives a full chapter, Involving Others [source: Psychology Tools MAP page]. The joint-session steps above are my synthesis [synthesis].
• Example language (illustrative): "When she starts to panic, what do you usually do? ... How might that make it harder for her to find out she's safe?"
• Watch for: A relative's own health anxiety, or their need to protect. Include them in planning.

Part B. Panic control treatment / MAP (Barlow & Craske), 5th edition.

The chapter list [source: Psychology Tools MAP page]. The chapter texts are not drawn on here. Detailed steps come from Schmidt & Trakowski (2004) and Craske et al. (2014), or are marked [uncertain] or [synthesis].

B1. Recording panic and anxiety (MAP ch. 5)

• Purpose: Replace a vague, global impression of anxiety with objective records, and give a baseline.
• Steps:
  1. The client records each panic attack: when, where, sensations, thoughts, and whether it was expected.
  2. The client also records daily levels of general anxiety and of worry about panic.
  3. Review the records every session and chart progress.
  • The exact MAP forms are not drawn on [uncertain].
• Example language (illustrative): "Write it down soon after, as an observer would: what happened, what you felt, what you thought."
• Watch for: Monitoring that increases body-checking. Keep entries brief.

B2. Psychoeducation: nature of panic, negative cycles, "panic attacks are not harmful" (MAP ch. 2, 6, 7)

• Purpose: Explain panic as a false alarm of the fight/flight system, kept going by fear of sensations and by avoidance, and correct beliefs that it is dangerous.
• Steps:
  1. Describe the physical, cognitive and behavioural parts of anxiety.
  2. Map the client's own cycle: sensation, then fearful thought, then more sensation, then avoidance.
  3. Go through the specific fears (heart, fainting, madness, control) and why each does not happen.
• Example language (illustrative): "Your body is running a fire drill when there's no fire. The alarm is loud, but it's still a drill."
• Watch for: Education used as reassurance. Move on to exposure to test it.

B3. Building the agoraphobia hierarchy (MAP ch. 8)

• Purpose: List the avoided situations to use for later exposure.
• Steps:
  1. List the situations the client avoids or endures with dread, including those done only with a safety signal.
  2. Rate the anxiety each would cause.
  3. Order them.
  4. Under the 5th edition's inhibitory-learning rules, use the list as a menu, not a fixed ladder. Vary the order and difficulty [source: Craske et al. 2014].
• Example language (illustrative): "Let's list everywhere panic has taken from you, including places you go only with someone or with your pills."
• Watch for: Situations left off because they are "not important". Ask about every area of life.

B4. Breathing skills (MAP ch. 9)

• Purpose: In PCT, teach slower, diaphragmatic breathing to reduce physical arousal [uncertain: the 5th-edition chapter text. This description reflects earlier PCT descriptions].
• Steps (general):
  1. Teach breathing with the abdomen rather than the upper chest, at a slow, even pace.
  2. Practise first when calm, then during anxiety.
• Example language (illustrative): "Let your stomach rise as you breathe in, and slow it right down."
• Watch for: The Clark & Salkovskis manual dropped breathing retraining because patients used it as a safety behaviour [source: CT manual 2009]. If a client relies on breathing to avert catastrophe, treat it as a safety behaviour and drop it during exposures [synthesis].

B5. Thinking skills (MAP ch. 10)

• Purpose: Challenge catastrophic misinterpretations of symptoms.
• Steps: PCT traditionally targets two errors [uncertain: the 5th-edition labels for these were not visible]:
  1. Overestimating probability. Treat the thought as a guess. Look at the evidence, including how many times the client has panicked without the catastrophe. Estimate a realistic probability.
  2. Catastrophic thinking. Ask "what if it did happen?" and work through how the client would cope, and how bad it would really be.
  • Under inhibitory learning, do this reappraisal after exposure trials to consolidate learning, not before them [source: Craske et al. 2014].
• Example language (illustrative): "You've had about 200 attacks and predicted a heart attack each time. What does that track record suggest about the odds?"
• Watch for: Mechanical answers that the client doesn't believe. Test the belief with exposure instead.

B6. Interoceptive assessment (MAP ch. 11; Schmidt & Trakowski)

• Purpose: Find which induced sensations frighten this client, to target in exposure.
• Steps:
  1. Get medical screening first. Run the exercises in a set order, with a few minutes' break between them, and the most demanding last. The standard set:
     • head shaking side to side, 30 seconds;
     • head between the knees, then sitting up;
     • running on the spot, 1 minute (stairs if needed);
     • holding the breath as long as possible;
     • touching the back of the throat (gag);
     • spinning in place or in a swivel chair, 1 minute, with a spotter;
     • push-ups;
     • hyperventilation, about 2 minutes;
     • breathing through a very thin straw (for example a coffee stirrer) with the nose pinched, up to 2 minutes.
  2. After each exercise, the client describes the sensations in their own words. They rate three things from 0 to 10: intensity, anxiety, and similarity to panic.
  3. Explain beforehand that sensations may linger afterwards.
• Example language (illustrative): "Spin until you're really dizzy. I'm right here as a spotter. Now rate the dizziness, how scary it was, and how much it felt like panic, separately."
• Watch for:
  • Half-hearted effort ("cheating"). Ask the client to repeat it more vigorously.
  • Confusing intensity with fear. Reinforce the distinction.
  • Leftover arousal skewing the next rating.

B7. Interoceptive exposure practice

• Purpose: Break the link between sensations and fear through repeated, deliberate induction.
• Steps:
  1. Choose the exercises rated as anxiety-provoking and similar to panic.
  2. In session, do repeated trials.
  3. Older habituation version [source: Schmidt & Trakowski 2004]:
     • aim for moderate anxiety (about 5 out of 10), adjusting duration if needed;
     • repeat until anxiety falls;
     • continue across sessions until the first trial of the day brings no fear.
  4. 5th-edition inhibitory version [source: Craske et al. 2014]:
     • before each trial, rate the probability of the feared harm (stroke, fainting and so on);
     • extend the trial past the point the client expects harm (for example, spin longer than they think safe);
     • stop when the expectancy has been tested, not when fear drops;
     • afterwards, ask: did it happen? How do you know? What did you learn?
  5. Assign daily home practice, alone.
• Example language (illustrative): "You're 85% sure spinning for a minute will give you a stroke. This time we go for 90 seconds. ... Did it happen? What does that tell you about dizziness?"
• Watch for:
  • The group or the therapist acting as a safety signal, so that practice alone at home feels harder. Plan for it.
  • Sensitisation instead of decline. Start easier, lengthen practice, and normalise slower progress [source: Schmidt & Trakowski 2004].

B8. Advanced and idiosyncratic interoceptive exercises

• Purpose: Match exposure to sensations the standard set misses.
• Steps: Choose according to the client's avoidance [source: Schmidt & Trakowski 2004]:
  • heat: a hot, steamy bathroom while wearing a heavy coat, or a sauna or hot car;
  • staring without blinking at a wall or a mirror, for unreality;
  • caffeine, for clients who avoid it, in amounts they normally tolerate and with medical clearance;
  • a snug belt around the chest, rolling up in a blanket, or eating dry or salty food without water, for choking or constriction fears;
  • paper-bag rebreathing, as an office substitute for the laboratory CO₂ challenge.
  • Laboratory CO₂ inhalation and medication-based inductions need specialist settings and are not covered here.
• Example language (illustrative): "You avoid hot rooms because the flushed, heavy feeling scares you. Let's make that feeling on purpose."
• Watch for: Subtle avoidance of exercise, caffeine, sex, anger, excitement or horror films. These count as interoceptive avoidance too.

B9. Facing agoraphobia situations: in vivo exposure (MAP ch. 12)

• Purpose: Re-enter avoided situations and learn that panic in them is tolerable and not dangerous.
• Steps:
  1. Choose items from the hierarchy.
  2. Write down the feared outcome and its probability.
  3. Enter the situation without safety signals, staying until the prediction has been tested.
  4. Combine with interoceptive inductions for deepened extinction. For example, induce a racing heart before entering a shop [source: Craske et al. 2014].
  5. Practise in several contexts: alone, unfamiliar places, different times of day.
  6. Record the outcome and what was learned.
• Example language (illustrative): "Before the bus, run up the stairs so your heart's already going. Then ride four stops alone, no phone in your hand. What do you predict?"
• Watch for:
  • Escape at peak fear.
  • A companion waiting just outside. In the Craske case, the wife stopped waiting outside and the client stopped bringing his medication [source: Craske et al. 2014].

B10. Inhibitory-learning enhancements for any exposure

Source for this item: Craske et al. 2014.

• Purpose: Make exposure learning stronger and longer-lasting.
• Steps (each is a separate tool):
  1. Expectancy violation: name and rate the feared outcome, then design the exposure to test it.
  2. Deepened extinction: practise cues one at a time, then combine them, for example caffeine, then jogging, then spinning, in one trial.
  3. Removing safety signals: phase them out, or drop them at once if the client agrees.
  4. Variability: vary the stimuli, length, intensity and order.
  5. Multiple contexts.
  6. Retrieval cues: recall a successful exposure as a relapse-prevention aid. Use sparingly so it does not become a safety signal.
  7. Affect labelling: during exposure, the client names their feeling ("I feel scared") without trying to change it.
  • The paper also describes "occasional reinforced extinction" (deliberately bringing on panic during situational exposure) as experimental. Not recommended for routine practice [uncertain].
• Example language (illustrative): "While you're on the escalator, just say out loud what you're feeling, 'I'm frightened, my heart's pounding', and keep going."
• Watch for: Fear that fluctuates rather than falls is expected. Judge progress by violated expectancies and expanded activity.

B11. Medications (MAP ch. 14)

• Purpose: Integrate CBT with SSRIs or benzodiazepines, and support later discontinuation [source: Psychology Tools MAP page].
• Steps:
  1. Map medication use, including "as needed" use and carried pills.
  2. Treat carried or "as needed" medication as a safety signal to fade during exposure.
  3. Coordinate any taper with the prescriber.
  • NICE advises against benzodiazepines for panic disorder [source: NICE CG113 1.3.20].
  • The chapter's detailed taper guidance is not drawn on [uncertain].
• Example language (illustrative): "Let's plan with your doctor how to reduce gradually, and treat any symptoms that come up as more exposure practice."
• Watch for: Rebound symptoms read as a relapse. Reframe them as sensations to test.

B12. Accomplishments, maintenance and relapse prevention (MAP ch. 15)

• Purpose: Consolidate gains and plan for lapses.
• Steps:
  1. Review progress against the baseline records.
  2. List the remaining avoidance and keep scheduling exposures.
  3. Separate a lapse (one attack) from a relapse.
  4. Plan what to do when symptoms return: re-start interoceptive and in vivo practice.
  • Detailed chapter content is not drawn on [uncertain].
• Example language (illustrative): "An attack next year doesn't wipe this out. It's a cue to go back to practice, not a sign you're back at square one."
• Watch for: Stopping practice as soon as symptoms ease.

B13. Brief six-session version for primary care (MAP ch. 16)

• Purpose: A condensed PCT for primary care and similar settings [source: Psychology Tools MAP page].
• Content: Session-by-session content was not visible [uncertain]. A practical approach is to keep psychoeducation, self-monitoring, interoceptive exposure and in vivo exposure, and to cut back the didactic time [synthesis].

5. Handling problems

Items tagged [source: …] come from the sources. Items tagged [synthesis] are my own clinical guidance and are not stated in those sources.

Problems the sources address

• Hidden safety behaviours during experiments, such as tensing the legs during a fainting test or gripping a chair. If the belief doesn't drop, ask exactly what the patient did to stay safe, then repeat the experiment without it [source: CT manual 2009; Salkovskis et al. 1999].
• Coping skills turning into safety behaviours, such as breathing, distraction, or reciting reassurances. Reframe them as things that block learning. Run the experiment again with the skill deliberately dropped, or with symptoms deliberately exaggerated (the bricklayer's apprentice) [source: CT manual 2009].
• Medication that blocks attacks completely, so nothing can be tested. Discuss a gradual, prescriber-led withdrawal once the patient feels ready, so the beliefs can be tested [source: CT manual 2009].
• Physical illness: cardiac disease, pregnancy, epilepsy, asthma.
  • Modify experiments or leave out the risky ones. Get medical clearance [source: CT manual 2009].
  • [synthesis] Do not use overbreathing or straw-breathing with asthma or epilepsy without the treating doctor's agreement.
  • [synthesis] For cardiac fears in a patient with known heart disease, agree the exercise limits with the cardiologist and test the catastrophic misreading within those limits.
• Another primary disorder, such as psychosis, severe depression or intoxication. Treat that first, or arrange detoxification [source: CT manual 2009].
• Half-hearted interoceptive exercises. Ask for a more vigorous repeat, or switch to a stronger related exercise [source: Schmidt & Trakowski 2004].
• Sensitisation instead of decline. Start with easier exercises, lengthen the practice, and explain that more trials may be needed. Under the inhibitory-learning approach, judge progress by expectancies that have been disconfirmed, not by falling fear [source: Schmidt & Trakowski 2004; Craske et al. 2014].
• Progress only with the therapist or the group present (a safety signal). Set the same task alone at home and in new places [source: Schmidt & Trakowski 2004; Craske et al. 2014].
• Belief drops in the room but stays high in attacks. Move to in-situation experiments with safety behaviours dropped [source: CT manual 2009].
• Persistent catastrophic images. Use image modification (A22) [source: CT manual 2009].
• Residual belief at the end of therapy. Re-rate every catastrophe. Use point/counterpoint role-play and more experiments [source: CT manual 2009].

Problems not covered in the sources (all [synthesis])

• Can't identify any thoughts ("it just comes out of the blue"):
  • Use paired associates or induction in session and catch the thought live.
  • Ask about images.
  • Ask what the patient did in response to the sensation. The safety behaviour shows what they feared.
• Nocturnal panic:
  • Treat waking sensations as the trigger. Map the sleep-time catastrophe, for example "I'll stop breathing in my sleep".
  • Test it, for example with breath-holding to show breathing restarts by itself. The manual mentions asking a patient whether they try to breathe during sleep; Full guidance on this is not given [uncertain].
• Repeated reassurance-seeking from the therapist, the GP, A&E or the internet:
  • Name it as a safety behaviour.
  • Agree a plan to reduce it, keeping one agreed GP check if there is a new medical concern.
  • Turn the question into a prediction to test.
• Non-adherence to homework:
  • Check that the patient understands the rationale. Ask them to explain the circle back.
  • Shrink the task.
  • Do the first trial together in session.
  • Use the record sheet's "what I learned" column, not anxiety, as the measure of success.
  • Ask what they predicted would happen if they did the homework. Often it is the same catastrophe.
• Overwhelm or refusal of an experiment:
  • Never force it. Re-negotiate a smaller version, a shorter duration, or the therapist modelling it first.
  • Keep the choice with the client.
  • Remember that a short-term rise in anxiety is expected, and say so beforehand.
  • If the client wants to stop, stop, then debrief on what they predicted and what happened up to that point.
• Rupture, for example the client feeling pushed, disbelieved or lectured:
  • Acknowledge it openly and ask for feedback.
  • Return to collaborative empiricism: you are testing their hypothesis together, not proving them wrong.
  • Lower the pace.
  • Check whether education has felt dismissive of real symptoms.
• "Yes, but this time it could be real" (head–heart split): rate belief while the patient is having the sensations, not only in the room. Then run in-situation experiments with the sensations exaggerated.
• Depersonalisation or derealisation as the main fear:
  • Reproduce it by staring (A15/B8).
  • Educate that it is a normal anxiety and attention effect, not madness.
  • Test "going mad" by deliberately inviting the feeling.
• Comorbid depression:
  • Use the magic-wand question to decide which problem comes first [source: CT manual 2009].
  • If panic is primary, add activity scheduling alongside.
  • Screen for suicidal thoughts at assessment and when mood drops. Follow your service's risk-assessment protocol. If means are a concern, include lethal-means counselling about safe storage or removal.
• Alcohol or cannabis used to dampen sensations: treat it as a safety behaviour. Agree on reduction before key experiments, because otherwise their results are confounded.
• Relapse after discharge: re-read the blueprint. Return to an interoceptive and in vivo practice block. Offer a booster session. The MAP 5th edition and CT manual details on boosters were not seen [uncertain].
• Choosing between CT and PCT procedures: both target the fear of sensations.
  • Lead with CT's belief-testing when the patient has clear, specific catastrophes and safety behaviours.
  • Lead with PCT interoceptive and situational exposure when the main problem is broad fear of arousal or extensive avoidance.
  • Do not teach breathing control as a rescue technique in either approach.

Sources

Sources

• Cognitive Therapy for Panic Disorder (Clark & Salkovskis, 2009, IAPT manual) – OxCADAT Resources
• Treating panic and social anxiety disorders – Oxford Neuroscience
• Salkovskis et al. (1999) safety-seeking behaviours in panic disorder with agoraphobia (BRAT 37:559–574), PDF
• Brief cognitive therapy for panic disorder: a randomized controlled trial (Clark et al., 1999) – University of Manchester research portal
• NICE CG113: Generalised anxiety disorder and panic disorder in adults – Guidance
• Mastery of Your Anxiety and Panic (5th ed.): Therapist Guide – Psychology Tools
• Mastery of Your Anxiety and Panic: Workbook (5th ed.) – Oxford Academic
• Schmidt & Trakowski (2004) Interoceptive assessment and exposure in panic disorder: a descriptive study, PDF
• Craske, Treanor, Conway, Zbozinek & Vervliet (2014) Maximizing exposure therapy: an inhibitory learning approach, PDF

Books and papers cited

• Clark, D. M. (1986). A cognitive approach to panic. Behaviour Research and Therapy, 24, 461–470. Not drawn on directly; known through the manual.
• Clark, D. M., & Salkovskis, P. M. (2009). Cognitive Therapy for Panic Disorder: Manual for IAPT High Intensity Therapists. University of Oxford / King's College London.
• Clark, D. M., Salkovskis, P. M., Hackmann, A., Wells, A., Ludgate, J., & Gelder, M. (1999). Brief cognitive therapy for panic disorder: a randomized controlled trial. Journal of Consulting and Clinical Psychology, 67(4), 583–589.
• Salkovskis, P. M., Clark, D. M., Hackmann, A., Wells, A., & Gelder, M. G. (1999). An experimental investigation of the role of safety-seeking behaviours in the maintenance of panic disorder with agoraphobia. Behaviour Research and Therapy, 37, 559–574.
• Barlow, D. H., & Craske, M. G. (2022). Mastery of Your Anxiety and Panic: Therapist Guide (5th ed.). Oxford University Press. Chapter list only; text not drawn on.
• Craske, M. G., & Barlow, D. H. (2022). Mastery of Your Anxiety and Panic: Workbook (5th ed.). Oxford University Press. Not drawn on.
• Schmidt, N. B., & Trakowski, J. (2004). Interoceptive assessment and exposure in panic disorder: a descriptive study. Cognitive and Behavioral Practice, 11, 81–92.
• Craske, M. G., Treanor, M., Conway, C. C., Zbozinek, T., & Vervliet, B. (2014). Maximizing exposure therapy: an inhibitory learning approach. Behaviour Research and Therapy, 58, 10–23.
• National Institute for Health and Care Excellence (2011). Generalised anxiety disorder and panic disorder in adults: management (CG113).