Bergen 4-day treatment

1. Introduction

Version followed: the Bergen 4-Day Treatment (B4DT) for OCD as described by its developers (Gerd Kvale, Bjarne Hansen) in the published trial reports of 2018–2020 (Kvale et al. 2018; Hansen et al. 2018; Launes et al. 2019, 2020), plus the developers' published adaptations for panic disorder (Hansen et al. 2018 pilot; Iversen et al. 2022; Eide et al. 2023) and social anxiety disorder (Hansen et al. 2024). The B4DT manual itself (a PowerPoint deck with an accompanying text file) is not public; everything below is reconstructed from the trial method sections. [uncertain: wording and slide-level content of the manual]

B4DT is concentrated, therapist-assisted exposure (exposure and response prevention, ERP, for OCD) delivered over four consecutive days. Its format is "individual treatment in a group setting": 3–6 patients are each paired 1:1 with a therapist inside one group run by a certified group leader. [source: Kvale et al. 2018; Hansen et al. 2018]

How change is thought to happen: the patient learns one transferable skill, the Leaning-in Technique (LET), "LEan into The anxiety". LET means deliberately approaching whatever triggers anxiety or discomfort, and noticing the exact moment one is tempted to avoid, ritualise or use a safety behaviour, then doing something incompatible with the disorder instead. Reducing anxiety and disconfirming catastrophic beliefs are explicitly not the goals. Anxiety is treated as the "raw material for change", and the patient leaves as their own therapist. [source: Kvale et al. 2018; Launes et al. 2019; Hansen et al. 2018 panic pilot; Eide et al. 2023] The panic version also cites Clark's cognitive model and inhibitory learning as influences. [source: Iversen et al. 2022] A process study found that leaning into anxiety on Day 3 predicted OCD improvement. [source: Miegel et al. 2025, abstract only]

2. Framework

Total contact is about 21–23 hours over four consecutive days, followed by 3 weeks of self-directed exposure and one booster session at about 3 months. [source: Hansen et al. 2018 panic pilot; Eide et al. 2023]

[source: Kvale et al. 2018; Hansen et al. 2018; Launes et al. 2019, 2020; Hansen et al. 2018 panic pilot; Iversen et al. 2022; Eide et al. 2023; Hansen et al. 2024]

Versions differ on these points:

• Day 2–3 length is 8–10 h for OCD and about 7–8 h for panic and SAD.
• The expectancy-rating threshold for follow-up is any score below 100% (Kvale et al. 2018) or below 70% (Hansen et al. 2018; Launes et al. 2019).
• The booster is 30 min (OCD) or 45 min (panic, Iversen et al. 2022).

Eligibility and staffing:

• Published exclusions are active suicidality, psychosis, active substance abuse, bipolar disorder in an unstable phase, a primary or severe eating disorder, and not speaking the treatment language. Misdiagnosis (for example OCD-like symptoms from a neurological condition) is also excluded. In the Bergen OCD trials, patients had to live within about one hour of the clinic so that exposures could happen in their own settings. [source: Kvale et al. 2018; Hansen et al. 2018; Eide et al. 2023; Hansen et al. 2024]
• Each group has a group leader with no patients of their own, often a "second in command" and a logistics person, and one therapist per patient. [source: Kvale et al. 2018]
• Certification: a therapist must have taken part in at least 2–3 groups and been judged competent, in Kvale et al. 2018 by two independent B4DT experts, in some trials from video. A group leader must have worked in at least six groups and be approved by a B4DT expert. A new clinic must train at least three therapists including one certified leader, deliver the manual faithfully, and be supported by certified therapists in its first groups. [source: Kvale et al. 2018; Hansen et al. 2018; Launes et al. 2020; Eide et al. 2023] Panic teams also took a two-day panic course. [source: Bergen panic study, BMC Psychiatry 2023 (s12888-023-05062-7)] The developers' training pathway is required to call a service B4DT. [synthesis: a clinician without certification can apply the principles below, but should not describe the result as B4DT]

3. Guiding principles

• Individual treatment in a group setting. Plan, coach and rate each patient individually, and use the group for norms, report-back and modelling. [source: Kvale et al. 2018; Hansen et al. 2018]
• Lean in; do not regulate anxiety. Never coach the patient to lower anxiety. The task is to approach it and drop control, so anxiety reduction is not a success criterion. [source: Eide et al. 2023; Hansen et al. 2018 panic pilot]
• Not behavioural experiments. In the panic version, do not set exposures up to disprove catastrophic predictions. Practise willingness and letting go of control. [source: Bergen panic study, BMC Psychiatry 2023] Iversen et al. 2022 describe "exposure and behavioural experiments" for panic. [uncertain: the sites' versions differ]
• Anxiety is the raw material. Treat discomfort as the opportunity. When it comes back after treatment, frame it as a new chance to practise. [source: Hansen et al. 2018; Launes et al. 2019]
• The moment of temptation is the point of change. Teach the patient to catch the exact second they want to avoid, ritualise or use a safety behaviour, and to choose the opposite there. [source: Launes et al. 2019; Hansen et al. 2018 panic pilot; Eide et al. 2023]
• Quality of leaning in over task difficulty. A simple task done with full engagement beats a hard task done with safety strategies. Move to harder tasks only once LET is applied correctly. [source: Hansen et al. 2024]
• Rate what the patient controls. The 1–6 LET rating measures not holding back, not anxiety level, so every patient can reach 6. [source: Hansen et al. 2018 panic pilot]
• Patients choose tasks, and the most disliked are often the most useful. Have patients propose tasks, steer them towards the ones their disorder "would appreciate the least", and start with those likely to change daily life most. [source: Kvale et al. 2018; Launes et al. 2019]
• Zero tolerance for rituals (OCD). Treat the externalised OCD as something whose demands the patient no longer obeys, including subtle and mental rituals. [source: Launes et al. 2019]
• Hunt for anxiety; leave none behind. When a task stops provoking anxiety, move to a new trigger rather than repeating it. [source: Hansen et al. 2018 panic pilot]
• Concentration and commitment. The patient makes an active decision to get rid of the disorder, keeps the four days clear and works full days. [source: Hansen et al. 2018 panic pilot; Helse Bergen]
• Gradual transfer of responsibility. The therapist coaches at first, then fades. By Day 3 the patient leans in alone, and by Day 4 they are their own therapist. [source: Kvale et al. 2018; Launes et al. 2019]
• Team, not dyad. Therapists rotate between patients. The most experienced therapist goes to the patient who is struggling most. Colleagues observe each other, and frequent short team meetings check fidelity. [source: Kvale et al. 2018; Launes et al. 2019; Iversen et al. 2022]
• Real-life context. Do exposures in the settings where the symptoms occur: home, shops, buses, work. [source: Hansen et al. 2018; Eide et al. 2023]
• Therapist tolerance of distress. Therapists must be able to stay with intense patient discomfort without rescuing, and accept being observed. [source: Launes et al. 2019]
• Fidelity to the manual. Deliver the manualised content without improvising, and use the adherence forms. [source: Hansen et al. 2018; Hansen et al. 2024]

4. Intervention inventory

All example language is illustrative unless it is in quotation marks with a citation. The published papers give the structure and core phrases. They do not give scripts.

4.1 Diagnostic screening and eligibility check

• Purpose and when: Before allocation to a group, to confirm a principal diagnosis of OCD, panic disorder or social anxiety disorder, and to rule out exclusions. [source: Kvale et al. 2018; Eide et al. 2023; Hansen et al. 2024]
• Steps:
  1. Run a structured interview (the trials used the MINI) and a clinician-rated severity measure: Y-BOCS for OCD, PDSS for panic, a structured interview plus self-report for SAD.
  2. Screen for active suicidality, psychosis, active substance misuse, unstable bipolar disorder and a severe or primary eating disorder.
  3. Check for OCD-like presentations of another cause, such as neurological illness or a primary GAD. [source: Kvale et al. 2018]
  4. Take a medication history. Psychotropics are kept stable (SAD: a stable dose for 4 weeks before starting). [source: Hansen et al. 2024]
  5. Collect baseline self-report measures (the trials used PHQ-9 and GAD-7).
  6. For exposure safety, ask about cardiac, respiratory, seizure and other medical conditions, and pregnancy, before any interoceptive exercise, and get medical clearance where indicated. [synthesis: standard interoceptive-exposure precaution; not described in the B4DT papers]
  7. Assess suicide risk using your service's standard structured risk assessment. If active suicidality is present, treat that first and reconsider B4DT later. [synthesis]
• Example language (illustrative): "Before we offer the four days, I need a clear picture of what you're struggling with and to check that this format is safe and right for you now."
• Watch for: Comorbid depression is not an exclusion in the published criteria. [uncertain: the papers do not say where the severity cut-off lies]

4.2 Standardised information package and commitment

• Purpose and when: Weeks before treatment, so that every patient starts with the same understanding and has made an active choice. [source: Kvale et al. 2018; Hansen et al. 2018 panic pilot]
• Steps:
  1. Send written information on the treatment and the focus of each day.
  2. Show video 1, which explains the rationale and stresses that the patient must make an active decision to get rid of the disorder and engage fully. (Bergen also used a short 2014 NRK broadcast.)
  3. Show video 2, which covers the content, structure and focus of the four days.
  4. State the practical commitments: no other appointments during the four days, including the evenings of Days 2 and 3. In the OCD trials, patients had to live within about one hour of the clinic.
  5. Get written informed consent. The consent should name the intensity, the deliberate provocation of anxiety, evening work and texting the therapist. [synthesis: consent detail; trials report written consent]
• Example language (illustrative): "These four days are demanding. We'll be deliberately looking for your anxiety, not avoiding it. The people who get most from it are the ones who decide beforehand to go all in."
• Watch for: Ambivalence or a plan to fit appointments around the days. Address these before the start date.

4.3 Expectancy and credibility check

• Purpose and when: After the information package, to find misunderstandings and obstacles early. [source: Kvale et al. 2018; Hansen et al. 2018; Launes et al. 2019]
• Steps:
  1. Have the patient rate, from 0 to 100%, how logical the treatment seems, whether they would recommend it, how likely they are to attend fully, and whether they expect it to help. The trials adapted the Borkovec and Nau scale. Use the licensed or published version; items are not reproduced here.
  2. Explore any low score. The threshold is below 100% in Kvale et al. 2018 and below 70% in Hansen et al. 2018 and Launes et al. 2019. Clarify misunderstandings and identify practical barriers.
• Example language (illustrative): "You rated attending all four days at 60%. What would get in the way?"
• Watch for: Low credibility that comes from believing the goal is to calm anxiety down. Correct this to the LET rationale.

4.4 Patient-prepared exposure task list

• Purpose and when: Before Day 1. The patient takes ownership of the targets. [source: Kvale et al. 2018; Launes et al. 2019]
• Steps:
  1. Ask the patient to write down situations, objects, thoughts and sensations they avoid or ritualise around.
  2. Explain that exposures "their OCD would appreciate the least" are often the most relevant (Kvale et al. 2018). For panic or SAD, use the equivalent wording about the disorder.
  3. Prioritise tasks that would make the biggest change in daily life. For OCD, encourage tasks that combine several symptom domains. [source: Launes et al. 2019; Hansen et al. 2024]
  4. Tell OCD patients not to begin exposures before Day 2. [source: Launes et al. 2019]
• Example language (illustrative): "Bring the list your OCD would least like you to show us. Those are usually the ones that free up the most life."
• Watch for: Lists made up only of mild items. Ask what the patient secretly hopes nobody will suggest.

4.5 Group leader pre-treatment phone call

• Purpose and when: About one week before Day 1. [source: Kvale et al. 2018; Hansen et al. 2018 panic pilot]
• Steps:
  1. The group leader runs a short standard interview.
  2. Clarify any remaining misunderstandings.
  3. Confirm that the exposure tasks are prepared and the four days are kept clear.
  4. Restate the medication rules (4.6).
• Example language (illustrative): "I'm calling to check you have your task list ready and to answer anything that's unclear before Tuesday."
• Watch for: An unprepared list or newly booked appointments. Problem-solve these on the call.

4.6 Medication rules

• Purpose and when: From referral through the four days, so that safety-seeking through medication does not undercut leaning in. [source: Kvale et al. 2018; Hansen et al. 2018 panic pilot; Iversen et al. 2022; Eide et al. 2023]
• Steps:
  1. Leave regular medication unchanged.
  2. Tell the patient not to use anxiolytics or benzodiazepines during or directly after exposures. Kvale et al. 2018 prohibited anxiolytics during and after treatment; Hansen et al. 2018 prohibited benzodiazepines during the four days; later panic sites "encouraged" avoidance, and some patients took them at night for sleep.
  3. Any change to medication goes through the prescriber. No dosing advice is given here.
• Example language (illustrative): "Taking a tablet to take the edge off would be a way of holding back. Please keep your regular medication as it is, and talk to your doctor before changing anything."
• Watch for: Undisclosed as-needed use. Ask about it at each group report-back.

4.7 Day 1 opening: introductions and group norms

• Purpose and when: At the start of Day 1, with all patients and therapists present. [source: Launes et al. 2019; Iversen et al. 2022]
• Steps:
  1. Patients and therapists introduce themselves.
  2. Agree group norms of mutual support and confidentiality.
  3. Explain the roles: group leader, therapists, team rotation.
• Example language (illustrative): "Over the four days you'll all be doing very personal work side by side. Let's agree how we'll support each other."
• Watch for: Comparing severity or making competitive remarks. Bring the focus back to each person's own leaning in.

4.8 Day 1 psychoeducation (disorder-specific)

• Purpose and when: On Day 1 (about 3–4 h), to give a shared model and rationale. The content is manualised as a slide deck with a text file, and nothing in it is improvised. [source: Launes et al. 2019, 2020]
• Steps:
  1. Cover the core features of the disorder, its maintaining factors (avoidance, rituals, safety behaviours, control efforts), the treatment principles and how exposure is done. [source: Hansen et al. 2018 panic pilot; Eide et al. 2023; Hansen et al. 2024]
  2. OCD: Externalise OCD as something that "craves attention" and tries to trick the patient into ritualising or avoiding. The key message is that the patient will show OCD they no longer comply, with "zero tolerance for rituals". [source: Launes et al. 2019]
  3. Panic: Present a case formulation in which harmless bodily sensations are misread as signs of serious illness, and trying to control the sensations keeps panic going. [source: Bergen panic study, BMC Psychiatry 2023] The developers cite Clark's cognitive model as an influence. [source: Iversen et al. 2022]
  4. SAD: Cover the features of SAD, its maintaining factors, and the role of safety behaviours and mental self-regulation. [source: Hansen et al. 2024] [uncertain: whether SAD psychoeducation includes self-focused attention models; not stated]
  5. Introduce LET (4.9).
• Example language (illustrative, OCD): "OCD feeds on attention and obedience. Every time it makes a demand and you comply, it learns it can ask again. From Wednesday you're going to stop complying, completely."
• Watch for: Patients hearing "exposure" as "getting used to it until it stops hurting". Restate that the aim is to lean in, not to wait for anxiety to fall.

4.9 Introducing the Leaning-in Technique (LET)

• Purpose and when: Day 1. This is the central skill and is practised in every later task. [source: Kvale et al. 2018; Hansen et al. 2018; Eide et al. 2023]
• Steps:
  1. Define LET as "LEan into The anxiety" (Kvale et al. 2018): actively approach what triggers anxiety or discomfort instead of avoiding it, controlling it or regulating it down.
  2. Explain that the moment of temptation, when the patient feels the pull to avoid, ritualise, reassure or use a safety behaviour, is where change happens. Teach them to spot it and choose a response incompatible with the disorder. [source: Launes et al. 2019; Hansen et al. 2018 panic pilot]
  3. Demonstrate examples of subtle avoidance: distraction, reassurance, mental neutralising, holding on to something, half-doing a task. [source: Hansen et al. 2018 panic pilot]
  4. Introduce the 1–6 self-rating, where 6 means fully leaning in and not holding back, and stress that every patient can reach 6 (4.15). [source: Hansen et al. 2018 panic pilot; Launes et al. 2020]
  5. Frame anxiety as "raw material for change" and as somewhere to "find gold". [source: Hansen et al. 2018; Launes et al. 2019]
  6. Introduce willingness: ask "Are you willing to let go?" [source: Hansen et al. 2024]
• Example language (illustrative): "When anxiety rises, you'll feel a pull to hold back somehow. That second is the most valuable moment of the day. Instead of stepping back, lean forward into it."
• Watch for: Patients who equate LET with gritting their teeth while secretly reassuring themselves. Check what is happening inside their head, not only what they are doing.

4.10 Exposure task planning in plenary

• Purpose and when: End of Day 1, and briefly at each later meeting. [source: Kvale et al. 2018; Eide et al. 2023; Iversen et al. 2022]
• Steps:
  1. Each patient presents their proposed tasks to the group.
  2. The group leader and the group test each task's relevance: will it trigger the core fear, and does it target what the disorder least wants?
  3. Order the tasks by impact on daily life, not by a fixed hierarchy. For SAD, start with simple tasks until LET is correct, then move to harder ones. [source: Hansen et al. 2024]
  4. Agree concrete settings and logistics for Day 2.
• Example language (illustrative): "Which of these would you most like to skip? Let's put that one early."
• Watch for: Plans built as slow, graded hierarchies. B4DT moves to the most relevant triggers quickly once LET is in place. [synthesis: inferred from "hunt for anxiety" and SAD guidance]

4.11 Therapist-assisted exposure with LET coaching (core procedure, Days 2–3)

• Purpose and when: Most of Days 2 and 3, 1:1, in real-life settings that matter to the patient. [source: Kvale et al. 2018; Hansen et al. 2018; Iversen et al. 2022]
• Safety and consent: Before each task, confirm the patient's consent to that specific exposure. Ensure nothing illegal, physically dangerous or medically unsafe is done, and that public exposures do not endanger or involve third parties without their consent. Keep the exclusion criteria and the medical clearance from 4.1 in force. [synthesis: standard exposure safety; the B4DT papers report screening and consent but no per-task procedure]
• Steps:
  1. Go to the setting with the patient. The therapist does exposures alongside the patient where relevant. [source: Iversen et al. 2022]
  2. Restate the LET goal: not to get through the task, but to lean in without holding back.
  3. Start the task. Watch for and ask about overt and covert avoidance (4.13).
  4. At each spike, prompt the patient to name the moment of temptation and pick the opposite response (4.12).
  5. Coach the patient to increase uncertainty or arousal rather than settle it. For example, in OCD, reminding oneself that one cannot be sure one is uncontaminated. [source: Hansen et al. 2018]
  6. When the task no longer triggers anxiety, move to a new trigger rather than repeating it (4.16).
  7. Fade prompts across the day so the patient initiates leaning in (4.17).
  8. Return to the group for report-back at the scheduled time (4.18).
• Example language (illustrative): "Notice what your mind is telling you to do right now. Whatever it is, do the opposite." "Are you holding anything back, even a little?"
• Watch for: Sustained engagement, a patient who seeks out the next trigger, and LET ratings of 6. Anxiety falling within a session is not the marker. [source: Hansen et al. 2018 panic pilot]

4.12 Catching the moment of temptation and choosing a competing response

• Purpose and when: In every exposure, whenever an urge to avoid, ritualise, reassure or self-regulate appears. [source: Launes et al. 2019; Hansen et al. 2018 panic pilot; Eide et al. 2023]
• Steps:
  1. Ask the patient to say out loud the instant the urge appears.
  2. Name the urge's function: protection, certainty or control.
  3. Have the patient choose and do something incompatible with the disorder. For OCD: touch again, leave the doubt open, add uncertainty. For panic: stay, move towards the sensation, increase arousal. For SAD: drop the safety behaviour, step forward, speak up.
  4. Have the patient briefly reflect on what they did at that moment, without reviewing how anxious they felt.
• Example language (illustrative): "There. That's the moment. What does the disorder want you to do? What's the opposite?"
• Watch for: Over time, the patient catching the moment without a prompt.

4.13 Identifying and dropping subtle avoidance, safety behaviours and mental self-regulation

• Purpose and when: Throughout, especially when a task seems "too easy" or the LET rating is below 6. [source: Kvale et al. 2018; Hansen et al. 2018 panic pilot; Hansen et al. 2024]
• Steps:
  1. Therapists demonstrate common forms of subtle avoidance on Day 1. [source: Hansen et al. 2018 panic pilot]
  2. During the task, ask what the patient is doing inside their head: counting, reassuring, reviewing, distracting, monitoring their body, rehearsing.
  3. Ask about external props: holding onto things, standing near exits, carrying medication or water, avoiding eye contact, rehearsed scripts.
  4. Agree to drop each one and repeat the task without it. For SAD, ask "To what extent are you holding back?" [source: Hansen et al. 2024]
  5. For OCD, include reassurance-seeking and mental neutralising as rituals, alongside overt behaviour. [source: Hansen et al. 2018]
• Example language (illustrative): "On a scale of how much you're shielding yourself right now, what's left? Let's take that away too."
• Watch for: Anxiety that rises when a safety behaviour is dropped, which shows it was doing work. A patient who reports nothing left to drop while still rating below 6 needs closer observation.

4.14 Interoceptive exposure (panic disorder)

• Purpose and when: Panic version, Days 2–3, aimed at feared bodily sensations. [source: Hansen et al. 2018 panic pilot; Eide et al. 2023; Iversen et al. 2022]
• Contraindications and safety: Get medical clearance before strenuous exercise or hyperventilation where there is any cardiac or respiratory disease, epilepsy or seizure history, pregnancy, or another relevant medical condition. Stop for genuine medical symptoms that differ from the patient's usual panic sensations, and seek medical assessment. [synthesis: standard interoceptive-exposure precautions; not detailed in the B4DT papers]
• Steps:
  1. Identify the sensations the patient fears most, such as racing heart, breathlessness, dizziness or derealisation.
  2. Choose arousal-raising exercises the sources name, physical exercise and hyperventilation, at an intensity agreed with the patient.
  3. Coach the patient to lean in: amplify rather than calm the sensations, and drop control efforts such as slow breathing, sitting down or checking the pulse.
  4. In the B4DT pilot, patients were encouraged to try mentally to "cause a heart attack" (Hansen et al. 2018 panic pilot) rather than down-regulate. Use this only after medical screening.
  5. Combine interoceptive with situational exposure (4.15).
• Example language (illustrative): "Instead of slowing your breathing, invite the dizziness. Let's see how strong you can make it."
• Watch for: Covert breathing control, hidden pulse-checking, and sitting down at the first symptom.

4.15 Situational and agoraphobic in-vivo exposure (panic disorder)

• Purpose and when: Panic version, alongside interoceptive work. [source: Eide et al. 2023; Iversen et al. 2022]
• Steps:
  1. Select avoided places tailored to the patient, such as shopping centres, buses, queues, crowds and distance from home.
  2. Practise in as many relevant settings as possible.
  3. Combine settings with interoceptive provocation, for example exercising and then boarding a bus. [synthesis: combination inferred from "as many settings as possible" and interoceptive work]
  4. The therapist accompanies at first, then fades.
• Example language (illustrative): "Let's choose the bus route you've avoided longest, and sit where getting off is hardest."
• Watch for: Sitting near exits, keeping a companion close, a phone in hand.

4.16 Hunting for anxiety and leaving none behind

• Purpose and when: Days 2–3, whenever a task loses its power. [source: Hansen et al. 2018 panic pilot]
• Steps:
  1. Once a task no longer triggers anxiety, stop repeating it.
  2. Look for other triggers across contexts, thoughts and situations, as the group motto "hunt for anxiety without holding back" puts it.
  3. Aim to cover every relevant trigger: "we do not want to leave any behind" (Hansen et al. 2018 panic pilot).
  4. Spend little time discussing anxiety reduction.
• Example language (illustrative): "This one's lost its bite. Where else is your anxiety hiding?"
• Watch for: Patients settling into comfortable tasks. Steer them to the next area of avoidance.

4.17 Graduated transfer of responsibility

• Purpose and when: Across Days 2–4, so that the patient becomes their own therapist. [source: Kvale et al. 2018; Launes et al. 2019]
• Steps:
  1. Day 2: the therapist coaches closely and prompts at each moment of temptation.
  2. Later on Day 2: the patient chooses the next task and names their own urges. The therapist only observes and asks.
  3. Day 3: the patient performs LET without therapist help for the first time. [source: Launes et al. 2019]
  4. Evenings: self-administered exposures (4.20).
  5. Day 4: the patient writes their own plan (4.23).
• Example language (illustrative): "This time I'll stay back. You pick the task, you catch the moments, and tell me afterwards how you rated it."
• Watch for: A rating that drops when the therapist steps back. Return briefly to coaching, then fade again.

4.18 Group check-ins with LET self-rating

• Purpose and when: Brief group meetings in the morning, at lunch and late afternoon on Days 2–3. [source: Hansen et al. 2018 panic pilot; Launes et al. 2019, 2020]
• Steps:
  1. Each patient reports what they did, with a focus on how they practised LET.
  2. The patient rates their own LET performance on the 1–6 scale, where 6 means fully leaning in and not holding back.
  3. The group leader asks whether the task was relevant for triggering anxiety, and whether the patient was holding back, leaning in, or doing something incompatible with the disorder. [source: Hansen et al. 2018 panic pilot]
  4. Explore any score of 5 or below to find the obstacle. The group discusses how to lean in on the next task, and the leader encourages the patient to "not hold back". [source: Launes et al. 2019, 2020]
  5. Weave repeated psychoeducation into these discussions. [source: Bergen panic study, BMC Psychiatry 2023]
• Example language (illustrative): "You gave yourself a 4. What were you still holding on to? What would make it a 6 this afternoon?"
• Watch for: All patients rated 6 by the end of Day 2 in the panic pilot. Ratings that stay below 6 call for the steps in section 5. Group cohesion may matter: it was a trend-level predictor of outcome in one study. [source: Miegel et al. 2025, abstract]

4.19 "Fake it until you make it"

• Purpose and when: When a patient struggles to lean in wholeheartedly. [source: Hansen et al. 2018 panic pilot]
• Steps:
  1. Encourage the patient to act as if they are fully leaning in, even when they do not feel willing.
  2. Treat behaviour as the target. Willingness can follow action.
• Example language (illustrative): "You don't have to feel ready. Act as if you're someone who isn't holding back, and let your body catch up."
• Watch for: Acting "as if" that turns into a new subtle avoidance, such as performing calm. Check the internal stance.

4.20 Evening self-administered exposure and 9 pm text

• Purpose and when: Evenings of Days 2 and 3. [source: Kvale et al. 2018; Launes et al. 2019; Iversen et al. 2022]
• Steps:
  1. At the end of the day, patient and therapist plan specific evening exposures.
  2. The patient carries them out alone.
  3. Around 9 pm, the patient texts the therapist what they did and their LET rating.
  4. Patients could phone their therapist if they struggled (Iversen et al. 2022).
  5. Day 4 opens with a summary of the evening. [source: Launes et al. 2019]
• Example language (illustrative): "Tonight, do the kitchen tasks we discussed with nobody checking. Text me at nine with what you did and your rating."
• Watch for: Vague texts or texts that seek reassurance. Reply briefly and ask about LET, without reassuring. [synthesis]

4.21 SAD-specific exposures

• Purpose and when: Social anxiety version, Days 2–3. [source: Hansen et al. 2024]
• Steps:
  1. Group exercises: giving presentations, being the centre of attention, speaking assertively.
  2. Individual tasks: mock job interviews, small talk, answering phone calls, shopping scenarios.
  3. Begin with simple tasks done with full LET. Move to harder tasks only once LET is applied correctly, then go promptly to the situations with the most impact on daily life.
  4. After each task, rate LET use and give peer and therapist feedback.
• Example language (illustrative): "Before you start the talk, drop your notes and your rehearsed opening. Speak to the person who looks least friendly."
• Watch for: Rehearsing, self-monitoring and avoiding eye contact. The paper does not mention video feedback or deliberate social-mishap tasks, so do not assume they are part of B4DT for SAD. [uncertain]

4.22 Team rotation and therapist team meetings

• Purpose and when: Throughout Days 2–3. These are therapist-side procedures that shape what the patient experiences. [source: Kvale et al. 2018; Launes et al. 2019; Iversen et al. 2022]
• Steps:
  1. The group leader assigns therapist–patient pairs for each block. Patients do not necessarily keep one therapist.
  2. Hold frequent short team meetings to review each patient's LET and obstacles.
  3. Send the most experienced therapist to the patient who is struggling most.
  4. Colleagues observe each other. The developers' sites hold daily leader meetings, and staff complete adherence forms after each group. [source: Hansen et al. 2024]
• Example language (illustrative, to the patient): "After lunch you'll work with Kari. She's heard where you got stuck."
• Watch for: A therapist who has started reassuring the patient or easing off. Swap pairs or have the leader step in.

4.23 OCD-specific ERP in the B4DT frame

• Purpose and when: OCD version, Days 2–3. [source: Kvale et al. 2018; Hansen et al. 2018; Launes et al. 2019]
• Steps:
  1. Choose tasks from the patient's list in the many OCD-relevant settings where they live. Favour tasks that cross several symptom domains.
  2. Apply "zero tolerance for rituals". This covers overt rituals, reassurance-seeking and mental neutralising.
  3. Deliberately add uncertainty where OCD demands certainty, for example reminding oneself one cannot be sure of being clean or safe. [source: Hansen et al. 2018]
  4. Use LET at each urge (4.12).
• Example language (illustrative): "OCD wants you to be certain. Let's give it the opposite: say out loud that you may never know, and carry on."
• Watch for: Hidden checking, later "undoing" in the head, and asking family for reassurance in the evening. The papers do not give exposure examples by subtype (contamination, checking, harm or taboo thoughts). Build these from standard ERP practice. [uncertain; synthesis]

4.24 Relatives and friends psychoeducation meeting

• Purpose and when: Afternoon of Day 3, about 1–1.5 h, one meeting per group, attended by people the patient invites. [source: Kvale et al. 2018; Launes et al. 2019, 2020; Iversen et al. 2022]
• Steps:
  1. Give a lecture on the disorder and on exposure and LET.
  2. Focus on how relatives can support the patient's change rather than accommodate the disorder. [source: Launes et al. 2019]
  3. Discuss reducing reassurance and accommodation, and how to respond when the patient asks for help with rituals or avoidance. [synthesis: the specific content is not described beyond the focus on supporting change]
• Example language (illustrative): "The most helpful thing you can do is to stop helping the disorder: kindly decline to reassure, and encourage your relative to lean in."
• Watch for: Relatives who feel blamed. Frame accommodation as understandable and changeable. [synthesis] The Bergen videoconference panic study does not mention a relatives' meeting. [uncertain: whether it is used in every version]

4.25 Day 4 consolidation: lessons learnt and becoming one's own therapist

• Purpose and when: Day 4, about 3–4 h. [source: Kvale et al. 2018; Hansen et al. 2018; Launes et al. 2019; Eide et al. 2023]
• Steps:
  1. Open with the previous evening's report.
  2. Summarise the "lessons learnt" and repeat the principles of change and how to do exposures.
  3. Discuss how to keep the change robust and integrate it into daily life.
  4. Cover relapse prevention: when anxiety returns, treat it as a new chance to lean in. [source: Hansen et al. 2018]
  5. Teach the patient to coach themselves with the leader's questions: Is this relevant? Am I holding back? What is incompatible with the disorder?
  6. Close with patient feedback to the group. [source: Launes et al. 2020]
• Example language (illustrative): "From today you're the therapist. When anxiety shows up next month, it's not a relapse, it's your next exposure task."
• Watch for: Patients who expect anxiety never to return. Correct this explicitly.

4.26 Three-week self-exposure plan

• Purpose and when: Written on Day 4 by patient and therapist together. [source: Hansen et al. 2018 panic pilot; Launes et al. 2019; Hansen et al. 2024]
• Steps:
  1. Write a day-by-day plan of exposures for the next three weeks.
  2. Include triggers not yet covered and the settings of everyday life.
  3. Agree daily reporting of LET practice: online in Bergen, or paper reports posted weekly (Launes et al. 2019).
  4. Explain that clinicians will read the reports but will not reply or make contact. [source: Kvale et al. 2018; Hansen et al. 2018]
• Example language (illustrative): "We'll read every report but won't reply. That's on purpose: you're doing this on your own now."
• Watch for: Plans that are vague or limited to easy tasks. Make each day specific.

4.27 Emergency contact information

• Purpose and when: Day 4, given orally and in writing. [source: Kvale et al. 2018; Hansen et al. 2018]
• Steps:
  1. Give clear instructions on how to reach health services in an emergency during the no-contact phase.
  2. Name the circumstances that count as an emergency, such as suicidal thoughts or acute mental or physical health crises, so the no-contact rule is not misread. [synthesis]
• Example language (illustrative): "We won't contact you about exposure, but if you're in crisis, here's exactly who to call."
• Watch for: Patients who think they cannot ask for help at all.

4.28 Three-month booster session

• Purpose and when: About 3 months after treatment, individual, 30 min (OCD) or 45 min (panic). No exposure is done. [source: Kvale et al. 2018; Launes et al. 2019; Iversen et al. 2022]
• Steps:
  1. Discuss the patient's experience since treatment.
  2. Repeat the LET principles and the core treatment components.
  3. Refresh the principles if needed. Do not do exposure in session.
• Example language (illustrative): "Tell me where you've leaned in since we met, and where you notice yourself holding back."
• Watch for: Re-emerging avoidance. Help the patient re-plan self-exposure. [synthesis: what to do beyond reviewing principles is not specified]

4.29 Outcome monitoring

• Purpose and when: Pre-treatment, about one week to 10 days after, and at 3 months. Some studies also measured at 12 months and 4 years. [source: Kvale et al. 2018; Hansen et al. 2018; Hansen et al. 2024]
• Steps:
  1. Use clinician-rated severity: Y-BOCS (OCD) or PDSS (panic), rated by an assessor independent of treatment where possible.
  2. Collect self-reports online (PHQ-9, GAD-7, CSQ-8 satisfaction) with automatic text reminders.
  3. Benchmark the results across clinics; this is part of the quality system. [source: Hansen et al. 2018]
• Example language (illustrative): "An independent colleague will phone you next week to rate your symptoms."
• Watch for: Bias when a group therapist rates their own patient. In one study, the group therapist did the 3-month rating. [source: Bergen panic study, BMC Psychiatry 2023]

4.30 Videoconference delivery (adaptation)

• Purpose and when: The Bergen clinic delivered much of the panic treatment by videoconference during COVID-19. [source: Bergen panic study, BMC Psychiatry 2023]
• Steps: Keep the same four-day structure with remote group meetings and therapist-guided exposure. [uncertain: the operational details of remote coaching]
• Watch for: Covert safety behaviours that are harder to see off camera. Ask about them explicitly. [synthesis]

5. Handling problems

The published papers describe few problems: dropout was near zero and no adverse events are reported. Most guidance below is therefore marked [synthesis], built from B4DT principles.

• Patient holds back (LET rating 5 or below). Explore the specific obstacle in the group report-back. Agree one concrete change for the next task. Give encouragement to "not hold back". [source: Launes et al. 2019, 2020] Simplify the task until LET is a full 6, then build up again. [source: Hansen et al. 2024] Use "fake it until you make it". [source: Hansen et al. 2018 panic pilot]
• Stuck: no anxiety triggered, or tasks are too easy. The task is probably irrelevant or protected by subtle avoidance. Ask the leader's relevance question, look for hidden safety behaviours (4.13), and switch to a trigger the disorder "would appreciate the least". [source: Kvale et al. 2018; Hansen et al. 2018 panic pilot]
• Patient wants to wait until anxiety drops. Restate that anxiety reduction is not a goal. The aim is to lean in, and anxiety is the raw material. [source: Eide et al. 2023; Hansen et al. 2018]
• Struggling patient within the team. The leader reallocates the most experienced therapist to that patient. [source: Kvale et al. 2018; Launes et al. 2019]
• Overwhelm, such as crying, shutdown or wanting to leave. Do not regulate the anxiety down or end the task as a rescue. Acknowledge it, ask about willingness ("Are you willing to let go?" [source: Hansen et al. 2024]), and offer a simpler task that can be done with full LET. If the patient becomes acutely unsafe, dissociated beyond responding, or medically unwell, pause and follow your service's safety and medical procedures. [synthesis]
• Wish to drop out mid-treatment. Bring in the group leader. Revisit the decision the patient made before treatment and their Day 1 commitment, and present leaving as the disorder's preferred option. Respect the patient's decision. [synthesis]
• Non-adherence: missed evening exposures or no 9 pm text. Raise it in the next morning's report-back as data about holding back, plan a smaller and specific evening task, and confirm the text time. [synthesis]
• Non-adherence in the three-week phase. No contact is made during this phase by design. Address it at the booster session and re-plan. [source: Kvale et al. 2018] Some services may add a check-in for high-risk patients, which departs from the protocol. [synthesis]
• Use of benzodiazepines or anxiolytics. Discuss it as a safety behaviour and check the rules agreed with the prescriber. Never advise stopping without prescriber input. [source: Hansen et al. 2018 panic pilot; Iversen et al. 2022] [synthesis: handling]
• Reassurance-seeking aimed at the therapist. Do not reassure. Name it as a ritual or safety behaviour and return to LET. [synthesis, consistent with "zero tolerance for rituals" (Launes et al. 2019)]
• Ruptures, such as anger at the therapist for pushing. Validate the difficulty, restate the shared goal the patient chose, and offer choice over which task comes next, not over whether to lean in. Rotating to another therapist is permitted within the team model. [synthesis]
• Group problems: comparing, competing or demoralisation. Return to individual LET ratings, since every patient can reach 6. Highlight peers' leaning in as models. [synthesis, based on Hansen et al. 2018 panic pilot; group cohesion from Miegel et al. 2025]
• Relatives undermining change. Use the Day 3 meeting to address accommodation, and offer a later conversation if needed. [synthesis]
• Relapse after treatment. Present returning anxiety as a new opportunity to lean in and restart self-exposure. Use the booster session to refresh. [source: Hansen et al. 2018; Kvale et al. 2018]
• Panic: medical worry during interoceptive work. Re-check the medical clearance from 4.1. If symptoms are new or atypical, stop and get a medical assessment. Do not use the "cause a heart attack" framing with anyone who has not been cleared. [synthesis]
• Suicidal ideation appears during treatment. Active suicidality is an exclusion criterion. If it appears, carry out a standard structured risk assessment, safety planning, and lethal-means counselling (safe storage or removal), and decide with the leader whether to continue. [synthesis]
• Therapist distress or wish to rescue. Use the team meetings and peer observation. The developers note that therapists must tolerate patient discomfort and accept being observed. [source: Launes et al. 2019]
• Clinic without certification. Using the B4DT name and format requires training and certification by the Bergen team. A clinician working alone can apply LET in standard weekly ERP, but this is not B4DT. [source: Hansen et al. 2018 (dissemination model)] [synthesis: the weekly application]

Sources

Sources:

• Launes et al. (2020). The Bergen 4-Day Treatment (B4DT) for OCD: outcomes for patients treated after initial waiting list or self-help intervention. Frontiers in Psychology 11:982
• Kvale et al. (2018). Successfully treating 90 patients with OCD in eight days: the Bergen 4-day treatment. BMC Psychiatry 18:323
• Hansen, Hagen, Öst, Solem & Kvale (2018). The Bergen 4-Day OCD Treatment Delivered in a Group Setting: 12-Month Follow-Up. Frontiers in Psychology 9:639
• Hansen, Kvale, Hagen, Hjelle, Solem, Bø & Öst (2018). The Bergen 4-Day Treatment for Panic Disorder: A Pilot Study. Frontiers in Psychology 9:1044
• Launes et al. (2019). The Bergen 4-Day Treatment for OCD: Does It Work in a New Clinical Setting? Frontiers in Psychology 10:1069
• Iversen et al. (2022). The Bergen 4-day treatment for panic disorder: replication and implementation in a new clinic. BMC Psychiatry 22:728
• Eide et al. (2023). The Bergen 4-day treatment for panic disorder: implementation in a rural clinical setting. BMC Psychiatry 23:305
• Bergen panic study (2023), BMC Psychiatry, doi 10.1186/s12888-023-05062-7 (title and authors not shown in the text retrieved)
• Hansen et al. (2024). The Bergen 4-day treatment for social anxiety disorder: a pilot study. BMC Psychiatry 24:145
• Miegel et al. (2025). Exploring the Black Box: What Happens during Brief Concentrated ERP for OCD? Psychotherapy and Psychosomatics 94(3) (abstract only)
• Helse Bergen: The Bergen 4-Day Treatment (B4DT)

Not drawn on: the Bergen repository record (bora.uib.no/handle/1956/21420). The B4DT treatment manual (PowerPoint and text, Kvale & Hansen) is unpublished and was not accessed. [uncertain]