CISD Good Intentions, Great Expectations

1
CISD Good Intentions, Great Expectations
Best Practice

• Doreen Tracey
• Frank McDonald
• Psychologists, Queensland Health,
• Townsville, Australia
• August 2003

2
Overview

• Rationale - current controversy
• Aim of presentation
• Preamble - our position on controversy
• Defining terms
• Past developments steady growth of sound CISD
  theory practice vs. rise of disaster industry
• Recent research - the fall of CISD?
• Future directions - back on firmer ground
  sticking with what we know

3
Rationale

• Current issue some recent widely publicised
  studies suggest one-off, early debriefing may do
  more harm than good (particularly via individual
  sessions with civilians after unexpected trauma
  e.g. medical problems)
• How? May invigorate traumatic memories that
  may be settling naturally (re-traumatising)
  May engender complacency May professionalise
  adversity -gtunder-use of normal networks

4
Rationale

• First do no harm
• Dilemma Do we tell our psychological
  debriefing teams to stay away?
• To do so- seems inhumane counter-intuitive -
  it runs counter to social expectations for
  counselling - may risk legal action

5
Aim

• Stimulate discussion about best practice of
  psychological debriefing

6
Preamble

• Natural tendency to talk things out after
  stressful events
• Arising from humane concern to facilitate this,
  CISD seeks to prevent people internalising
  traumatic stress prevent PTSD
• Obeys social injunction to Get it off your
  chest psychological maxim of better out than
  in
• Meets widespread social/organisational
  expectations for counselling

7
Preamble

• However there are commonplace natural models of
  processing that we should not try to supplant or
  transcend. Nor should we, despite good
  intentions, rush in, push people into one size
  fits all debriefing sessions rush out. We need
  to tailor our inputs

8
Preamble

• CISD does not stand for Critical Incident
  Sheep Dipping!

9
Preamble

• We are sympathetic to view that one-off trauma
  counselling can be a negative experience
• We welcome 2002 study by Dutch team lead by
  Arnold van Emmerik ( others like it) as an
  overdue wake-up call to practitioners public
• Done insensitively, with no option of
  follow-up, without person returning to a team who
  shared trauma or to other environmental supports,
  may open boxes that cant be closed

10
Preamble

• However, no evidence at all that programmatic
  CISD/CISM (vs. uni-dimensional) intervention is
  harmful! Reverse is true, with reduction in human
  suffering well established
• Even then, conceivable situations when one-off
  intervention should not be dismissed as an
  option, solely on basis of meta-analyses RCTs
  (refer to limitations of these evaluating
  psychological therapies) on special populations

11
Preamble

• CISD should be applied sensitively
  comprehensively with guidance from whole of
  research adhering to basic principles of
  counselling psychotherapy - with due regard for
  individual differences e.g. not everyone may need
  to externalise. Not talking may allow use of more
  appropriate defence mechanisms in some

12
Literature Terms

• Critical Incident Stress Debriefing (CISD)
• Critical Incident Stress Management (CISM)
• Defusing session
• (Formal) Debriefing session
• Counselling/psychotherapy
• Terms should not be used interchangeably,
  especially Debriefing (session). Confusion
  results between brief component / stand-alone
  interventions more extensive interventions that
  have sound support

13
Research base to our position

• Past - Original CISD/CISM models that have been
  developing since early 80s
• Present - a closer look at recent controversial
  papers
• Future - complex conflicted area -
  early days/more to know - see
  Conclusions for ideas on future
  research

14
Past Developments

• 1983 CISM (a.k.a. Mitchell Model) developed as
  comprehensive, multi-phasic program response to
  disaster, large scale incident emergency
  service worker needs
• Often misrepresented misunderstood. CISD just
  one interlocking component thats been
  decontextualised (e.g. Snelgrove,1998)
• Further developed into standard of care for
  schools, families, small functional groups,
  organisations communities

15
Past Developments

• Effectiveness empirically validated, through
  qualitative analyses, controlled investigations
  meta-analyses (Flannery, 1998 Everly Mitchell,
  1997 Everly Boyle,1997 Dyregrove, 1997)
• Other group debriefing models Raphael model
  (Raphael,1986) Process Debriefing (Dyregrove,
  1989). Both designed for post-disaster
  intervention

16
Current Research

• Over recent years several trials systematic
  reviews have been carried out to consider the
  evidence for efficacy of single session
  debriefing
• Brisson et al. (1997) conducted randomised trials
  on patients admitted to Cardiff Burns Unit
  found significant increase in PTSD rates in those
  who received debriefing
• Mayou et al. (2000) found road accident victims
  who had been debriefed also reported an adverse
  effect
• Wessely S. (2000) - Systematic review of brief
  psychological interventions. Found these failed
  to show any advantage to debriefing

17
Current Research

• Van Emmerik et al (2002) Meta-analyses of single
  session debriefing within 1 month after trauma,
  using pooled effect size PTSD ratings as
  outcome measure
• 7 studies were included
• Types of trauma
• Burns
• Misc (police officers)
• Road traffic accident x 2 studies
• Early miscarriage
• Violent crime
• Combat exposure

18
Current research

• Interpretation Single session debriefing
  non-debriefing interventions do not improve
  natural recovery from psychological trauma
• Results show that they have no efficacy in
  reducing symptoms of PTSD other trauma-related
  symptoms and in fact suggest that it could have a
  detrimental effect

19
Current research

• Rose S., Bisson, J., Wessely, S. (2003)
  Systematic review using standard Cochrane
  Collaboration methodology
• This review included 11 trials of individuals or
  couples
• Types of trauma
• MVA victims x 3
• relatives of seriously ill/injured
• miscarriage
• MVA, assault or dog bite
• acute burns victims
• mothers following childbirth x 2
• accident emergency presentations
• victims of violence

20
Current research

• Results of this Review
• 3 studies reported positive outcomes
• 6 found no difference between intervention or
  non-intervention
• 2 showed negative outcomes in intervention group
• Conclusion an absence of evidence that
  single-session debriefings are effective

21
Current Research

• Differences between these last 2 studies which
  yielded similar results
• Van Emmerick et al. included studies which
  required both pre post measures did not
  involve randomisation
• 22 studies excluded because they failed to meet
  stringent criteria e.g. intervention consisted of
  gt one session, gt than 1 month since trauma etc.
• Major criteria for inclusion in the Rose et al.
  Study were that studies should have had
  randomised controlled trials subjects had been
  debriefed within 1 month of trauma
• Disparity noted between subjective objective
  findings in Rose study. Subjective views of
  intervention were usually positive
• Only 4 studies were included in both reviews

22
Current Research

• In two earlier RCTs suggesting that debriefing
  may be harmful, (Hobbs, Mayou et al., 1996
  Bisson et al., 1997), both failed to achieve
  equivalent group membership at pre-test
• In both studies debriefed groups had marked
  pre-test differences between them
• The debriefed groups had more severe injuries
• Such factors may have influenced the outcomes. As
  well, the deterioration, though statistically
  significant, was reported to be very slight from
  a clinical perspective

23
Comments on research

• RCTs are the level 1 of good evidence and are
  considered the best most reliable method of
  knowing whether a treatment does more good than
  harm. This rigorous search for evidence of effect
  is now the dominant paradigm of treatment outcome
  studies
• However, other valid means of evaluation of
  treatments include naturalistic studies,
  observational or case series and these have
  implications for ways in which we judge the
  quality and value of research
• Group debriefing in naturalistic settings was the
  original intention for which psychological
  debriefing was designed - as part of an ongoing
  multi-component intervention that included
  preparation for crisis, follow-up and ongoing
  referral as required. (van Emmerick 2002)

24
Comments on research

• These recent papers have highlighted the need for
  caution in applying strategies indiscriminately
• Possible explanations for findings on one-off
  sessions of debriefing
• might interfere with the natural processing of
  trauma event
• might lead to victims bypassing family, friends
  or other social supports
• probably increases awareness/expectancy of normal
  responses to distress
• that these normal reactions should require
  professioonal care may be maladaptive -
  pathologising normal distress
• interference with time to habituate
• general inclusion of victims in debriefing may
  include victims not at risk of chronicity
• quality of delivery
• secondary traumatisation e.g. shame reaction

25
Conclusions

• Need to urgently reassess how we respond to
  trauma
• Publicity given to recent studies about harm
  should be welcomed as wake-up call to
  practitioners participants to adjust their
  expectations about universal quick fixes
• If only doing single-shot intervention, then must
  warn participants of potential harm as well as
  good (especially if conditions match those in
  relevant Cochrane Base RCTs or meta-analyses)

26
Conclusions

• An over-reaction to conclude Do nothing for
  victims. (Sends wrong message traumatised are
  beyond help)
• Rather, simply reminds us No quick means of
  prevention! (Contrary to idea implied by some in
  debriefing industry)

27
Conclusions

• Stand-alone sessions without an appreciation of
  individual circumstances potential needs were
  never intended to be the norm should never be
  considered as such
• Not to say there wont be times when single
  session should be (carefully) considered e.g.
  someone just been through cyclone (not
  advisable)vs.nurse attacked by pt. (possibly
  yes)- Future research needs to discriminate
  populations that can benefit from single sessions

28
Conclusions

• Further question for future research that will
  have heuristic value Exactly how is harm
  done?
• All studies that suggest CISD/CISM is harmful to
  recipients refer to investigations that simply
  did not use comprehensive models prescribed by
  CISD developers (such as Mitchell Model)
• No evidence at all that programmatic CISD/CISM
  (vs. uni-dimensional) intervention is harmful!
  Reverse is true, with reduction in human
  suffering well established

29
Conclusions

• Anyone attempting to aid suffering of trauma
  victims needs to be well-versed well-resourced
  for potentially multi-phasic interventions
• In spirit of earlier point about tailoring
  interventions, perhaps screen for Acute Stress
  Disorder sufferers offer multi-session CBT
  (Bryant et al. 1999) rather than see everyone
• Compulsory debriefing should desist never be
  the rule