Atlas›Domain V · Systemic, relational, group & community›Guide 151
Guide V-151

Critical incident stress debriefing

Jeffrey T. Mitchell1983
A 27 · L 61 · M 1214 interventions in the Trinity explorer9 described for recognition only
Client cardCompare
[source: …] where a claim comes from[uncertain] not confirmed in a primary source[synthesis] the writer's reconstruction

1. Introduction

Version followed: Jeffrey T. Mitchell's seven-phase group Critical Incident Stress Debriefing (CISD), introduced in 1983 and later placed inside the Critical Incident Stress Management (CISM) continuum taught by the International Critical Incident Stress Foundation (ICISF). The details come from ICISF and agency training handouts and from a published law-enforcement review of the Mitchell model (Malcolm et al. 2005). Verification pass (October 2026): the two harm trials were checked against their own abstracts (Bisson et al. 1997; Mayou, Ehlers & Hobbs 2000); WHO's recommendation against debriefing was checked in WHO guideline records; NICE NG116's recommendations were checked against the NICE draft for consultation (June 2018) and a clinical summary of the final guideline (GPnotebook). Mitchell's 1983 paper and the Mitchell & Everly CISD operations manual are sold through ICISF and are not openly published; the phase content below rests on handouts derived from them [uncertain: secondary]. The sections below describe the model as its developers and users publish it. A Harm account before Sources covers the evidence against single-session debriefing. Read that first if you are deciding whether to use CISD at all.

  • What it is. CISD is a structured, single-session, small-group crisis intervention. It was built for close-knit operational groups such as fire, EMS and police crews after a shared critical incident. Mitchell, a former Baltimore Fire Department paramedic, introduced it in 1983 after about nine years of groundwork [source: Utah CISM History 2023]. Its promoters describe it as a crisis-intervention and educational process, not psychotherapy and not a substitute for it [source: Utah CISM History 2023]. A UK occupational provider describes it as a seven-phase, small-group, supportive process meant to strengthen resistance to traumatic stress, build resilience and support return to normal functioning [source: Oldham/PAM Assist].
  • Core idea of change, as proponents state it. The group moves from describing facts (cognitive), to thoughts, to emotional reactions, and then back to the cognitive level through symptoms and teaching. Normalising information is meant to give a "cognitive anchor" and psychological closure [source: GACC CISD Process]. Peer presence and shared accounts are assumed to support recovery [source: Oldham/PAM Assist; Humintell].
  • Status (contested). The Cochrane review found no evidence that single-session individual debriefing prevents PTSD, and said compulsory debriefing should cease [source: Rose et al. 2002 Cochrane]. WHO guidelines recommend against it: psychological debriefing "should not be used in an attempt to reduce the risk of post-traumatic stress" (strong recommendation, very low-quality evidence) [source: BIGG records of WHO 2017 and WHO 2019 guidelines]. The 2019 record marks this recommendation as maintained and notes that it also appears in WHO's 2013 guideline on intimate partner and sexual violence. WHO's psychological first aid (PFA) guide calls debriefing ineffective and not recommended [source: WHO PFA 2011]. NICE advises against psychologically focused debriefing to prevent or treat PTSD [source: March on Stress summary of NICE NG116]. A UK research impact case study lists other bodies that advise against it, including the IASC (2010), the Sphere Project (2011) and the US VA/DoD (2010) [source: REF 2014 impact case study]. Proponents reply that most negative trials tested one-to-one hospital sessions with primary victims, which did not follow the Mitchell group model [source: Malcolm et al. 2005]. The Cochrane review made no recommendation on group debriefing, which is the format Mitchell specified [source: Rose et al. 2002].

2. Framework

CISD is one session inside a wider programme (CISM), not a course of therapy [source: Utah CISM History 2023; ICISF comparison chart].

CISM continuum (as listed by sources). The components are:

  • pre-incident stress education
  • on-scene support
  • peer and significant-other support
  • defusings
  • debriefings
  • follow-up services
  • referral procedures [source: Utah CISM History 2023]

A fuller component list. A 2005 law-enforcement review of the Mitchell model lists ten CISM components, with timing and format [source: Malcolm et al. 2005, Table 1]: strategic planning (before any crisis); demobilizations and staff consultations (shift- or event-driven, large groups); assessment of need (any time after the crisis); defusing (within 12 h, small groups); CISD (1–10 days after, or 3–4 weeks after mass disasters, small groups); individual crisis intervention (any time, one-to-one); family CISM; community and organisational consultation; pastoral crisis intervention; and follow-up referral. The review states that none of the components is meant to be used alone [source: Malcolm et al. 2005]. An ICISF-certified two-day Basic CISM (Group Crisis Intervention) course trains participants in demobilizations, defusings and CISD [source: Shift Wellness course page]. No procedural detail for strategic planning, family CISM or pastoral crisis intervention was found [uncertain].

ICISF's chart compares four group or individual interventions, summarised below [source: ICISF comparison chart; Humintell].

InterventionWhoWhenLengthTeam
Crisis Management Briefing (CMB)Mixed (heterogeneous) groupsWhenever new information is available20–30 minAn administrator plus a crisis-team member
DefusingHomogeneous groupWithin about 8 h20–45 min2+ trained team members
CISDHomogeneous, close-knit group1–3 days after the event (24–72 h described as ideal)1–3 h (about 2 h typical)2+ trained members, at least one a mental-health professional, plus trained peers
Individual crisis interventionOne personAny time15–20 min1 trained team member

Timing differs by version [uncertain]. ICISF, Utah and Humintell sources give 24–72 h for CISD. Malcolm et al. (2005) give 1–10 days, or 3–4 weeks after mass disasters, and defusing within 12 h (the ICISF chart says within about 8 h). A UK provider says CISD is recommended at least 72 h after the incident, with on-site defusing within 24–48 h [source: Oldham/PAM Assist]. In the Oxford road-traffic trial, the individual sessions were mostly given within 24–48 h [source: Mayou et al. 2000]. The figure in Mitchell's own manual could not be checked. If you follow the ICISF version, use its chart's 24–72 h, and never run a debrief while the incident is still operational [synthesis].

The seven phases (one session) [source: GACC CISD Process; Oldham/PAM Assist]. The phase names (Introduction, Fact, Thought, Reaction, Symptom, Teaching, Re-entry) are confirmed by Malcolm et al. (2005), who describe Re-entry in police use as aimed at returning the officer to duty:

  1. Introduction. Introduce the team, explain the process, set expectations and ground rules. Participation is voluntary and the discussion is confidential.
  2. Fact. Each participant describes the event from their own point of view, at a cognitive level. The group builds a timeline.
  3. Thought. Participants give their first or most prominent thoughts. This phase is the bridge toward emotional content.
  4. Reaction. Participants identify the most traumatic aspect for them and the emotional reactions it caused. The UK provider calls this the core phase.
  5. Symptom. Participants identify personal signs of distress (physical, emotional, behavioural). The group moves back to the cognitive level.
  6. Teaching. The team explains normal reactions and adaptive coping, including stress management, and gives a cognitive anchor.
  7. Re-entry. The team clarifies ambiguities, answers questions, summarises and prepares the group for closing. Handouts may be given.

After the session, refreshments are served and one-to-one contacts often follow. Follow-up and referral are part of the process [source: Oldham/PAM Assist; Utah].

Moving between phases. No published phase-exit criteria were found. One handout gives a single rule: go around the room to each person for the first two steps only [source: GACC CISD Process]. The movement is cognitive, then affective, then cognitive. The team moves on once the phase goal has been met by those who choose to speak [synthesis].

3. Guiding principles

These are the principles the developers and users publish, with practical meaning. Additions are marked.

  • Crisis intervention, not psychotherapy [source: Utah]. Do not treat the session as treatment. Anyone with significant symptoms is referred to assessment and treatment.
  • Homogeneous, close-knit groups only [source: ICISF chart; Humintell]. Include only people with the same role and comparable exposure, such as one crew. Mixed groups get a briefing (CMB), not a CISD.
  • Voluntary participation and confidentiality [source: Oldham/PAM Assist]. No one is required to attend or to speak. What is said stays in the room, and supervisors do not use it for performance or investigation purposes.
  • Right technique, right time, trained personnel [source: ICISF chart]. Match the intervention to the timing and the group. The techniques are not mutually exclusive.
  • Peer plus professional team [source: Utah; ICISF chart; Malcolm et al. 2005]. At least one mental-health professional, plus trained peers who know the work culture. In police teams this means at least one peer officer and at least one mental-health professional.
  • Structured sequence, cognitive then affective then cognitive [source: GACC]. Do not end the session while participants are in heightened emotion. Close on education and practical next steps.
  • "CISD may not be for everyone" [source: Utah]. Expect that some people will not benefit. Offer other routes.
  • Follow-up and referral are part of the process [source: Utah; Oldham/PAM Assist]. Make contact again after the session, and know your referral pathways before you start.
  • Do no harm: never compulsory, never the only provision, never for highly distressed primary victims [synthesis from Rose et al. 2002; Del Mar 2002; WHO PFA 2011]. Offer PFA and active monitoring as the default. Screen people and refer to evidence-based treatment those whose symptoms persist.

4. Intervention inventory

The published handouts give only each phase's goal and very few prompts. All step detail and example wording below is illustrative [synthesis] unless a source is tagged. Risk flags mark points where the harm evidence applies (see Harm account).

4.1 Pre-session assessment and group selection

  • Purpose. Decide whether a group intervention is appropriate at all, and for whom.
  • Steps.
    1. Talk with the requesting manager about the incident, who was involved and each person's exposure.
    2. Confirm the group is homogeneous: same unit and role, comparable exposure [source: ICISF chart; Humintell].
    3. Confirm the incident is over and the group is not still operational at the scene [synthesis].
    4. Send mixed audiences to a CMB instead [source: ICISF chart].
    5. Arrange a private room, cover for staff so they can attend, no media, and two or more facilitators [synthesis].
    6. Agree with management in writing that attendance is voluntary and that content will not be reported back [synthesis; source: Oldham/PAM Assist on voluntariness].
    7. Identify anyone directly injured, bereaved, or showing severe acute distress. Offer them individual PFA and assessment instead of the group [synthesis from Del Mar 2002 and Rose et al. 2002].
  • Example language (illustrative). "Before we plan anything, I'd like to know who was there and what each person was exposed to. Attendance needs to be optional, and nothing said in the room will come back to management."
  • Watch for.
    • Risk flag: management wants everyone "put through" a debrief. Compulsory debriefing is advised against [source: Rose et al. 2002].
    • Groups whose members had very different exposure levels.
    • Ongoing disciplinary or legal investigations.

4.2 Introduction phase

  • Purpose. Introduce the team, explain the process and set expectations [source: GACC]. Make voluntariness and confidentiality explicit [source: Oldham/PAM Assist].
  • Steps.
    1. Each team member gives their name and role.
    2. Explain that this is a structured discussion, not therapy and not an operational review [source: Utah on not psychotherapy].
    3. State the ground rules [synthesis]: speaking is optional, everyone speaks for themselves, no rank in the room, no notes or recordings, confidentiality and its limits (risk to self or others), and people may step out with a team member.
    4. Say how long the session lasts and that the team stays afterwards.
  • Example language (illustrative). "You don't have to say anything today. If you prefer just to listen, that's fine. What's said here stays here. The one exception is if someone is at risk, and then we'd talk with that person first."
  • Watch for.
    • Body language that suggests people were told to attend.
    • Supervisors present in a way that inhibits speech.
    • If people were told to attend, re-state the right to leave without consequences [synthesis].

4.3 Fact phase

  • Purpose. Each participant describes the event from their own vantage point at a factual, cognitive level. The group builds a shared timeline [source: GACC; Oldham/PAM Assist].
  • Steps.
    1. Go around the room once [source: GACC].
    2. Ask for role, where they were and what they saw or did, briefly.
    3. Allow people to pass.
    4. Keep accounts factual. Gently redirect requests for graphic detail [synthesis].
  • Example language (illustrative). "Going round once: tell us who you are, what your role was, and in a sentence or two what you saw or did. Passing is fine."
  • Watch for.
    • Risk flag: listeners are exposed to details they did not witness. People who were less exposed may take on others' images [synthesis; mechanism untested, uncertain].
    • Keep the phase brief and non-graphic. Stop any one person going into prolonged sensory detail.

4.4 Thought phase

  • Purpose. Elicit first or most prominent thoughts. This is the bridge from cognition to emotion [source: GACC; Oldham/PAM Assist].
  • Steps.
    1. Ask an open question to the whole group. No go-round is required [source: GACC: the go-round applies to the first two phases only].
    2. Accept short answers.
  • Example language (illustrative). "What was your first thought when you realised what was happening, or once you were off autopilot?"
  • Watch for.
    • Self-blame statements ("I should have…"). Note them for the teaching phase and for individual follow-up [synthesis].

4.5 Reaction phase

  • Purpose. Identify the most traumatic aspect of the event for participants and the emotional reactions it caused. The UK provider calls this the core phase [source: GACC; Oldham/PAM Assist].
  • Steps.
    1. Ask about the worst part for each person who wishes to answer.
    2. Let silence happen.
    3. Do not press individuals.
    4. Have a second team member track anyone who becomes overwhelmed and quietly offer support or a break outside [synthesis].
  • Example language (illustrative). "For those who want to say, what has been the hardest part of this for you?"
  • Watch for.
    • Risk flag: this is the phase where people are encouraged to recount and express emotion. Del Mar raises concern that encouraging people to relive the event may be harmful [source: Del Mar 2002]. WHO PFA describes debriefing's emotional ventilation as ineffective [source: WHO PFA 2011].
    • Dissociation, panic, intense guilt, or a participant who goes silent and withdrawn.
    • Never require anyone to speak. Move toward the cognitive phases if arousal is escalating [synthesis].

4.6 Symptom phase

  • Purpose. Identify personal signs of distress (physical, emotional, behavioural) and move the group back to the cognitive level. The symptoms raised become the starting point for teaching [source: GACC; Oldham/PAM Assist].
  • Steps.
    1. Ask what people have noticed in themselves since the incident.
    2. List common categories neutrally: sleep, appetite, irritability, intrusive images, concentration, alcohol use.
    3. Note which ones are raised.
  • Example language (illustrative). "Since it happened, what have you noticed in yourselves? Things like sleep, concentration, irritability, or the scene coming back to mind?"
  • Watch for.
    • Persistent re-experiencing, avoidance, numbing, heavy drinking, hopelessness or suicidal thoughts. Each needs an individual conversation and risk assessment after the session, not discussion in the group [synthesis].

4.7 Teaching phase

  • Purpose. Explain normal stress reactions and adaptive coping, including stress management. Give a cognitive anchor [source: GACC; Oldham/PAM Assist].
  • Steps.
    1. Link back to the symptoms people named.
    2. Explain that most people recover over time, especially when basic needs are met and support is available [source: WHO PFA 2011 on natural recovery].
    3. Give practical coping steps: rest, regular eating, contact with trusted people, limiting alcohol and caffeine [source: WHO PFA 2011 helper self-care].
    4. Name the warning signs that mean someone should seek help. WHO PFA uses persistence beyond about one month [source: WHO PFA 2011].
  • Example language (illustrative). "What you've described is very common in the first days and usually eases over the next few weeks. If sleep problems, upsetting memories or heavier drinking are still there after about a month, please get in touch with us."
  • Watch for.
    • Do not suggest that symptoms are inevitable or that this session will prevent PTSD. The evidence does not support that [source: Rose et al. 2002].
    • Avoid over-medicalising normal reactions [synthesis].

4.8 Re-entry phase

  • Purpose. Clarify ambiguities, answer questions, summarise and prepare for closure. Handouts may be given [source: GACC; Oldham/PAM Assist].
  • Steps.
    1. Invite questions.
    2. Correct rumours with known facts only.
    3. Summarise themes without naming individuals.
    4. Give written information with contact routes for follow-up and professional help.
    5. Each team member gives a brief closing comment [synthesis].
  • Example language (illustrative). "Before we finish: is anything about what happened still unclear? The team will be here for a while afterwards, and the sheet has the numbers to call over the coming weeks."
  • Watch for.
    • People leaving abruptly.
    • Anyone who has not spoken but appeared distressed. Make sure a team member checks in with them [synthesis].

4.9 Post-session contact, follow-up and referral

  • Purpose. Informal one-to-one contact, monitoring and referral [source: Oldham/PAM Assist on refreshments and one-to-one sessions; Utah on follow-up and referral].
  • Steps.
    1. Serve refreshments and let team members circulate.
    2. Hold brief one-to-one contacts with people flagged during the session.
    3. Hold a team meeting afterwards to identify follow-up needs [synthesis].
    4. Recontact the group or individuals within days to weeks.
    5. Refer anyone with persisting symptoms for assessment and, if indicated, trauma-focused treatment. The UK provider names CBT and EMDR [source: Oldham/PAM Assist].
  • Example language (illustrative). "I noticed today was hard for you. Could I check in with you on Thursday to see how sleep is going?"
  • Watch for. This step does the most to protect people [synthesis]. It is consistent with NICE's active monitoring for people within a month of a traumatic event who do not have clinically important symptoms [source: NICE NG116 draft for consultation 2018, rec. 1.6.3; March on Stress summary of the final guideline].

4.10 Defusing (CISM component)

  • Purpose. A shorter, earlier contact with a small homogeneous group, within about 8 hours (ICISF chart) or within 12 hours (Malcolm et al. 2005), lasting 20–45 minutes and led by two or more trained team members [source: ICISF chart; Humintell; Malcolm et al. 2005]. It is taught in ICISF's Basic CISM (Group Crisis Intervention) course [source: Shift Wellness course page].
  • Steps. No published internal structure was found: Malcolm et al. (2005) and an ICISF-certified course page give only timing and format [uncertain]. Practical adaptation [synthesis]:
    1. Check basic needs (rest, food, transport home).
    2. Give brief normalising information and a contact sheet.
    3. Do not elicit detailed accounts.
  • Example language (illustrative). "Before you head home: you may notice some reactions over the next day or two. Here's what's normal and how to reach us."
  • Watch for. An early contact focused on safety and information is close to PFA. Steer it that way [synthesis; source: WHO PFA 2011].

4.11 Crisis Management Briefing (CMB)

  • Purpose. An information-giving session for large or mixed (heterogeneous) groups whenever new information becomes available. It lasts 20–30 minutes and is co-led by an administrator and a crisis-team member [source: ICISF chart].
  • Steps. No published detailed structure was found; Malcolm et al. (2005) do not list the CMB by name [uncertain]. Practical adaptation [synthesis]:
    1. The administrator gives verified facts.
    2. The clinician describes common reactions and coping.
    3. Explain how to access support.
    4. Take questions.
    5. Invite no personal disclosure.
  • Example language (illustrative). "Here is what we know so far, what we don't yet know, and where support is available."
  • Watch for. Rumours. Do not let a CMB turn into a group debrief of a mixed audience [synthesis].

4.12 Individual crisis intervention

  • Purpose. One-to-one support at any time, about 15–20 minutes, by one trained team member [source: ICISF chart; Malcolm et al. 2005 on timing]. ICISF teaches a named individual crisis model in its own courses; its steps are not openly published [uncertain].
  • Steps. Use PFA's Look, Listen, Link structure as the safer evidence-informed format [source: WHO PFA 2011]:
    1. Check safety and urgent needs.
    2. Listen without pressuring disclosure.
    3. Help the person sort priorities.
    4. Connect them with practical help and their own supports.
    5. Refer urgently anyone at risk of harming self or others.
  • Example language (illustrative). "You don't need to tell me what happened unless you want to. What would help most right now?"
  • Watch for. Grounding needs: WHO PFA suggests slow breathing, feeling the feet on the floor, or naming things seen or heard [source: WHO PFA 2011].

4.13 Pre-incident stress education

  • Purpose. Teach staff in high-exposure roles about stress reactions, coping and support routes before incidents happen [source: Utah CISM component list; course content not published, uncertain].
  • Steps (illustrative) [synthesis].
    1. Run a short training on common reactions, self-care and buddy systems [source: WHO PFA helper self-care].
    2. Explain when and how to seek help.
    3. Explain that support after incidents is optional.
  • Example language (illustrative). "After a bad job, most people feel rough for a while and then recover. Here's how to look after yourself and each other, and when to call us."
  • Watch for. Content that implies everyone must be debriefed [synthesis].

4.14 Demobilization (CISM component)

  • Purpose. A contact for large groups after large-scale incidents, timed to the end of a shift or of the event. Malcolm et al. list it as "demobilizations and staff consultations" [source: Malcolm et al. 2005; Shift Wellness course page]. It is taught alongside defusing and CISD in ICISF's Basic CISM course [source: Shift Wellness course page].
  • Steps. No published internal structure was found [uncertain]. Practical adaptation [synthesis]:
    • Meet crews as they come off shift, before they go home.
    • Offer food, drink and a chance to rest.
    • Give brief information on common reactions and on how to get support.
    • Do not ask for accounts of the incident or for emotional disclosure.
    • Note anyone who is visibly struggling, for individual follow-up.
  • Example language (illustrative). "You've had a long shift. Get something to eat. This sheet covers what people often notice over the next few days, and who to call."
  • Watch for. Keep it brief and informational. A demobilization that turns into a group recounting by a mixed crowd carries the same risks as debriefing a heterogeneous group [synthesis].

5. Handling problems

The published CISD sources give almost no problem-handling guidance. Everything below is [synthesis] unless tagged.

  • Employer requests a compulsory or "everyone must attend" debrief. Decline the compulsory element. Explain that the Cochrane review concluded compulsory debriefing should cease [source: Rose et al. 2002]. Offer instead:
    • a CMB for information
    • PFA available on request
    • active monitoring with a check-in at about one month [source: WHO PFA 2011; March on Stress/NICE NG116]
    • a clear referral route for treatment
  • Heterogeneous or mixed-exposure group requested. Do not run a CISD. Use a CMB for the whole group, and offer individual contact to the most exposed [source: ICISF chart on CMB for heterogeneous groups].
  • Primary victims, such as injured survivors or bereaved relatives. Do not offer single-session debriefing. The trials showing harm were in this population [source: Bisson et al. 1997; Mayou et al. 2000; Rose et al. 2002]. Use PFA, practical help and monitoring.
  • Participant overwhelmed during the Reaction phase (sobbing, dissociation, panic). A second team member offers a quiet break outside and uses grounding (slow breathing, feet on the floor, naming surroundings) [source: WHO PFA 2011 grounding]. Lead facilitator moves the group toward the cognitive phases. After the session, do an individual risk assessment and follow-up.
  • Silent or reluctant participants. Respect silence; PFA explicitly warns against pressuring disclosure [source: WHO PFA 2011]. Check in individually afterwards, briefly and without pressure.
  • High initial distress. In the road-traffic trial, people with high initial intrusion and avoidance scores improved without debriefing but stayed symptomatic with it [source: Mayou et al. 2000]. Steer highly symptomatic people away from group recounting and toward individual assessment [synthesis].
  • Disclosure of suicidal thoughts, or of risk to others. Follow the confidentiality limits stated in the Introduction phase. See the person privately the same day for a standard risk assessment and safety planning, and refer urgently [source: WHO PFA 2011 urgent referral criteria].
  • Anger at management, blame of a colleague, or operational criticism. Name it as understandable. Say that the session is not an operational review or investigation and cannot change findings. Signpost the proper channel and protect named individuals from group blame.
  • Legal or disciplinary process under way. Postpone the group, or limit it to education. Discussion of facts in a group could affect witness accounts and expose participants. Get local legal and HR advice.
  • Repeat incidents, or a culture of routine debriefing after every job. Shift resources to pre-incident education, peer support, monitoring and access to treatment [source: Rose et al. 2002 on resourcing identification and treatment].
  • Client or participant reports worsening after a debrief. Validate the report. Assess for PTSD, depression and substance misuse. Offer evidence-based treatment, and record and report the adverse effect to the service.
  • Rupture with the team ("you made it worse"). Acknowledge it without defensiveness, and apologise for any pressure to talk. Offer an individual contact with a different clinician if preferred.
  • Clinician's own distress. Use buddy systems, breaks and supervision. Seek help if symptoms persist beyond about a month [source: WHO PFA 2011 helper self-care].

Harm account

What was actually done

  • Mitchell-model CISD in emergency services, 1983 onward. Crews were brought together after a critical incident, typically 24–72 hours later, for a single session of about two hours [source: Utah CISM History 2023; ICISF chart]. A team of a mental-health professional and peers led them through the seven phases. Each person went round the room giving their account. They then voiced their thoughts, the worst part and their emotional reactions, before receiving stress education [source: GACC]. Use spread through state teams. In Utah, a team was formed after the 1987 Kearns mid-air collision and ran its first debriefing that November, after a police shooting. That session drew about 60 people, many of whom, by the team's own account, should not have been there [source: Utah CISM History 2023].
  • Spread to primary victims and individual sessions. The randomised trials in the Cochrane review tested single-session debriefing: recalling and processing the trauma, usually one-to-one, with adults exposed within the previous month [source: Rose et al. 2002]. Two examples:
    • Oxford road-traffic trial (Hobbs 1996; Mayou, Ehlers & Hobbs 2000). 106 consecutive road-traffic accident admissions aged 16–65 at the John Radcliffe Hospital were randomised. The intervention was one individual session of about an hour, led by a research worker and given mostly within 24–48 hours of the accident. It covered a detailed review of the accident, encouragement of emotional expression and cognitive appraisal, and patients were given a leaflet. Follow-up was at 4 months and 3 years [source: Mayou et al. 2000 abstract]. Del Mar adds that the session included advice on who could help, including the GP [source: Del Mar 2002].
    • Burns trial (Bisson, Jenkins, Alexander & Bannister 1997). 133 adults with acute burn trauma completed an initial questionnaire and were then randomised to individual or couple debriefing or to no intervention. Assessors blind to group interviewed 110 (83%) at 3 and 13 months [source: Bisson et al. 1997 abstract]. The abstract does not state how long after the injury the session was given [uncertain].
    • Other uses. Midwife-led debriefing after operative childbirth was trialled [source: Del Mar 2002, citing Small 2000]. Counsellors were flown in to offer similar sessions after the Port Arthur mass shooting [source: Del Mar 2002].
  • Compulsion. The Cochrane authors' call to stop compulsory debriefing shows that it was being imposed on trauma victims and staff [source: Rose et al. 2002]. How this was enforced, for example by roster or by employer policy, is not documented in the cited sources [uncertain].

How it caused harm

  • No benefit. Across trials, debriefing did not reduce PTSD at up to 3, 3–6 or 6–12 months. It gave no benefit for depression, anxiety, general morbidity or functioning, and did not outperform simple education [source: Rose et al. 2002].
  • Higher PTSD at 13 months. In the burns trial, 16 people in the debriefed group (26%) had PTSD at 13 months, against 4 controls (9%) [source: Bisson et al. 1997]. The Cochrane review gives an odds ratio of 2.51 (95% CI 1.24–5.09) for PTSD at 12+ months [source: Rose et al. 2002]. Caveats: the debriefed group had higher initial questionnaire scores and more severe burns, both linked to poorer outcome, and loss to follow-up differed between groups [source: Bisson et al. 1997; Rose et al. 2002]. The authors said the findings "seriously question the wisdom of advocating one-off interventions post-trauma" [source: Bisson et al. 1997].
  • Worse outcomes at 4 months and 3 years. In the road-traffic trial, intrusion and avoidance scores (Impact of Event Scale) rose slightly in the debriefed group but fell in controls at 4 months [source: Del Mar 2002]. At 3 years, 61 of the 106 were followed up. The overall difference in intrusion and avoidance was not significant. Among people with high initial scores, controls improved while debriefed patients stayed symptomatic. The debriefed group had worse general psychiatric symptoms, pain, physical problems, everyday functioning and financial problems, and debriefed passengers enjoyed travelling less. The authors concluded that debriefing "is ineffective and has adverse long-term effects" [source: Mayou et al. 2000]. Caveats: the groups were imbalanced at baseline, and attrition was high [source: Del Mar 2002; Rose et al. 2002].
  • Other trials. Midwife-led debriefing showed a trend toward more harm than good [source: Del Mar 2002]. More participants dropped out of debriefing arms [source: Rose et al. 2002].
  • Mechanisms.
    • Encouraging people to relive and voice the event early may be harmful, and may disrupt protective distancing or "denial" [source: Del Mar 2002].
    • Debriefing's focus on detailed recounting and emotional release does not target the factors linked to recovery: safety, calm, connectedness, hope and practical support [source: WHO PFA 2011].
    • Further mechanisms, untested [synthesis; uncertain]:
      • group members are exposed to each other's worst details
      • normal reactions are pathologised
      • a one-off session replaces natural supports and real follow-up
      • organisations treat a debrief as having discharged their duty of care
  • Deaths. No deaths attributable to debriefing are documented in the cited sources.
  • Most at risk. People with high early intrusion and avoidance symptoms [source: Mayou et al. 2000]. Injured primary victims, such as burns and accident patients [source: Bisson et al. 1997; Mayou et al. 2000; Rose et al. 2002]. Anyone made to attend [synthesis].

Recognising it today

These features are [synthesis] unless tagged.

  • Policy phrases. "Mandatory debrief", "all staff involved must attend", "CISD within 72 hours", "get it all out", "talk through what you saw and felt" as a standard response to every incident.
  • Outsourced "trauma debriefs". Single-session debriefs run by employee-assistance or occupational providers, sometimes advertised as building resilience [source: Oldham/PAM Assist wording].
  • Rebrandings. "Psychological debrief", "trauma debrief", "critical incident debrief" and "hot debrief" may be used for emotional-processing sessions. NICE's term is "psychologically focused debriefing" [source: March on Stress/NICE NG116].
  • Red flags in a colleague's or service's practice.
    • Mixed-exposure groups.
    • Primary victims or bereaved people in groups.
    • Pressure to speak.
    • A go-round asking for the "worst part".
    • No screening, follow-up or referral pathway.
    • Claims that the session prevents PTSD.
  • How legitimate practices differ.
    • Operational or technical debriefs review what happened for learning and are distinct. WHO PFA explicitly separates routine operational debriefing of aid workers [source: WHO PFA 2011].
    • PFA does not ask for a detailed account [source: WHO PFA 2011].
    • Trauma-focused therapy uses planned, consented exposure within a course of treatment after assessment, not a single group session.

What to do instead and how to help those harmed

  • Immediately: psychological first aid.
    • Look: safety, urgent needs, serious distress.
    • Listen: ask about needs, do not pressure anyone to talk, help them feel calm.
    • Link: practical help, information, loved ones and supports.
    • Refer urgently anyone seriously injured, unable to care for self or children, or at risk to self or others [source: WHO PFA 2011]. WHO's mhGAP guideline group concluded that PFA, not debriefing, should be offered to people in severe distress [source: quoted in WHO PFA 2011]. WHO's 2013 guidelines on conditions specifically related to stress, which underpin the mhGAP stress module, are known only from a publication page [uncertain]. WHO's later guidelines pair the recommendation against debriefing with psychosocial support: psycho-education, help with stress management and support for daily functioning [source: BIGG record of WHO 2019].
  • Weeks 1–4: active monitoring. Do not give routine psychological intervention, and do not offer psychologically-focused debriefing to prevent or treat PTSD. Check in proactively and screen people who are struggling [source: NICE NG116, as given in the 2018 draft for consultation (recs 1.6.3, 1.6.19), GPnotebook's summary of the final guideline and March on Stress]. The Cochrane authors favoured a "screen and treat" model [source: Rose et al. 2002].
  • Persisting symptoms or PTSD. Give evidence-based trauma-focused treatment. For adults presenting more than a month after the event, NICE NG116 recommends an individual trauma-focused CBT intervention (for example cognitive processing therapy, cognitive therapy for PTSD, narrative exposure therapy or prolonged exposure), usually over 8–12 sessions; for non-combat trauma, EMDR is offered after 3 months and considered at 1–3 months if the person prefers it [source: GPnotebook summary of NICE NG116; Oldham/PAM Assist].
  • Assessing someone who went through a debrief [synthesis].
    1. Ask whether attendance was voluntary, what was asked of them, and how they felt afterwards.
    2. Screen for PTSD, depression, substance misuse and suicidality.
    3. Distinguish symptoms from the incident from any added distress caused by the session, such as others' details now intruding or shame at being made to speak.
    4. Restore control: explain what they will and will not be asked to recount, and get informed consent for any exposure-based work.
  • Ethical, reporting and legal duties [synthesis; jurisdiction-specific, check local rules].
    • Do not deliver or endorse compulsory debriefing.
    • Give organisations evidence-based advice in writing.
    • Obtain informed consent that includes the known lack of benefit.
    • Record and report adverse effects through your service's incident system.
    • Raise concerns about a colleague's or provider's harmful practice through clinical governance or your professional regulator.
    • Follow mandatory reporting duties for risk to self or others as usual.

Sources

Main sources

Added in the verification pass (October 2026)

Further references (cited, not the basis of specific steps) [uncertain]

  • Mitchell, J.T. (1983). When disaster strikes: the critical incident stress debriefing process. Journal of Emergency Medical Services.
  • Mitchell, J.T. & Everly, G.S. Critical Incident Stress Debriefing: An Operations Manual (ICISF).
  • NICE (2018). Post-traumatic stress disorder (NG116), https://www.nice.org.uk/guidance/ng116/chapter/recommendations. Checked here via the NICE draft for consultation, June 2018 and GPnotebook, CBT in post-traumatic stress disorder; recommendation numbers are the draft's [uncertain].
  • WHO (2013). Guidelines for the management of conditions specifically related to stress (publication page listed above).
  • Hobbs et al. (1996), BMJ (road-traffic trial, 4-month results), as reported by Del Mar (2002) and Rose et al. (2002).
  • Schimelpfenig T. (2005), NOLS review of critical incident stress debriefing (background only).
← Previous · V-150Psychological first aidNext · V-152 →Mental Health First Aid