Ready! Fire! Aim! The Status of Psychological Debriefing and ...

Ready! Fire! Aim! The Status of Psychological Debriefing andTherapeutic Interventions: In the Work Place and After Disasters

Grant J. DevillySwinburne University

Richard GistKansas City, Missouri Fire Department and

University of Missouri-Kansas City

Peter CottonSwinburne University and Health Services Australia

Psychological debriefing (PD) is a brief, short-term intervention aimed at mitigatinglong-term distress and preventing the emergence of posttraumatic stress. In recentyears, it has become a ubiquitous intervention, one which has evolved as almostprescriptive following harrowing events and grew through a practical need to offerassistance to those who are exposed to severe trauma. Despite disturbing data from therecent refereed literature of psychology, it is still referred to as the “standard of care”for disaster and crisis response and its use in many quarters continues. This articlecritically reviews the evidence for and against its use and outlines the weaknesses in theresearch. The emphasis of this review is on the appropriateness of debriefing inorganizations. This article also proposes a set of hypothesized constructs that may, inpart, be responsible for the paradoxical effects found in some outcome studies ondebriefing. Guidelines are also proposed to help organizations and professionals reactappropriately using evidence-based interventions.

Keywords: debriefing, CISD, early intervention, trauma, resilience

Psychological debriefing (PD) has become awidespread and expected intervention followingexposure to trauma. Employers, aid organiza-tions, and other authorities frequently default toorchestrating “debriefing” services, both to helpmitigate the psychological consequences ofthese disruptions and to meet duty of care re-quirements under workplace health and safetylaws. This article reviews currently availableempirical data to assess the wisdom of provid-ing such services using existing models of in-tervention and to offer hypotheses regarding thefindings reported. We then recommend guide-lines for organizational and clinical interven-

tions that we suggest are more likely to meetstandards of empirically supported practice. Wefinally discuss in some depth the implications ofthe “debriefing debates” for both academic andapplied psychology.

Traumatic events are considered endemic inpsychiatric populations (McFarlane, Bookless,& Air, 2001) and have been reported to increasethe likelihood of psychological dysfunction(Brickman, Garrity, & Shaw, 2002; Kessler,Sonnega, Bromet, Hughes, & Nelson, 1995).Those who develop posttraumatic stress disor-der (PTSD) show high rates of comorbidity(Creamer, Burgess, & McFarlane, 2001) andhave demonstrated higher utilization of healthcare services (Kessler et al., 1999). The lifetimeprevalence rate for significant traumatic lifeevents (e.g., rape, assault, natural disaster, wit-nessing murder, etc.) has been estimatedat 60.7% for men and 51.2% for women withlifetime prevalence of PTSD (using Diagnosticand Statistical Manuel of Mental Disorders–III–Revised [DSM–III–R] criteria) estimatedat 7.8%—clearly far below the rate of exposure

Grant J. Devilly, Brain Sciences Institute, SwinburneUniversity; Richard Gist, Kansas City, Missouri Fire De-partment and University of Missouri-Kansas City; PeterCotton, Brain Sciences Institute, Swinburne University andHealth Services Australia.

Correspondence concerning this article should be ad-dressed to Grant J. Devilly, Director of Clinical & ForensicResearch Unit, Brain Sciences Institute, Swinburne Univer-sity, P.O. Box 218, Hawthorn, Vic 3122, Australia. E-mail:[email protected]

Review of General Psychology Copyright 2006 by the American Psychological Association2006, Vol. 10, No. 4, 318–345 1089-2680/06/$12.00 DOI: 10.1037/1089-2680.10.4.318

318

(Kessler et al., 1995). Indeed, epidemiologicdata following the terrorist attacks in New YorkCity found probable PTSD in approxi-mately 7.5% of those exposed (and 9.7% metcriteria for current depression—i.e., withinlast 30 days), though that rate was more thandoubled for those in greatest proximity (Galeaet al, 2002). However, that the 7.5% incidencehad resolved to 0.6% six months after the firstwave of data collection (Galea et al, 2003)strongly suggests that exposure alone is insuf-ficient to stimulate PTSD in a substantial ma-jority of cases and that many early manifesta-tions spontaneously resolve without orches-trated intervention.

The likelihood of PTSD appears moderatedby such trauma-specific variables as personalinvolvement with the traumatic event; by eventcharacteristics such as whether the event was ofnatural, technological, or volitional origin; andby both degree and proximity of one’s exposureto the event and its sequelae. The likelihood ofpathological outcomes has also been shown tobe affected by person-specific factors such associoeconomic status, coping styles, and bothlevel and quality of perceived social support(Norris, Kaniasty, & Thompson, 1997). Currentevidence is somewhat inconsistent regardingwhich coping styles (e.g., practical vs. emo-tional) may prove most advantageous at partic-ular intervals following traumatization, al-though Norris (2001) has noted that minimizing(or distancing) the event appears to be adaptivewhile avoidant coping strategies and the assign-ment of blame have consistently been related topoorer outcomes.

Kessler et al. (1995) also noted that the rateof PTSD was higher among women (10.4%)than among men (5.0%) and was higher amongthe previously married. Australian data from a12-month prevalence study (Creamer et al.,2001) have replicated the finding regardingmarital status yet found a much smaller differ-ence regarding gender. The Australian data alsoappeared to reflect a lower 12 month prevalencerate of PTSD overall (1.33%) than did reason-ably contemporaneous U.S. data (3.9%; Kessleret al., 1999). Considering that the Australiandata indicated a slightly higher incidence oftraumatic exposure, this could be hypothesizedto reflect more resilience toward manifestationof PTSD as a product of culture, life experience,or life expectations—or, perhaps more likely,

an artifact related to differences in methodolog-ical and diagnostic stringency that have consis-tently plagued studies in this arena.

PTSD is not, however, the only nor even themost likely pathological outcome associatedwith traumatic exposure. History of traumaticexposure has been indicated as a risk factor fordepression (Zlotnick, Warshaw, Shea, & Keller,1997) with one study (Lopez, Piffaut, & Seguin,1992) reporting that 71% of raped women suf-fered from major depression while only abouthalf that number (37.5%) developed chronicPTSD of 1–3 years duration. It has been esti-mated that at least 30–40% of those who expe-rience a significant stressful event go on todevelop some significantly distressing reactionsby one year follow-up (Raphael, 1986), thoughstandard rubrics with established indices of re-liability and validity for assessing severity ofthe stressor and degree of resultant distress havealso remained somewhat elusive.

Efforts to mitigate the consequences of trau-matic exposure have commanded both collo-quial and professional concern. PD serviceswere argued to provide a simple and effectiveprophylactic for application immediately fol-lowing virtually any traumatic event (e.g.,Mitchell, 1983) and their use rapidly became awidespread practice. Employers, governments,and public policymakers rallied to calls for rea-soned and humane support of those potentiallyaffected. Most have depended solely upon thosemarketing these services for what informationthey may hold regarding advisability and utilityand have been given neither clear nor unbiaseddirection as to the wisdom of providing suchinterventions (Kenardy, 2000). It is only quiterecently that these consumers have been ex-posed to the controversies and questions sim-mering in the literature of academic research(cf. Kadet, 2002).

Early and consistent proclamations weremade of intervention efficacy in preventingPTSD, and arguments were made that the prac-tice was essentially devoid of iatrogenic riskand represented the only responsible avenue fora competent and compassionate response (cf.Mitchell, 1992). In marked contrast, reportsemerging more recently from the refereed sci-entific literature of the psychological disciplineshave increasingly suggested that preventativeeffects are limited at best (Bisson, McFarlane,& Rose, 2000; Raphael, 1999) and that the

319DEVILLY, GIST, AND COTTON

practice should be approached with caution(Bledsoe, 2003) or be treated as contraindicatedand curtailed (Mayou, Ehlers, & Hobbs, 2000;NATO, 2002; Parry, 2001; Rose, Wessely, &Bisson, 2001). Accordingly, those charged withmarshalling assistance in the aftermath of po-tentially traumatizing events have found them-selves in a cross-fire of data and assertions thatcan prove difficult for even a reasonably in-formed laity to decipher.

We attempt here to bring these current dilem-mas into sharpened focus through evaluation oftwo recent reviews of the PD literature, whichrealized diametrically opposing conclusions.We attempt to explain these divergent results byproviding a broadened and systematic review ofboth the literature related to the effectiveness ofdebriefing and of the social history of the de-briefing movement. We then propose some hy-potheses to explain the derived results and rec-ommend preliminary guidelines for appropriateaction by organizations and psychologists fol-lowing traumatic events as the enterprise beginsa long overdue shift from caveat emptor toevidence based practice.

Definitions

The area of trauma research has becomefraught with domain-specific jargon that can beeasily confused and is often misused by thoseoutside the research arena. PD and “CriticalIncident Stress Debriefing” (CISD), for exam-ple, are often used interchangeably. The formeris best described as a generic term for a class ofimmediate interventions following trauma (usu-ally within three days) that seeks to relievestress with the goal of mediating or avoidinglong term pathology. PD relies predominantlyon ventilation/catharsis, normalization of dis-tress, and psycho-education regarding presumedsymptoms. CISD, on the other hand, is a pro-prietary PD variant originally articulated byMitchell during the 1980s (Mitchell, 1983)through trade magazines, trade conferences, andproprietary seminars. It centers predominantlyaround group based interventions, though indi-vidual (or one-on-one) debriefings have alwaysbeen advocated as an acceptable and expectedvariant and relies heavily on reconstruction ofthe traumatic event, ventilation, and normaliza-tion. It also includes a structured “teaching”component.

CISD advocates have more recently envel-oped the debriefing component within an amal-gam of other self-help activities. These addi-tional components share a similar colloquialfamiliarity but, like debriefing, lack establishedempirical grounding. The revamped product, inwhich CISD remains the central and defining“signature intervention,” has collectively beendubbed “Critical Incident Stress Management”(CISM; Everly & Mitchell, 1997). It has nowbecome a frequent argument that the efficacy ofdebriefing applied within this context is some-how materially distinct from its efficacy as asingular intervention.

Devilly and Cotton (2003) have argued thatCISD and CISM have yet to be sufficientlydifferentiated to represent distinct interventions(to wit, not mutually dependent upon one an-other), nor have they been contrasted to deter-mine any differential efficacy. Other compo-nents of CISM stand similarly untested regard-ing efficacy in these applications, whetherjointly or severally, and interaction effects have,therefore, not been evaluated. No reliable evi-dence has been encountered to demonstrate thatsuch conjunction improves in any demonstrableway the efficacy of any component or mitigatesthe paradoxical impacts associated with the de-briefing component (which we describe below).CISD proponents nonetheless claim that thisscheme of interventions “mitigates the acutepsychological distress associated with psycho-logical crisis that may arise from violent acts,and . . . [will]. . . prevent or mitigate the inten-sity of adverse posttraumatic sequelae” (Everly,Flannery, & Mitchell, 2000, p. 23–24).

It also is important to differentiate prophy-lactic debriefing from early intervention forassessed pathological responses. PD usuallyinvolves wholesale provision of professionalservices, often through private debriefing com-panies, contracted employee assistance pro-grams, or volunteer “CISD teams” and “peerproviders,” immediately following a traumaticevent—often as a matter of organizational man-date. Early intervention, on the other hand, isthe provision of what may be called “restorativetreatment” to individuals who request psycho-logical help following trauma and manifest clin-ically significant presentations (Devilly, 2002).Recent evidence appears to support use of earlyCognitive Behavior Therapy (CBT) interven-tions for those diagnosed with Acute Stress

320 PSYCHOLOGICAL DEBRIEFING & OUTCOME

Disorder (ASD; Bryant, Harvey, Dang, Sack-ville, & Basten, 1998; Bryant, Sackville, Dang,Molds, & Guthrie, 1999; Foa, Hearst-Ikeda, &Perry, 1995), although the specificity of currentdiagnostic criteria and over reliance on disso-ciative symptoms within the classification hasbeen questioned in subsequent research (Harvey& Bryant, 1998). Likewise, CBT for those whoprogress to develop PTSD is demonstrably ef-ficacious, particularly techniques which pro-mote the graded and progressive processing ofinformation from the trauma and exposure tocorrective information (e.g., Devilly & Spence,1999; Foa et al., 1999; Foa, Rothbaum, Riggs,& Murdoch, 1991).

Analysis of Debriefing Review Studies

CISD Reviews

Everly, Flannery, and Mitchell (2000) of-fered a review of the literature related to PDwith specific emphasis on CISD. The CISDmodel encompasses seven explicit phases.Since Everly and colleagues stressed the impor-tance of strict fidelity to the CISD model as adeterminant of its efficacy, it is relevant tobriefly outline these stages: 1) the introductoryphase (rules, process, and goals outlined); 2) the“fact” phase (recitation of what participantssaw, did, and heard); 3) the “thoughts” phase(recounting of participants’ first thoughts asawareness of the event and its magnitude devel-oped); 4) the “reaction” phase (emotional reac-tions to the experience, sometimes labeled the“feelings” phase); 5) the “symptoms” phase(global assessment of physical or psychologicalsymptoms based on participant disclosures); 6)the “teaching” phase (educating the participantsabout common, likely, or possible stress re-sponses); 7) the “reentry” phase (referral infor-mation provided). CISD sessions generally lastone to three hours, are usually delivered togroups of individuals (though “one-on-one”sessions have always been presented as an ac-ceptable method of delivery), and are typicallyconducted within 24 to 72 hours after the event.The “process goal” of the intervention is de-clared to be “psychological closure subsequentto the crisis” (Everly et al., 2000, p. 26).

These authors argued that the inherent unpre-dictability of traumatic events renders con-trolled research difficult and that the need for

immediate assistance has generally precludedtraditional randomized controlled trials (RCTs).They also argued that assigning individuals tocontrol groups could be seen as “withholdingassistance” (Everly et al., 2000, p. 29), an argu-ment that would seem to beg the very questionefficacy studies are designed to address. Theyalso contended that studies should require as-sessments relating to adherence of the testedintervention to CISD protocols in order to clar-ify any procedural anomalies that could, inde-pendent of the debriefing, be responsible for thederived results. With these caveats in place,Everly et al. (2000) contended that only robustempirical studies had been included, with em-phasis placed on peer-reviewed articles andconference presentations. The authors then splittheir review first into studies with “comparison”groups and those with “no comparison groups,”then into those yielding “positive” or “negative”outcomes.

Their published results have been summa-rized in Table 1. Of particular interest are thosestudies that purportedly used the CISD model,utilized a comparison group, had positive out-comes, and had been used as the basis for anearlier meta-analysis by Everly and Boyle(1997). Everly and Boyle’s (1997) review ex-pressly stated that only group debriefings ex-plicitly employing the CISD model had beenincluded in their analyses.

The authors reported an averaged Cohen’s deffect size (Cohen, 1992) for each treated groupthat met these criteria and then averaged theseeffect sizes to obtain an averaged effect size forCISD interventions, which they reported to bed � 0.86. Since these investigations providedthe entire basis for the Everly et al. (2000) andEverly and Boyle (1997) claims that CISD is aneffective and advisable preventative interven-tion, a critical and detailed examination is war-ranted.

Table 2 summarizes the studies cited by Ev-erly and Boyle (1997), and subsequently byEverly et al. (2000), as evidence for the efficacyof CISD; we have added comments relating topertinent issues for each of the studies. Of thefive studies cited, data from only three wereavailable for evaluation, despite Everly et al.’sexplicit stipulation that emphasis had beenplaced upon empirically robust studies reportedthrough peer-reviewed journals and confer-ences. The studies by Wee, Mills, and Koehler


(1993) and Nurmi (1997) were conference pre-sentations at consecutive conferences sponsoredby Mitchell and Everly’s CISD organization.The original presentations were not availablefrom Dr. Everly (personal communication, Oc-tober 3, 2001), although later articles by both ofthese authors were forwarded by Dr. Everlyfrom the inaugural issue of the InternationalJournal of Emergency Mental Health1. Thesearticles are reviewed in place of those directlycited in the published Everly et al. analyses.

The citation for the Bohl (1991) article ref-erenced a government document, which wasalso unavailable. It was eventually obtainedwith the assistance of the United States FederalBureau of Investigation. That report did notpresent any actual data (means or standard de-viations) and, hence, did not allow effect sizesto be computed. This research hailed from adoctoral dissertation, which could not be ob-tained from either the author or the sponsoringschool. This left only two of the original articlesto review, followed by two articles publishedlater in outlets other than those actually cited inthe Everly et al. review.

Jenkins (1996) assessed 36 emergency work-ers following a mass shooting. Repeated mea-sures were taken at one week postevent and atone month follow-up, with the SymptomChecklist-90-Revised (SCL-90-R) being thepredominant measure of psychological distress

(n � 29). Of the 29 data sets available, 15subjects had attended at least one group basedCISD session. Jenkins concluded that those whoreceived CISD were more likely to have re-duced their anxiety and depressive symptoms.

Major methodological weaknesses limit theutility of this study. Pre-event SCL-90-R mea-sures were obtained by asking participants to“remember how you were feeling a week beforethe shooting” (Jenkins, 1996, p. 481), with thismeasure obtained after participants had com-pleted the SCL-90-R with respect to the weekfollowing the shooting. Such retrospective ap-proaches introduce a clear and considerable po-tential for reconstructive memory bias, particu-larly among those most distressed (cf. Keuler &Safer, 2001) and may also reflect nonspecific

1 The International Journal of Emergency Mental Healthis a publication of Chevron Publishing Company of EllicottCity, MD, that was initiated in 1999 to “promote the growthof CISM.” Chevron Publishing is a proprietary enterprisebegun by principals of the International Critical IncidentStress Foundation (ICISF), the organization created byMitchell and Everly. Chevron Publishing produces paper-back manuals and books on CISD related topics, generallyfrom ICISF principals and contract trainers and marketsother CISD related titles and merchandise. Our purpose fordrawing attention to this point is to provide a context withinwhich to appraise the independence of the peer-reviewprocess that is usually associated with scientific journals andassociated texts.

Table 1Summary of Results From Everly et al. (2000)

Study type Positive Negative

Compared Ersland et al. (1989) McFarlane (1988)Yule (1992) Deahl et al. (1994)Bohl (1991) Kenardy et al. (1996)Wee et al. (1993) Matthews (1998)Jenkins (1996) Bisson et al. (1997)Nurmi (1997) Hobbs et al. (1996)Chemtob (1997) Lee et al. (1996)

Uncompared Sloan (1988) Creamer et al. (1989)Lanning & Fannin (1988) Weisaeth (1989)Hytten & Hasle (1989) Searle & Bisson (1992)Shapiro & Kunkler (1990) Dyregrov et al. (1996)Feldman & Bell (1991) Griffiths & Watts (1992)Smith & de Chesnay (1994)Turner et al. (1993)Robinson & Mitchell (1995)Stallard & Law (1993)

Note. Compared � has a comparison group; Uncompared � no comparison group withinresearch design.


halo effects arising from visible displays ofconcern, presence, and promised assistance attimes of high perceived need. Further, morethan 45 correlation coefficients appear to havebeen calculated using 29 subjects, of which onlytwo correlations—–anxiety and depression(both p � .05)— proved significant. Since analpha level of .05 allows that one in every 20calculations may appear significant when notrue relationship exists, two significant resultswithin 45 analyses is precisely the value ex-pected to occur at random. If the alpha level hadbeen protected in these serial calculations, eventhese two would not have achieved significance.

Participants also self-selected participation indebriefing, raising a number of concerns withrespect to internal validity. Neither the debrief-ing procedure applied nor the characteristicsand training of the debriefers were outlined,raising additional questions regarding the natureof the intervention delivered. Given the stipula-tion that “only studies purporting to specificallyassess the CISD model of group crisis interven-tion (Mitchell, 1983) were utilized, consistentwith the narrative review and recommendationsof Everly and Mitchell (1997)” (Everly &Boyle, 1997; p. 2), this lack of detail raisesadditional questions regarding objective andsystematic application of exclusion criteria. Agreater concern, however, arises from our in-ability to derive Everly and Boyle’s estimatedeffect size of d � 0.93 from Jenkins’ publisheddata.

Chemtob, Thomas, Law, and Cremniter(1997) investigated the effect of a brief psycho-logical intervention given to 43 Participants sixmonths following a hurricane. They reportedsignificant improvements over time for thosewho were treated. The approaches employed inthis study, however, are difficult to reconcilewith standard tenets of CISD as an immediatecrisis intervention strategy. The interventionwas provided much later than normal debriefing(six months rather than the recommended24–72 hours), five hours of interaction wasinvolved, and the seven stage CISD model ofMitchell and Everly (1997) was not followed—conditions that, again, should have excluded thestudy according to the stated criteria for thereview. These major concerns not withstanding,application of Everly and Boyle’s method ofestimating effect size yields a Cohen’s d ofbetween 0.57 and 0.68 (depending on parameterestimates), approximately half the size(d � 1.35) reported by Everly and Boyle in theirreview (1997).

Wee, Mills, and Koehler (1999) reported anaturalistic study in which a convenience sam-ple of emergency medical and fire personnelwas sent a nonstandardized questionnaire con-cerning involvement with and reaction to a riotin Los Angeles following the acquittal of fourpolice officers tried for the video-taped assaultof Rodney King. Questionnaires were sent tomedical agencies asking employees to completeand return the forms anonymously. Question-

Table 2Meta-analytic Summary of Everly and Boyle (1997) With Extra Comments Regarding Studies

Study Population Effect size Comments

Bohl (1991) Police personnel .86 Non-random assignment, no pre-interventionmeasurement, effect size unjustified fromstudy details (should be 0.77).

Wee et al. (1993) Emergency medical techs. .47 Unpublished & unavailable for evaluation.Jenkins (1996) Emergency workers .93 Pre-intervention measures obtained by

“memory‘, participants self-selected intoconditions, no treatment integrity outlineor adherence checks, heightenedlikelihood of type I errors, effect sizeunjustified from study details.

Nurmi (1997) Emergency workers .89 Unpublished & unavailable for evaluation.Chemtob (1997) Hurricane victims 1.37 Intervention was 6 months post-trauma, five

hours of intervention, no control groups,no treatment adherence, not explicitlyCISD model, effect size unjustified fromstudy details (should be 0.68).

Note. Effect size � Cohen’s d (convention: 0.2 � small, 0.5 � medium, 0.8 � large).


naires sought information regarding physicaland psychological stress, CISD participation,exposure to traumatic scenarios, and work per-formance during the riot.

Much critical detail regarding methods, in-strumentation, and sampling cannot be deci-phered from the published article. No informa-tion is provided regarding reliability or validityof the instruments. Demographics of respon-dents, as contrasted to the population of emer-gency service workers involved, in response tothis event were not provided. The timing ofsubsequent mailings is unclear and does notappear to have been uniform; some participantswere apparently solicited at three months fol-lowing the riot with other respondents added upto a year later. Sixty-five participants returnedquestionnaires, of which 42 had attended CISDsessions up to two weeks following the riot.CISD therapist specifics were not described,except that they held from two to five yearsexperience. Participants were said to have beenself-selected, though CISD attendance was alsosaid to have been designated as mandatory byvarious services.

Differences between the two groups showeda moderate effect size on their measure ofstress, based on one-tailed significance testing.Given the absence of any a priori evidencepredictive of a directional result in favor ofCISD efficacy—and especially given the rangeof reported findings of neutral to paradoxicalimpact—two-tailed testing would have beenmore appropriate; had this been done, the re-sults would not have achieved significance (� �.05). This anomaly, coupled with the range ofmethodological flaws or omissions noted (e.g.,self-selection bias, respondent bias, memorybias, therapist nonspecifics, no fidelity informa-tion, etc.) render the results essentially uninter-pretable for the purpose of metaanalysis.

Nurmi (1999) used CISD to debrief firemen,rescue workers, and Disaster Victim Identifica-tion (DVI) teams following the sinking of theEstonia ferry. The author was one of the de-briefers and administered a host of psychomet-rically sound questionnaires (e.g., Impact ofEvent Scale—Revised, Weiss & Marmat, 1997;Penn Inventory, Hammarberg, 1992; SCL-90-R, Derogatis, 1992). Timing of the adminis-tration for the questionnaires was not reported.The control group was comprised of nurses, allfemale, potentially representing a decidedly dif-

ferent sample. It is not clear from the articlehow, when, and from where this comparisongroup was recruited. Analyses were conductedbetween debriefed and nondebriefed groups(DVI, firemen, and rescuers vs. nurses) andsignificant differences between these twogroups were reported, favoring those who hadreceived debriefing. The lack of comparabilitybetween treatment and control conditions, alongwith other design flaws, again precludes anyconclusions regarding the efficacy of debriefing.

Detailed review raises serious questions as towhether any of the studies from the Everly andBoyle (1997) meta-analysis met their own in-clusion criteria for either of their reviews (Ev-erly & Boyle, 1997, 2000). Analysis of thestudies utilized also raises serious concerns re-garding validity and applicability of these re-ports. As a consequence, the conclusions as-serted by these authors must be considered un-substantiated. This assessment is not ours alone,but has been independently reported by Bledsoe(2003); Fullerton, Ursano, Vance, and Wang(2000); Litz, Gray, Bryant, and Adler (2002);and van Emmerik, Kamphuis, Hulsbosch, andEmmelkamp (2002).

Cochrane Review of PD

Rose, Wessely, and Bisson (2004) conducteda meta-analytic review of the PD literature un-der the aegis of the Cochrane Collaboration, anextensive initiative to support evidence basedpractice. Their inclusion criteria limited studiesto RCTs and operationalized debriefing as asingle session intervention administered lessthan one month posttrauma that included “nor-malization” and “ventilation” components. Ex-clusion criteria included crisis intervention forpsychiatric patients, treatment of PTSD, de-briefing of research participants, support/bereavement counseling, N � 1 studies, andinterventions aimed at children. Of the 11 stud-ies that satisfied these criteria, none were in-cluded in the Everly and Boyle (1997) review. Itshould be noted that Rose et al. (2004) did notrestrict their review to CISD, but included allPD methods. A weakness of this evaluation isthat, as a result of their stricter inclusion criteria,none of the included studies utilized group de-briefing. Group debriefing is a method of inter-vention most commonly applied in disaster andorganizational settings and, therefore, their con-


clusions are not necessarily generalizable tomany debriefing situations. Cochrane and thescientific world await the first randomized con-trolled trial of group debriefing, a point weaddress later.

Studies included in the review predominantlyfound that debriefed participants were either nobetter off following debriefing or were deleteri-ously affected (see Table 3). Two studies (Bis-son, Jenkins, Alexander, & Bannister, 1997;Hobbs, Mayou, Harrison, & Worlock, 1996)reported that those who were debriefed weremore likely to develop PTSD than nondebriefedparticipants. Rose et al. (2001) concluded thatcompulsory debriefings should cease and thatresources would be better utilized by focusingon those who go on to develop diagnosablepsychiatric disorders.

One major obstacle in conducting researchwith traumatized populations is the difficulty ofacquiring participants willing to enter the re-search and/or employers willing to allow theresearch, and this limitation is apparentthroughout the studies included. These prob-lems are compounded further when strict ran-domization is sought. Most participants wereobtained either through the accident and emer-gency department within a hospital or throughanother hospital department. The only excep-tion was Rose, Brewin, Andrews, and Kirk(1999), who contacted victims of violent crimevia letter and requested their participation.These self-selected subjects may well have beenmore distressed and, hence, less likely (as dem-onstrated by Mayou et al., 2000, and describedbelow) to benefit from any intervention. Ac-cordingly, they may not represent the full rangeof variance in response types likely to be foundin non self-selected populations following manytypes of traumatic events.

Such samples may not, in particular, proverepresentative of an organization’s workforcefollowing a major event and may not take intoaccount the potentially quite larger pool of in-dividual reactions likely to be encountered insuch circumstances. Epedemiologic data dis-cussed above (e.g., Galea et al., 2002) suggestthat such groups may well be dominated byindividuals who might show no long-lastingimpact following the event if not debriefed, butwhose recovery could stand inhibited by thepotential paradoxical impacts of debriefing seenin multiple studies of individual applications.

The absence from this review of studies exam-ining the most typical applications (i.e., work-place or common exposure groups) in the mosttypical settings (e.g., workplace trauma; disasterand mass casualty events) has left the questionof differential impacts somewhat open, thoughthe burden of demonstrating any such differen-tial effect now clearly rests with debriefing pro-ponents.

Studies Omitted From Both Reviews

Two major problems plague the conduct ofliterature reviews and particularly the calcula-tion of meta-analytic representations of infor-mation in this arena. First, it is difficult to equil-ibrate the quality of the studies included. Qual-ity control (“Q”) statistics have been attemptedby designating a study as high, medium, or lowquality, based on various “gold standard” crite-ria (e.g., Foa & Meadows, 1998). The utility ofsuch designations is generally compromisedsince the methodological rigor of a study tendsto vary inversely with usual clinical practice.This often results in a technically excellent re-view, such as that by Rose et al. (2001), whichdoes not include any applications of the domi-nant mode of delivery (to wit, group debriefing).This is not necessarily a fault of the review, butstems from the lack of appropriate studies toinclude. The alternative is to specify very laxcriteria or criteria which seem to bend ratherthan break. This approach, however, results in areview that is both unreliable and nonspecific(as evidenced in the Everly et al. reviews).However, three recent RCTs have reported nulleffects following group debriefing, althoughtwo of these studies (Devilly & Annab, in press;Devilly, Varker, Hansen & Gist, in press) areanalogue in nature, with the second study fo-cusing on misinformation effects and not emo-tional outcome. In the first study, the research-ers inspected the effect of providing group de-briefing (CISD model) or “tea and coffee”following viewing of a very stressful video.Results indicated that while the participantsrated the video as very distressing, there wereno incremental positive effects from debriefing.The second study utilized a similar design, buthad a confederate introduce subtle misinforma-tion during the debriefing. It was found that thismisinformation was incorporated into eyewit-ness testimony at one month follow-up and that,


Table 3Meta-analytic Summary of Rose et al., (2004) With Extra Comments Regarding Studies

Study Population Comparisons Findings - post & follow-up Comments

Bunn & Clarke(1979)

Relatives of illhospital patients

1). PD2). Assessment only

Assessment taken withinminutes of debriefing.Statistics uninterpretable.

Doubtful whethercriterion A fulfilled.

Bordrow &Porritt (1979)

Road accident 1). Extended emotional& practical support

2). Minimal emotionalsupport.

Statistics uninterpretable. Unstandardised, 2–12hours contact.

Hobbs et al.(1996)

Road accident 1). PD2). Assessment only

IES: No significant differencebetween debriefed andnon-debriefed at post-treatment and follow-up.Trend for less associatedSx for treated.

Fewer PTSD diagnosesin controls.

Lee et al. (1996) Post miscarriage 1). PD2). Assessment only

IES: No significant differencebetween debriefed andnon-debriefed at post andfollow-up.

Hobbs &Adshead (1997)

Casualty attendees 1). PD2). Assessment only

46% treated & 56% controlhad “any psychiatricdisorder” by follow-up.

Bisson et al.(1997)

Acute burns 1). PD2). Assessment only

IES: No significant differencebetween debriefed andnon-debriefed. Butsignificant adverse effecton IES at 13 monthfollow-up for debriefed.

Fewer PTSD diagnosesin controls.

Lavender &Walkinshaw(1998)

Postnatal mothers(allprimigravidas)

1). PD/counselling(midwife)

2). Assessment

PD group significantly betteron Anxiety & Depression.

Listening, empathy,support, and anopportunity to askquestions about whathad happened.

Conlon et al.(1999)

Road accident 1). PD & education2). Assessed only

No difference betweendebriefed and non-debriefed at three monthfollow-up. Decrease insymptoms for bothconditions over time.

No significantdifference in PTSDdiagnosis betweenconditions.

Rose et al. (1999) Violent crime 1). PD & education2). Education3). Assessed only

All groups improved overtime with no significantdifference between them.

Participantsinterviewed in ownhomes and all self-selected to take partin research. Atfollow-up someparticipants receivingpsychologicaltreatment.

Small et al.,(2000)

Postnatal mothers(operativeintervention)

1). PD (midwife)2). Given pamphlet

PD group significantly worseon emotional functioning(SF36) and non sig onmost other measures. PDSs predominantly morelikely to suffer PND.

Measured mainlyDepression. NoPTSD measures.

Dolan et al., (inpress)

Hospital traumaclinic (generaltrauma)

1). PD(Mitchell/Dyregrov)

2). Assessment

N/A Unpublished.

Note. PD � Psychological Debriefing; IES � Impact of Event Scale; PTSD � Posttraumatic stress disorder; SF36 �Medical Outcome Study 36-item Short-Form Health Survey; PND � Postnatal Depression; Sx � Symptoms; Ss � Subjects.


overall, participants were more confident intheir incorrect responses. As yet unpublisheddata from this study also found no positiveeffects on emotional state following the debrief-ing or at one month follow-up. In the third study(Litz et al., 2004) a RCT was conducted ofgroup debriefing with another specific popula-tion—returned peacekeepers. Likewise, theyfound no discernable emotional effects from thedebriefing in comparison to stress inoculationand no treatment. So, while a full organizationalRCT of group debriefing has yet to be con-ducted, first signs tend to suggest that groupdebriefing may be a waste of resources. Onestudy (Devilly & Annab, in press) even suggeststhat those who receive debriefing are morelikely to report wanting to have spoken abouttheir experiences directly after the event, whilethose who did not receive debriefing remem-bered not wanting to have spoken about theevent they witnessed. Such a cognitive disso-nance explanation of satisfaction rates may,however, have negative effects on one’s percep-tion of safety and danger, as explained laterunder the section of priming and prepping. Fur-thermore, in occupations where eye witness tes-timony is of importance (e.g., emergency ser-vices), use of group debriefing is of concernwhen delivered before evidence has been ob-tained as there is evidence that debriefing cantaint recall when misinformation is introducedby one of the group members (Devilly et al., inpress).

The second major problem in conductingmeta-analyses comes as the converse of thefirst: Many studies are omitted from reviewsand valuable information is lost. No study ap-pears in both the meta-analyses discussed,though a quick look through the literature re-veals many other studies (experimental, theoret-ical, or observational) into the utility of PD thatdid not meet criteria for either review or weresimply overlooked in the searches (e.g., Andre,Lelord, Legeron, Reignier, & Delattre ,1997;Armstrong et al., 1998; Bierens de Haan, 1998;Brailey, Vasterling, & Sutker, 1998; Brom, Kle-ber, & Hofman, 1993; Carlier, Lamberts, vanUchlen, & Gersons, 1998; Creamer, Burgess,Buckingham, & Pattison, 1989; Cremniter etal., 1997; Deahl, Gillham, Thomas, Dearle, &Strinivasan, 1994; Doctor, Cutris, & Isaacs,1994; Ford et al., 1993; Gist, Lubin, & Redburn,1999; Hovens & Van de Weerd, 1998; Kenardy

et al., 1996; Matthews, 1998; Shalev, Peri, Ro-gel-Fuchs, Ursano, & Marlowe, 1998; Turner,Thompson, & Rosser, 1995; Viney, Clarke,Bunn, & Benjamin, 1985). Moreover, in anyrapidly evolving research area, new informationappears regularly, creating the need for ongoingassessments as the research lines expand (as isthe practice with the Cochrane Reviews).

Mayou, Ehlers, and Hobbs (2000), for exam-ple, have reported a three year follow-up of theHobbs et al. (1996) study regarding individuallydebriefed motor vehicle accident victims. Thosewho initially scored in the higher range on post-traumatic symptomatology were more likely tohave maintained their pathological presentationat both four month and three year follow-upintervals if they received the debriefing inter-vention, while those who received no such in-tervention tended to exhibit resolution. The au-thors concluded that, while mandatory debrief-ing should cease, practical and immediatesupport to those who are distressed should notbe denied. They suggested instead that interven-tion and support be tailored to individual needsand that follow-up treatment should use CBTinterventions with demonstrated empirical effi-cacy (e.g., see Bryant et al., 1998). It shouldagain be noted, though, that group debriefingswere not employed in their research.

Further, the thrust of the sentiment deliveredby the Rose et al. (2001) meta-analysis wasemphasized again in a recent meta-analysis byvan Emmerik et al. (2002). These authors like-wise conducted a literature search to find studiesthat had used debriefing techniques within onemonth following a trauma, and where symp-toms were assessed pre- and postdebriefing us-ing psychometrically acceptable assessment in-struments. Seven studies met their criteria, fiveof which used CISD as one intervention, sixused no-intervention control conditions, andthree used other PD-like interventions (i.e., “30minute counseling,” “education,” and “histori-cal group debriefing”). The results suggestedthat while people have a disposition to improveover time when they receive no intervention (onboth measures of PTSD and other trauma re-lated domains), neither CISD nor non-CISDbased interventions made a significant differ-ence in the outcomes reported. The authorsnoted, however, that although confidence inter-vals overlapped, the effect size was moderatefor nonintervention and moderate to strong for


non-CISD interventions. The interval for CISD,unlike the other conditions, included zero andnegative values, indicating no effect or a possi-ble paradoxical impact on resolution. Put moredirectly, the provision of CISD would appear toinhibit or even reverse the normal inclinationtoward resilience and resolution while the pro-vision of non-CISD interventions had no nega-tive effect and may, at least to some degree,work to enhance normal patterns of recovery.For this reason, yet other authors have sug-gested that debriefing should cease as a practice(McNally, Bryant, & Ehlers, 2003). However,this does not mean that early intervention for arecognizable disorder (e.g., ASD) should be de-nied and neither does it mean that people aredenied practical help and emotional supportwhen they are self-requested, a point we shallreturn to later.

Possible Negative Impacts of DebriefingModules

In order to look at what may be useful toorganizational settings, it is necessary to firstlook at what may be harmful or counterproduc-tive. Why should it be that in some studies thosewho were debriefed were more likely to developPTSD? What parts of the debriefing processmight be inhibiting to natural resolution? Whichsubsets of individuals may be most susceptibleto paradoxical impacts? It is most important toquestion if we can remove these aspects andscreen participants, yet still have an interventionwith high satisfaction ratings and practical util-ity? Will it ultimately help to mitigate long termdistress?

Most debriefing interventions have beenmodeled on the seven-phase CISD process de-scribed by Mitchell (1983; Mitchell & Everly,2000). The first phase, consisting of processintroductions, would appear relatively benignfrom a structural perspective, though complica-tions may arise from representations respectingnature and expected efficacy of the intervention,failure to inform of foreseeable risks, and thelike. Indeed, any failure to provide clear andcomplete information regarding current findingsrespecting limitations of efficacy and indicatorsof paradoxical inhibition of recovery—muchless any representation to the contrary—wouldimmediately raise major issues surroundingfully informed consent. Each of the central in-

tervention phases to follow, however, presentsone or more potential sources of difficulty andthe interactions between them collectively en-hance their possible impact.

Serial Revivification and HeightenedArousal

The “facts” phase and event reconstruction.Most debriefing protocols encourage specificreporting of what one saw and heard during theevent, moving from there to articulation of whatone was thinking and feeling—often specifi-cally inquiring as to the worst moments andmost intense emotions encountered. While thegoal of this exercise is often described in termsof creating a calibrated perspective of the event,such reconstruction may serve to a) modify theeye witness memory of the event, as outlinedabove (Devilly et al., in press) and b) intensifyalready disturbing reactions by reconnecting theindividual with the sources of discomfort wellbefore sufficient distancing has been achieved.In such cases, this revivification is unlikely toserve its intended cathartic end and may bemore likely to arrest than to accelerate the pro-cesses inherent in normal resolution.

Group applications of debriefing, rather thancreating a shared picture of circumstances andevents, may further compound these issues byexposing individuals struggling to keep theirown arousal in check to additional, potentiallyeven more vivid and arousing, constructions ofthe event and its images. Especially when theprocess is invoked within the frequently recom-mended 24–72 hour postimpact envelope, thepotential for these paradoxical impacts may beheightened as one progresses from this elementthrough the “thoughts” and “reactions” phasesto follow. Given that Charlton and Thompson(1996) found only positive reappraisal and dis-tancing to be coping strategies predictive ofsuccessful adaptation, this early insistence onreconstruction may well run counter to the veryprocesses most likely to promote eventual res-olution, problems that may be systematicallycompounded in the following two phases of theclassic CISD rubric.

The “thoughts” phase and cognitive reap-praisal. The “thoughts” phase of the tradi-tional CISD model asks participants to articu-late their first thoughts as the impact and mag-nitude of the traumatic event first came to their


awareness. While perhaps intended to establisha sort of cognitive baseline from which subse-quent reappraisals could emerge, it may para-doxically serve to further solidify the negativeelements of revivification associated with thenarrative reconstruction of the “facts” phase.

This may again be compounded in groupapplications, where some people may not havebeen fully aware of the level of danger to whichthey were exposed. This postevent processcould lead to a reappraisal of their memory ofthe event in a way that could increase subjectiveestimation of threat (e.g., “I thought the gun thatthe small guy carried was probably a fake – Ididn’t even realize that the second guy also hada gun, let alone that it was real”). Such reeval-uation of a situation has been posited as centralin the derivation of a fear response (Davey,1993) and others have shown that such an in-creased subjective appraisal of danger correlateswith pathologic outcomes (Solomon, Miku-lincer, & Benbenishty, 1989; Stallard, Velle-man, & Baldwin, 2000). This may be exacer-bated in vulnerable individuals as these cogni-tions are again paired with arousal sensations asthe process moves into the “feelings” or “reac-tions” phase.

The “reactions” phase and cathartic ventila-tion. In this phase of the CISD intervention,participants are asked to articulate their emo-tional responses to the event, often through que-ries such as “What was the worst part of theexperience for you?” The serial progressionfrom narrative reconstruction of events (the“facts” phase), through cognitive retrieval ofproximal perceptions (the “thoughts” phase), toreconnection with the immediate emotional im-pact of the experience in the “reactions” phasepresents an effective completion of the revivi-fication, returning those who fully submit them-selves to the process to subjective states thatcould well prove intolerably close to thosestates of terror, helplessness, and confusionfrom which distancing is most vital.

The process of revivification represented inthese phases, delivered to people who mightnormally have gone on to process the informa-tion successfully if left unassisted, runs the riskof sensitizing such persons to the stimuli in-volved at a time when desensitization is vital toresolution. The processes of desensitizationnecessary to address pathological elements in-herent in diagnosable PTSD require systemati-

cally graded exposure to defined stimuli andprogressive habituation to those stimuli to ex-tinguish the fear response and provide correc-tive information to challenge aberrantly heldbeliefs (Foa & Kozac, 1986). This clearly can-not be accomplished in a “one-off,” ostensiblyprophylactic group or individual encounter. In-deed, such short-term and short-lived exposureto memories of threat in people who may notcurrently have a pathological condition, butwho present with pronounced subjective dis-tress could quite conceivably run the risk ofgeneralizing the memories and priming certainstimuli which are, or could become, triggers forconsolidating the fear response (McNally et al.,1987). Regardless of any intent to “normalize”this condition of subjective arousal, those al-ready struggling to regulate hyperarousal andintrusion may find these elements of the debrief-ing rubric to reinforce and exacerbate, ratherthan to mitigate and diminish, their subjectivediscomfort.

Priming and Prepping of Symptoms(Modeling Dysfunctionality)

The “symptoms” phase and attribution.While the rhetorical justification for this phaseis to “normalize” whatever reactions may befelt, there is a subtle but possibly very profounddifference to be drawn between discussing com-mon manifestations of postimpact distress andpriming people to consider these discomfituresas if pathological symptoms. Moreover, repeat-edly labeling the event “traumatic” superim-poses a set of attributions and expectations thatmight not otherwise occur. Such attributionsmay dispose vulnerable individuals to interpretthe inescapable disequilibrium of disruptive lifeevents as pathological anxiety, becoming, ineffect, a self-fulfilling prophecy of despair.

Indeed, the very labeling of subjective expe-riences which are, in most cases, signs of ines-capable disequilibrium as if they are “symp-toms” of pathology may contribute to a “medi-calization” of the experience—to wit, “I didn’tthink of myself as sick until you sent for aremedy” (Gist, 2002). Here again, the combinedimpact of one phase (in this case, the “symp-toms” phase) with that of its succeeding phase(the “teaching” phase) holds even further poten-tial to compound complications for vulnerableparticipants. There is even preliminary evidence


that providing trauma patients with pamphlets(education) regarding trauma responses doesnot help follow-up presentation and appears tohave a paradoxical effect on depression andPTSD “caseness” (Turpin, Downs, & Mason,2005).

The “teaching” phase and psychoeducation.Debriefing protocols generally include a “psy-choeducation” element intended to providemodeling and information respecting adaptiveapproaches to addressing the trauma and itssequelae. However, these generally center oncolloquialized discussions of PTSD. Debrieferstypically attempt to accomplish this by distrib-uting lists of problems (e.g., increased irritabil-ity, avoiding reminders of the trauma, disturbedsleep, intrusive memories of the event, etc.)which participants are told they may expect toexperience and then provide suggestions, oftensimplistic at best, regarding coping strategiesand approaches. A narrow focus on the coreconstructs of PTSD may lead one somewhatastray in dealing with disaster as a social expe-rience (see Staab, Fullerton, & Ursano, 1999,for an alternative construction). Social compar-ison under threat, however, may prove a moresalient construction for understanding both suc-cessful adaptation and paradoxical impacts.

Perceived threat lends a unique urgency tothe search for affiliation and social comparison(Kulik, Mahler, & Moore, 1996), and these con-tacts follow particular patterns that underscorethe need for specifically appropriate models (seealso Taylor, 1983). The models preferred arethose seen to be similarly situated, and espe-cially those offering clear indications of havingevolved and sustained successful adaptation tosimilar demands (Taylor & Lobel, 1989). These“upward contacts” are contrasted against othermodels, whether known or imagined, which areperceived to have fallen short of acceptableadaptation (dubbed “downward evaluations”).The more abrupt, unexpected, novel, or ambig-uous the experience, though, the less likely thatsuitable models will be readily available toserve as effective upward contacts.

CISD approaches frequently prescribe utili-zation of “peer” debriefers, precisely to fulfillsuch a modeling role. These are often personswhose prior experience with traumatic expo-sures has encouraged their participation in thiscapacity. Those whose prior exposure to traumahas left them with unresolved issues for which

vicarious rumination may be sought might wellrelish opportunities to enter settings where suchreprocessing can be offered as if a therapeuticcontribution to others. This can result in anunwitting dispatch of responders who serve asinadvertent downward evaluation targets whenthese factors demand instead a very specifictype of upward contact to provide effectivemodeling and support—a type unlikely to befound in conjunction with persistent cathexistoward reprocessing those very events thatshould have been adequately distanced and re-framed in the adaptive process.

Taylor (1991) presented a hypothesis regard-ing this seeming paradox of resolution, consis-tent in many ways with her earlier argumentsregarding the role of “positive illusions” inmental health (Taylor & Brown, 1988). Theessential premise would suggest that profoundlynegative events require major mobilization ofpersonal resources to respond effectively and toweather their impacts. But successful resolutiondemands subsequent minimization of the re-called impact of the event and of those veryresponses initially commanded. Those bestadapted, and hence best suited, to provide up-ward contact modeling are, of course, mostlikely to be found among those seasonedthrough occupational experiences that have de-manded prior resolution and accommodation.This is certainly consistent with the recurrentfinding that experience is among the most ro-bust protective factors mitigating postexposuresymptomatology (McCarroll, Fullerton, Ursano,& Hermsen, 1996;McCarroll, Ursano, & Fuller-ton, 1993; McCarroll, Ursano, Fullerton, &Lundy, 1993; McCarroll, Ursano, Ventis, &Fullerton, 1993). If, however, the essence ofsuccessful accommodation entails minimizationand its functional analogues (to wit, distancingand positive reappraisal), the most effectivemodels would be expected to specifically avoid(rather than to proactively seek) visible inter-ventionist roles. This would certainly stand con-sistent with Redburn’s (1992) findings regard-ing a strong inverse relationship between expe-rience and participation in debriefing exercisesand may suggest another plausible hypothesisregarding the paradoxical findings respectingthe objective efficacy of the intervention in oc-cupational group settings (Gist, Lubin, & Red-burn, 1998).

Given that Gump and Kulik (1997) found


that settings comprised of persons who sharetraumatic exposure contain demonstrable ele-ments of social contagion, blanket applicationof an indiscriminate group process may standparticularly prone to stimulation of negativeoutcomes, especially when invoked before con-structive coping strategies have had time tofully evolve in the affected individual or popu-lation. This may be paradoxically enhancedwhere models portray courses or styles of ad-aptation inconsistent with the coping predilec-tions of recipients. Particularly within the fieldof teacher-student interactions, it has been notedthat higher-status supervisor expectations influ-ence lower-status subordinate performance inthe direction of the expectation (see Kierein &Gold, 2000, for a review). Furthermore, self-fulfilling prophecies and self-verification ap-peared to occur simultaneously in a contextwhere supervisors and subordinates apparentlyhad highly valid information on which to basethese initial expectations (Madon et al., 2001).Add to this the hypothesis that an overestima-tion of threat and fear expectation plays a causalrole in the origins and maintenance of anxiety(Wiedemann, Pauli, & Dengler, 2001) and that,in people with panic disorder, the expectation ofpanic is associated with actual panic occurrence(Kenardy & Taylor, 1999), and the potential forselective misadventure again increases.

The “reentry” phase and appropriate refer-ral. Appropriate referral may be operationallydefined as the timely direction of clients indiscernable need of intervention toward provid-ers trained and competent in techniques withempirically demonstrated efficacy in resolvingtheir presenting problems. For most issues as-sociated with exposure to disasters and similardistressing events, this would entail referral ofthose demonstrating ongoing indications of corePTSD symptoms—most specifically arousaland intrusion symptoms persisting at 4 – 6weeks postimpact (Brewin et al., 2002)—forshort-series CBT variants employing exposure(see Litz et al., 2002, for overview of effectiveearly interventions). Such referrals should alsoentail avoidance of ordinary supportive coun-seling techniques (cf. Bryant et al.,1998; Bryantet al., 1999), especially for manifestations ofASD. Various curricula taught to debriefersthrough the primary training organ of CISD(International Critical Incident Stress Founda-tion, 2003b) do not reflect these modalities, but

do reflect training opportunities in such mar-ginal approaches as Thought Field Therapy(TFT) and similar “power therapy” techniques(see Lohr, Hooke, Gist & Tolin, 2003; Devilly,2005, for an overview of TFT and other contro-versial trauma treatments). While no formalstudy of referrals emanating from CISD ses-sions can be found, nothing in the curriculaexamined would indicate that appropriate refer-ral for evidence based intervention is taught,much less systematically monitored or evalu-ated.

Individual Factors Related to DifferentialImpact

Certain symptom presentations in the imme-diate postimpact period may dispose individualstoward differential impacts from debriefing in-terventions. Mayou et al. (2000) reported thatthose with high intrusion and avoidance symp-toms as measured by the Impact of Events Scale(IES) at intake fared particularly poorly withdebriefing, remaining symptomatic at three yearfollow-up, while those of similar presentationwho did not receive the intervention displayed aproclivity toward resolution. A similar findingwas reported in a dismantling study of CISD inwhich the seven-phase process with either the“feelings” (emotional ventilation) or the “teach-ing” (psychoeducation) phases omitted wascompared to a nonintervention control group.Neither approach to debriefing proved effica-cious at 24 weeks follow-up, though those withlow hyperarousal showed a mild intermediatebenefit from the educational debriefing (emo-tional ventilation component omitted), whilethose with two or more hyperarousal symptomsshowed better resolution without either form ofintervention (Sijbrandij, 2002).

Debriefers trained in CISD protocols aretaught that those showing higher symptom lev-els are those most in need of participation. Whatcomparative data have been reported, however,suggest strongly that these are instead the per-sons most likely to experience paradoxical out-comes. Indeed, while the intervention continuesto appear inert overall, comparative and dis-mantling studies available suggest that whatpalliative impact it may hold is limited to thosewith the least subjective distress—to wit, whatlimited benefit it may offer accrues to those who


need it least, while those in most distress maybe most inclined toward paradoxical impacts.

Broader Considerations RegardingIntervention

Perhaps the most salient cause for concern inall the interventionist zeal is captured in Gilbertand Silvera’s (1996) concept of overhelping.They demonstrated that immediate and highlyvisible attempts to “help” a target individualwith processes that the target would, in fact,have successfully executed without aid servedto defeat perceptions of self-efficacy centralboth to personal and interpersonal assessmentsof mastery on the part of the target. Theseassessments of self-efficacy, however, may becrucial to successful adjustment (Major, Cozza-relli, Sciacchitano, Cooper, & Testa, 1990). Ac-cordingly, given the consistent finding that mostindividuals confronted with disaster resolve itsimpacts with or without intervention (Cook &Bickman, 1990; Helzer, Robins, & McEvoy,1987; McFarlane, 1988; Redburn, Gensheimer,& Gist, 1993; Rubonis & Bickman, 1991; Sal-zer & Bickman, 1999), the very essence of ourcurrent trend toward rapid, highly promoted,highly visible intervention may be, at its mostbasic level, counterproductive for those wemost intend to aid.

Seeking Empirical Guidance

So what do we really know about debriefingand what should a responsible organization/practitioner do? The first step is to look at thoseareas where most parties seem to agree, seekingto find the threads of some general consensusfrom which to frame an informed and meteredresponse with the best prospects for meaningfulassistance.

Areas of Agreement

A. The first area of agreement appears,predictably, to be that agreement islacking. While early proclamations re-garding the effectiveness of the ap-proach frequently asserted its pur-ported “scientific” grounding (cf.Mitchell, 1983, �1992; Mitchell & Ev-erly, 1997), independent researchers

have continued to note the weaknessesin the data proffered (see Gist et al.,1997; Gist, Woodall, & Magenheimer,1999, for discussion). Indeed, at aNorth Atlantic Treaty Organization(NATO) – Russian workshop on ter-rorism, the general agreement was that“there is still no consensus on the role,if any, of very acute interventions.Classic CISD debriefing can no longerbe recommended. The balance be-tween getting people to talk to people,and getting people to talk to profes-sionals, has not been established”(NATO, 2002). While it must be notedthat proponents of this intervention,particularly those with direct interestin its proliferation, continue to disputethis position (cf. Mitchell, 2003), it isfair to say that the limitations of cur-rent data are widely acknowledged.

B. Debriefed parties generally seem toappreciate the gesture. Client satisfac-tion with the procedure has beenwidely reported as strong (e.g., Arm-strong et al., 1998; Robinson & Mitch-ell, 1993). But more critical assess-ments of satisfaction data sometimesreveal an endorsement that is less thanoverwhelming. Bunch and Wilson(2002), in a fire service trade maga-zine, reported that “critical incidentstress debriefing was considered to behelpful at some level by no less than70% of Oklahoma City firefighter sur-vey respondents” (p. 48). When thosedata were presented in tabular format,though, approximately three times asmany were found to have rated theintervention “not helpful” as had ratedit “very helpful” and more than twothirds had rated it below the midpointof the four point Likert-type scale theyhad been presented (options of “veryhelpful,” “helpful,” “somewhat help-ful,” or “not helpful”).


While expectancy for change in people diag-nosed with psychiatric disorders and attendingtreatment is sometimes related to actual change(Devilly & Borkovec, 2000), high levels of sat-isfaction with debriefing have not necessarilybeen reflected in positive outcomes (e.g., Car-lier, Voerman, & Gersons, 2000). Gist, Wood-all, and Magenheimer (1999) compared this re-lationship to customer service surveys found ina neighborhood doughnut shop:

One can determine very precisely that people (espe-cially people who choose to come to such shops) tendto like doughnuts—that tells us, however, absolutelynothing about their nutritional worth. Many people likeand even crave doughnuts for precisely those proper-ties that render them nutritionally undesirable. Wecertainly wouldn’t accept an argument that preferencesof the palate translate into dietary superiority (p. 279)

Certain operationalizations of low satisfac-tion, such as having received no form of psy-chological intervention, could conceivably cor-relate to higher end state functioning, particu-larly given the finding that debriefing in someindividuals has been less potent than the naturalproclivities toward resilience and may, in fact,inhibit its progress (Gist & Devilly, 2002; vanEmmerik et al., 2002). While it is generallyagreed that traumatized individuals both expectand appreciate some form of visible aid, not allforms of help turn out to be equally helpful(Gist, Lubin, & Redburn, 1999). Indeed, there issome evidence that this “satisfaction” with de-briefing may be nothing more than the manifes-tation of cognitive dissonance (Devilly & An-nab, in press).

C. It is generally agreed that most organi-zations earnestly desire to provide somekind of assistance to their employees orclients. It is also generally agreed that thereasons for this desire to intervene arebroadly based and include the domains ofsocial concern (“I don’t want my employ-ees to be hurt or sick, and I want to help”),legal concerns (“If I don’t do something, Icould be sued for negligence”) and orga-nizational considerations (“A healthy andsatisfied workforce is a productive work-force”). It is, therefore, generally agreedthat some kind of disaster response plan isuseful to an organization, particularly con-sidering workplace health and safety con-cerns and possible litigation (cf. Devilly &

Cotton, 2003). That plan, however, mustreflect sound empirical information if it isto be maximally instructive and avoid del-eterious, if unintended, side effects.

D. Most researchers and clinicians wouldalso agree that those who are distressedfollowing a traumatic event should be de-nied neither practical nor emotional sup-port, although the best method of deliver-ing such support remains unclear. Everly etal. (2000) contended that CISD stood asthe only proven immediate interventionmodality but, as made clear in the abovediscussion, CISD is a method of very ques-tionable utility, particularly with thosemost distressed (Mayou et al., 2000). Roseet al. (2001) recommended, given the ab-sence of quality data regarding group de-briefings and the possibly iatrogenic ef-fects of individual debriefings, that re-sources should instead be focused onidentifying and treating those who developdiagnosable disorders following trauma.These recommendations, however, fallshort of the needs and requirements oforganizations by failing to specify reason-able approaches to assistance or outlinemethods for their delivery. A more prudentapproach for those arenas in which CISDfound its strongest footing (response orga-nizations and other corporate entities) maybe to frame the assistance needed withinbroader organizational contexts, as op-posed to these somewhat tangential quasi-clinical models of intervention that havefailed to demonstrate utility or efficacy.Such an approach, coupled with peer-riskassessment training, is currently beingtrialed in the Royal Navy in the U.K.(Jones, Roberts, & Greenberg, 2003).

The Organizational Context

Evidence has been emerging in the work psy-chology and organizational behavior literaturethat the organizational context may exert amuch stronger influence on outcomes related toemployee well being than has hitherto beenrecognized. Hart, Wearing, and Heady (1994)found that organizational experiences (e.g.,management practices, decision-making, careeropportunities, clarity of roles, coworker rela-


tions, performance feedback, etc.) were morestressful for a nonclinical population of servingpolice officers than operational pressuresunique to police work (including exposure todanger, threats, and attending the aftermath ofincidents with fatalities). More recently, Hartand Cotton (2003) replicated this finding withanother sample of police officers. They alsofound that a low level of positive affect (whichthey termed “morale”) was a much strongerdeterminant of police withdrawal behaviors(e.g., stress-related absenteeism and intention tosubmit a stress-related workers compensationclaim) than levels of overt psychological dis-tress. With this in mind, it is very possible thatoperational pressures and stressful events mayappear to be the triggers of traumatic reactions,when in fact they act purely as a vehicle withwhich to express more pervasive organizationaldissatisfaction.

Work-related clinical and stress research hashitherto focused almost exclusively on negativeworkplace events and their impact on indices ofnegative emotional responses in the workplace(Hart & Cotton, 2003). Beaton and Murphy(1993), however, found that the impact of oc-cupational “critical incidents” failed to contrib-ute significantly to job satisfaction predictors ina large population of firefighter/paramedics andbarely achieved significance among firefighter/EMTs, with more mundane (but also more per-vasive) occupational strain factors such as com-pensation, supervision, and sleep deprivationemerging as the principal influences. Hart andCotton (2003) similarly found that leadershipbehaviors (particularly relating to people man-agement skills) and organizational climate (e.g.,organizational structures and work team pro-cesses) were the strongest determinants of lev-els of positive affect, accounting for approxi-mately 70% of the variance in levels of morale.These daily factors in organizational climateand leadership may provide a more significantset of foundations for organizational resilience(Gist & Woodall, 1999).

This line of research suggests that, at theorganizational or work group level, employersshould accord priority to workplace strategiesthat maintain employee morale and improve thequality of people management practices as op-posed to implementation of routinized, quasi-clinical interventions implemented in staccatofashions following disruptive workplace events.

Organizational interventions must also be re-sponsive to a distinctly different client—the or-ganization and its management rather than theindividual employees (whether taken jointly orseverally)—holding objectives somewhat dif-ferent and often distinct from the interests of theindividual employee (e.g., work team integrity,maintenance of productivity, limitation of lossand liability). While these interests often over-lap and frequently intersect, the routes forachieving them are commonly quite disparate.

There is gathering evidence that positive af-fective responses from employees contribute toincreased discretionary performance (Borman& Motowildo, 1993), as well as reduced absen-teeism (George, 1989, 1996). Workers’ com-pensation costs have also been reported to bereduced where these positive affect impacts areachieved (Hart & Cotton, 2003). Field tests ofapplications’ contrasting approaches based inorganizational development against traditionalCISD models have shown them strongly fa-vored in career fire service organizations(Woodall, 1994; see also Gist & Woodall, 1995,1999). These approaches seek to strengthen theorganization’s preparedness to deal with thedemands of challenging workplace events, astrategy also receiving increasing attention inother corporate and organizational settings(Blythe, 2002).

Organizations seeking to provide meaningfulassistance find themselves caught on the hornsof a dilemma that their ordinary approaches toproblem solving may be ill equipped to address.Recent cases, based on the prevailing practiceof immediate debriefing, have prompted orga-nizations to be concerned of litigation for notproviding immediate intervention (e.g., Howellv. State Rail Authority of New South Wales,Australia2). Now, in light of increasing empiri-cal evidence of inefficacy and paradoxical im-pacts, these same organizations are beingwarned that they could conceivably be sued forproviding a noxious intervention that has beendemonstrated to increase the risk of developinga pathological outcome for some employees(Bledsoe, 2002; Devilly & Cotton, 2003).

Indeed, in a landmark class action case inwhich a number of former military combatants

2 S6/1997 & 93400071 Geoffrey Clarence Howell v StateRail Authority of NSW.


sued the British Ministry of Defense for allegedfailure to adequately foresee, prevent, diagnose,and treat psychological sequelae of traumaticexposure, the Court entered specific findingsregarding the inefficacy of PD and the possibil-ity of paradoxical outcomes indicated in empir-ical studies (PTSD Claimants v. Ministry ofDefense). The Court also noted that, while theclaimants had amended their cause of action todelete their original claims that failure to pro-vide PD was negligent (due, the Court opined,to awareness that such a claim could not prevailgiven the clear evidence questioning the prac-tice), the implications of the evidence regardingPD must be considered as probative across therange of similar interventions and approaches. Itbecomes critical at this juncture that well in-tended employers receive some more definitiveguidance regarding appropriate measures theymight take.

Considerations for Organizational Action

There are several avenues of theory and re-search that can lend both insight and direction tothe emerging practice of organizational andcommunity assistance in disaster, and whichmay help to frame and potentially resolve someof the dilemmas raised above. Their effectiveconsideration, however, demands that we firstretreat to the bedrock of our explanatory framesand consider disaster as a developmental chal-lenge, rather than as a pathogenetic threat (cf.McCrae, 1984). A number of suggestions can beproffered that may serve to accomplish an or-ganization’s crisis management and employeesupport objectives while still reflecting empiri-cally supported best practices.

Proactivity. Higher self-mastery and asense of control within organizational settingstend to predict less negative affect when peopleare faced with stressful tasks (Hoffman, 2001).Increasing a sense of mastery is, of course,context specific. In emergency services organi-zations, for example, the implementation ofconsistent incident management systems hasbeen suggested as more influential in mitigatingincident stress than programs of psychologicalintervention (Gist, Lubin, & Redburn, 1999).Police officers involved in protracted body re-covery and identification work showed no del-eterious impacts where their activity was sys-tematically and sensitively managed to yield a

sense of optimal performance under duress (Al-exander & Wells, 1991); indeed, several do-mains showed actual improvement compared topreincident baseline values. In an uncontrolledstudy, even practiced general approaches tostress management were found to be effective inmitigating impact of a catastrophic occupationalevent involving an air ambulance mishap, whilepostincident CISD exercises showed no signif-icant impact (Macnab, Russell, Lowe, & Gag-non, 1999).

Just as premorbid functioning has been agood predictor of longer term posttraumatic im-pact for individuals (McFarlane, 1988), thestrength of an organization and its employees atthe time of a crisis may well be the most salientpredictor of its resilience in the aftermath (Gist& Woodall, 1995). The most basic elements ofsuch a foundation may be detected less in mea-sures directed toward crisis anticipation andabatement than in measures directed towarddaily functioning and effectiveness. EmployeeAssistance Programs that are well integratedand well utilized to deal with the ongoingstrains of both workplace and daily living pro-vide a solid foundation from which postcrisisresilience can emerge and encourage a solidrelationship between the organization, its em-ployees and its providers of psychosocial assis-tance prior to the occurrence of a major disrup-tive event. These must be coupled, however, toother elements of business planning and humanresource management if their intersection is toprove timely and effective at junctures charac-terized by stress and disruption.

Organizational and incident specificity.One frequently identified problem with debrief-ing interventions has been the attempt to applya rigidly uniform approach to responses acrossan increasingly wide range of organizations,industries, settings, and events (Gist, 2002). Re-cent ICISF trade show marketing displays haveadvocated the training and approach for lawenforcement agencies, fire and EMS providers,airlines, EAPs, clergy, schools, and social ser-vice providers, with courses offered on theirwebsite including applications targeted towardcorporations, hospitals, children, and families.Yet the generalizability of the approach, evenacross its originally intended public safety tar-gets, has been found lacking in direct examina-tions (Gist & Woodall, 1999; Woodall, 1994).Variability between organizational missions,


cultures, expectations, structures, and commu-nication patterns, as well as the interaction ofthese with incident, employee, and work groupcharacteristics all demand careful tailoring ofapproaches to fit the circumstances actually en-countered.

Organizational preplanning. While the ex-act nature, location, or timing of workplacecrises cannot be reliably predicted, most orga-nizations can reasonably anticipate that suchevents will at times occur. The development,documentation, and rehearsal of effective crisismanagement plans helps ensure that the organi-zation will be reasonably prepared to respond ina measured, timely, and effective way to assertcontrol of circumstances and implement neces-sary measures to ensure the safety and supportof its affected staff.

Successful responses will generally deal withpractical and instrumental needs of affected em-ployees, provide structure and continuity, vali-date concerns and provide a sense of presenceand care, and take measures to ensure that anyin need of specialized attention are recognizedand receive timely referral to appropriate ave-nues of effective intervention. Such plans con-centrate on crisis communication, informationflow, business continuation, and similar mattersof concern to both the organization and its em-ployees (see Blythe, 2002, for one set of sug-gested templates). Further, and in light of acompany’s nondelegable duty of care and itspotential vulnerability to claims based on neg-ligent failure to plan, its policy should be reg-ularly revised in consultation with a recognizedexpert in crisis planning and response andshould reflect changing evidence from the em-pirical literature.

Immediate instrumental support. Recom-mendations regarding immediate efforts at as-sistance have become much more understated,practical, and nonintrusive. These interventionsfocus instead on instrumental support throughexisting (i.e., nonclinical, nonpsychological)programs and relationships and emphasize nat-ural avenues of social support and resilience inthe immediate aftermath period (cf. Bledsoe,2003; Gist, 2002; NATO, 2002; Ritchie, 2001).These again require close integration into theoverall schemata of corporate and organiza-tional responses and will many times be bettermediated by familiar, visible corporate repre-sentatives dealing with tangible needs who

communicate a genuine concern for concomi-tant emotional impacts.

While validation of experiences and demon-stration of concern has been shown to increasemorale (e.g., Hart & Cotton, 2003), it is alsoimportant to be careful of “terminology slip-page” and the creation of a trauma myth where,for example, the issue of concern to the employ-ees in a workplace was organizational in nature,such as the dismissal of a unit manager orsimilar events of less than crisis magnitudes(Devilly & Cotton, 2003). In the past, encour-aged by the expansive claims of debriefing pro-ponents regarding the benefits of such interven-tion, sessions have been orchestrated to addresseverything from executive misconduct to loss oflibrary books (Kadet, 2002). Such applicationsare unlikely to impact the actual sources of anydisaffection, but, rather, open the organizationto claims of injury by providing an interventionwidely claimed to address traumatic sequella ofinjurious exposures. The employer has then, byproviding or endorsing such intervention, effec-tively stipulated that a workplace event has oc-curred from which psychiatric morbidity canreasonably be foreseen as one possible out-come. Unfortunately, the intervention profferedhas shown no efficacy in preventing such im-pacts and some potential to retard their normalresolution. Rather than representing a protectivemeasure respecting organizational liability, itbecomes instead a recipe for claims and litiga-tion.

Service integration. Most crisis interven-tion plans have focused on establishing helpdelivery systems for exporting more or lesstraditionally conceived intervention objectivesthrough relocated and repackaged interventionmechanisms. However, it is likely to prove farmore productive to focus instead on the helpseeking patterns of affected persons and groups,endeavoring to enhance the capacity of estab-lished relationships and to extend and expandtheir impact to address psychological and socialwelfare needs borne of the crisis experience(Yates, Axom, Bickman, & Howe, 1989; Yates,Axom, & Tiedeman, 1999).

Most people, for example, first crave solid,reliable information following a traumaticevent—questions such as “What exactly hap-pened? Who has been affected? How badly?What is being done? What will happen tomor-row?” all require accurate, reliable answers


from recognizable resources who can win andmaintain trust. From the perspective of the cli-ent organization (as well as from that of theemployee), this is much more effective ifbranded to the structure and culture of the or-ganization itself. Such information should bemade available as soon as possible and throughan ordered and official process. At a communitylevel, this is frequently achieved through a toll-free “hotline” service, providing details, practi-cal information, and support as informationcomes to light; at an organizational level, thiswould require regular and official meetings withthe staff, designated organizational spokesper-sons and contacts, and other pertinent vehiclesof communication. This can, of course, includereferral options for practical or emotional sup-port.

Practical instrumental aid. Practical con-siderations may include facilitating family sup-port and the provision of facilities which ensurethat people have access to needed services topromote physical safety, health care, sanitation,and, of course, food and comfort. In addition tothe necessity of such sustenance, the gesturesimplicit in these acts can be seen as increasingboth received and perceived social support andas maintaining social embeddedness, factors as-sociated with improved long-term outcome(Kaniasty & Norris, 1993; Kaniasty, Norris, &Murrell, 1990; Norris, 2001). They also serve toprovide a foundation for an enhanced sense ofconnectedness and commitment between orga-nization and employee as the crisis progressestoward resolution.

Establishing contacts and relationships.The first defined response objective, rather thanprophylactic intervention, might be more pro-ductively conceived as establishing contacts andinitiating relationships that can later serve toenhance perceptions of familiarity and access toservices as ongoing recovery issues emerge.When practical, instrumental activities, such asthose outlined above, are facilitated or assistedby representatives of those organizational com-ponents or adjuncts that are likely to be ac-cessed at future junctures, the threshold for ini-tiating later contact is likely to be loweredsomewhat through such prior familiarity. Inother words, in order to receive support, peoplewill seek out those they are already familiarwith and will do so with fewer reservations.Since most recovery issues emerge well beyond

the 24–72 hour window commonly associatedwith debriefing exercises, such contacts and re-lationships are likely to prove more significantin facilitating longer term outcomes than brief,intensive early contacts with providers of ser-vices directed toward needs not yet fully real-ized as necessary or salient.

Resiliency focus. Sound organizational con-sultation and assistance generally works moreeffectively when it enlists established organiza-tional structures and agents to promote resil-ience in both individuals and the organizationalculture. Resilience is the natural human processof positive adaptation in the face of adversity,trauma, tragedy, or stress. Individuals cope withtrauma in varying ways and with varying de-grees of success, but perhaps the most consis-tent finding in disaster research is that the vastmajority of individuals recover from a traumaticexperience without experiencing significantpsychopathology (cf. Cook & Bickman, 1990;Rubonis & Bickman, 1991; Salzer & Bickman,1999).

It has been suggested that the lack of efficacyof PD might be explained at least in part by itsinterference with the natural processing of atraumatic event, and by inadvertently leadingvictims to circumvent the support of family,friends, or other sources of social support—aprimary factor in resilience—in favor of a mis-guided notion that professional help is more aptto aid their resolution (van Emmerik et al.,2002). Rather than importing an interventionstructure to be imposed upon the organizationand circumstance, competent organizational ad-juncts may function as agents and extenders ofthe established organizational structure, bothformal and informal, to empower its effectiveaddress of employee and organizational needs.

Early assessment and referral. While iden-tification of those in need of referral is an oftenstated objective of debriefing, such early at-tempts at predicting morbidity have proven nei-ther sensitive nor specific. However, there isemerging evidence relating to treatment forthose who go on to develop pathological reac-tions, such as ASD. For example, about 80% ofthose who have ASD progress on to developPTSD at 6 months (75% at two years) if they areleft untreated, although it should also be notedthat 70% of “subclinical” ASD (i.e., those thatdid not meet all the dissociative criteria) had


PTSD at two year follow-up (Bryant & Harvey,2000).

The currently most replicated treatment resultis that CBT treatments (particularly those in-volving exposure to corrective information) ev-idence the best gains, with around 8% (at posttreatment) and 17% (at 6 month follow-up)meeting criteria for PTSD (Bryant et al., 1998,1999; Foa et al., 1995). Therefore, once some-one has been identified who a few days to fourweeks later is still experiencing inordinate dis-tress due to the event, care should be taken toprovide a tiered structure that involves theseindividuals in more structured, individual andindividualized treatment.

However, yet further research into more spe-cific assessment approaches applied at aboutfour to six weeks postimpact (when reactionstend to stabilize) is now showing promisingresults (Brewin et al., 2002). These approachesallow those at increased risk to be identifiedwithout extensive or intrusive testing, but stillwith reasonably high levels of specificity andsensitivity, through simple inquiries scored byrote tally of affirmative responses. Such ap-proaches can be delivered many ways with theobjective of directing those at demonstrablyheightened risk toward competent and effectiveassistance.

Reasonably strong evidence is accumulatingregarding treatments of choice for the minoritywho will ultimately require focused profes-sional intervention, as are reliable indicatorsregarding timelines for such intervention. Bestevidence supports short course CBT (betweenfive and nine sessions, depending upon diagnos-tic presentation) using graded exposure, com-mencing two to six weeks after exposure (seeLitz et al., 2002, for an overview). This requiresadvanced training and credentialing, but iswidely available among legitimate psychologi-cal providers.

Stepped care. Most responsible systemplanners are now exploring strategies based onstepped care—approaches in which graded lev-els of assistance are made available based onongoing assessments of individual and collec-tive needs (cf. Bisson, 2001). These generallyhinge on effectively addressing issues such asaccessibility, affordability, and applicability ofassistance with respect to the divergent needs ofparticular individuals and groups. Uniformity oftreatment protocol ultimately proves contrary to

treatment efficacy and is avoided in favor ofimproved assessment and carefully targeted as-sistance. This requires ongoing monitoring ofissues and impediments and ongoing organiza-tional attention to the resolution process overtime.

Ongoing accessible support. Resolution isan ongoing process that takes place over monthsand even years, rather than hours, days, orweeks. Once again, intensive early interventionthat dissipates and disappears just as recovery isbeginning cannot effectively address the ripen-ing of concerns that occurs across the course ofresolution (Gist & Lubin, 1999). One promisinginnovation has been the beta testing of a resil-iency focused website providing a wide range ofinformation, empirically supported self-helpstrategies, referral contacts, and such for use bycorporate employees following workplace cri-ses (Crisis Management International, 2003).The site is branded when accessed to the orga-nization contracting for consultation and assis-tance, and, if available and indicated, to itscontracted EAP. Availability is designed to con-tinue for one year, with specified periodic fol-low-up contacts for those providing specificconsent. Mechanisms are provided for employ-ees to track their own resolution over time, withlocal referral information available wheneverindicated or desired.

Moving Forward

The broader challenge emerging from thesedebates and investigations centers most aroundthe peculiarly daunting task of refocusing andredirecting the energy of a charismatic socialmovement toward this somewhat more circum-spect view of how best to assist. While theserious literature of psychology and related dis-ciplines has yielded strong warnings regardingtraditional debriefing practices and a range ofindications regarding more viable and effectivealternatives, information flowing to the purvey-ors and consumers of debriefing has been al-most solely in the domain and control of inter-vention proponents. The result has been con-flicting bodies of information, one presentingobjective, refereed, independent assessments ofmeasured efficacy and the other dominated by asocial movement attempting to argue those ac-cumulating data away. The fact that such anintense discrepancy not only continues, but has


indeed intensified over time is a discomfitingindicator of a schism between research andpractice that grows progressively more trou-bling.

If, as is often asserted, academic psycholo-gists become detached from the realities of ap-plication and practice, it also seems evident thatmany practitioners have become progressivelymore estranged from the empirical underpin-nings of their discipline. This would be seriousenough in itself, were it not for an emergingcaste of nouveau trauma responders who, evermore pointedly, eschew empirical analysis withthe assumption that “clinical intuition” is notonly sufficient evidence, but is indeed somehowsuperior (Gist, Woodall, & Magenheimer,1999). As a consequence, many risk mistakingappreciation of responsiveness for efficacy ofresponse, misapprehending the nonspecific im-pact of a concerned presence as if some specificimpact of a routinized process, and confusingthe illusory correlation between early activityand subsequent natural recovery with a quanti-tative indication of effect. These foibles becomeall the more difficult to discern when repetitionof “accepted practice” supercedes the cautiousand objective reporting of controlled research inthe information venues most directly accessedby providers and consumers.

Reconciliation becomes all the more difficultwhere the majority of adherents fall outsideeven the progressively eroding boundaries ofregulated psychological practice. Mental healthprofessionals are defined by CISD prescripts asessentially anyone with a master’s degree ineven a remotely related field (Mitchell, 1983 etseq.), and the process is, in many settings, dom-inated by “peer” providers with no more prep-aration than attendance at a few days of propri-etary workshops at which no mechanisms orstandards for evaluating competence or perfor-mance are advocated, much less applied.

Neither researchers nor practitioners can es-cape accountability by displacing blame to theother. We share, at the end of the day, a collec-tive responsibility to ensure that what assistancewe offer indeed translates to meaningful, mea-surable help for those who avail themselves ofour ministrations; at the very least, we share amoral obligation that first, we shall do no harm.Our intentions are surely quite honorable andour efforts strong and sincere, but it is theoutcomes that ultimately matter and it is those

outcomes that must therefore be subjected toour strongest empirical scrutiny.

Objective empirical analysis remains ourbest, if not our only, hope to hear the murmur-ing of Nature amid the din and clamor of ourown interests, motives, expectations, and needs.The entire scientist-practitioner archetype isconstructed on a fundamental belief that ourknowledge is always imperfect and our tech-niques perpetually evolve. It is that essentialpremise which drives the evolution of empiri-cally based practice, one in which research in-forms practice and application refines theory.

Caveat emptor — “Let the buyer beware”—may be a defensible tenet when selling used carsor real estate, but the purveyor of professionalservices, whose authority in the eyes of a trust-ing laity is couched in the presumption of sci-entific rigor and responsibility, must hold him-self or herself to a much higher standard.Should that fail, the profession in aggregatemust step forward to ensure that its integrity ismaintained. Distressingly, however, this has yetto happen with respect to the debriefing debates.

References

Alexander, D. A., & Wells, A. (1991). Reactions ofpolice officers to body-handling after a major di-saster: A before-and-after comparison. BritishJournal of Psychiatry, 159, 547–555.

Andre, C., Lelord, F., Legeron, P., Reignier, A., &Delattre, A. (1997). Effectiveness of early inter-vention on 132 bus drivers’ victims of aggressions:A controlled study. Encephale, 23, 65–71.

Armstrong, K., Zatzick, D., Metzler, T., Weiss, D. S.,Marmar, C. R., Garma, S., et al. (1998). Debriefingof American Red Cross personnel: Pilot study onparticipants’ evaluations and case examples fromthe 1994 Los Angeles earthquake relief operation.Social Work in Health Care, 27, 33–50.

Beaton, R., & Murphy, S. (1993). Sources of occu-pational stress among firefighter/EMTs and fire-fighter/paramedics and correlations with job-related outcomes. Prehospital and Disaster Medi-cine, 8, 140–150.

Bierens de Haan, B. (1998). Emotional group de-briefing of humanitarian aid workers: The experi-ence of ICRC. Schweizer Archiv fuer Neurologieund Psychiatrie, 149, 218–228.

Bisson, J., Jenkins, P., Alexander, J., & Bannister, C.(1997). Randomised controlled trial of psycholog-ical debriefing for victims of acute burn trauma.British Journal of Psychiatry, 171, 78–81.


Bisson, J. I., McFarlane, A. C., & Rose, S. (2000).Psychological debriefing. In E. B. Foa, T. M.Keane, & M. J. Freidman (Eds.), Effective treat-ments for PTSD (pp. 39–59). New York: GuilfordPress.

Bledsoe, B. E. (2002). CISM: Possible liability forEMS services? Best Practices in Emergency Med-ical Services, 5, 66–67.

Bledsoe, B. E. (2003). Critical Incident Stress Man-agement (CISM): Benefit or risk for emergencyservices? Prehospital Emergency Care, 7, 272–279.

Blythe, B. T. (2002). Blindsided: A manager’s guideto catastrophic incidents in the workplace. NewYork: Portfolio.

Bohl, N. (1991). The effectiveness of brief psycho-logical interventions in police officers after criticalincidents. In J. Reese, J. Horne, & C. Dunning(Eds.), Critical incidents in policing (pp. 31–88).Washington, DC: U.S. Government Printing Of-fice.

Bordrow, S., & Porritt, D. (1979). An experimentalevaluation of crisis intervention. Social Scienceand Medicine, 13, 251–256.

Borman, W. C., & Motowidlo, S. J. (1993). Expand-ing the criterion domain to include elements ofcontextual performance. In N. Schmitt, W. C. Bor-man, & Associates (Eds.), Personnel selection inorganisations (pp. 25–38). San Francisco: Jossey-Bass.

Brailey, K., Vasterling, J. J., & Sutker, P. B. (1998).Psychological aftermath of participation in the Per-sian Gulf War. In A. Lundberg (Ed.), The environ-ment and mental health: A guide for clinicians.Mahwah, NJ: Erlbaum.

Brewin, C. R., Rose, S., Andrews, B., Green, J., Tata,P., McEvedy, C., et al. (2002). A brief screeninginstrument for post-traumatic stress disorder. Brit-ish Journal of Psychiatry, 181, 158–162.

Brickman, A. L., Garrity, C. P., & Shaw, J. A.(2002). Risk factors for psychosocial dysfunctionamong enrollees in the state children’s health in-surance program. Psychiatric Services, 53, 614–619.

Brom, D., Kleber, R. J., & Hofman, M. C. (1993).Victims of traffic accidents: Incidence and preven-tion of post-traumatic stress disorder. Journal ofClinical Psychology, 49, 131–140.

Bryant, R. A., & Harvey, A. G. (2000). Acute StressDisorder. Washington DC: American Psychologi-cal Association.

Bryant, R. A., Harvey, A. G., Dang, S. T., Sackville,T., & Basten, C. (1998). Treatment of acute stressdisorder: A comparison of cognitive-behavioraltherapy and supportive counseling. Journal ofConsulting and Clinical Psychology, 66, 862–866.

Bryant, R. A., Sackville, T., Dang, S. T., Moulds, M.,& Guthrie, R. (1999). Treating acute stress disor-

der: An evaluation of cognitive behavior therapyand supportive counseling techniques. AmericanJournal of Psychiatry, 156, 1780–1786.

Bunch, K. P., & Wilson, T. G. (2002). Counselors oncall. Fire Chief, 46.

Carlier, I. V. E., Lamberts, R. G., van Uchlen, A. J.,& Gersons, B. P. R. (1998). Disaster related post-traumatic stress in police officers: A field study ofthe impact of debriefing. Stress Medicine, 14, 143–148.

Carlier, I. V. E., Voerman, A. E., & Gersons, B. P. R.(2000). The influence of occupational debriefingon post-traumatic stress symptomatology in trau-matized police officers. British Journal of MedicalPsychology, 73, 87–98.

Charlton, P. F. C., & Thompson, J. A. (1996). Waysof coping with psychological distress after trauma.British Journal of Clinical Psychology, 35, 517–530.

Chemtob, C. M., Thomas, S., Law, W., & Cremniter,D. (1997). Postdisaster psychosocial intervention:A field study of the impact of debriefing on psy-chological distress. American Journal of Psychia-try, 154, 415–417.

Cohen, J. (1992). A power primer. PsychologicalBulletin, 112, 155–159.

Conlon, L., Fahy, T., & Conroy, R. (1999). PTSD inambulant RTA victims: Prevalence, predictors anda randomised controlled trial of psychological de-briefing in prophylaxis. Journal of PsychosomaticResearch, 46, 37–44.

Cook, J. D., & Bickman, L. (1990). Social supportand psychological symptomatology following anatural disaster. Journal of Traumatic Stress, 3,541–556.

Creamer, M., Burgess, P., Buckingham, W., & Pat-tison, P. (1989). The psychosocial aftermath of theQueen Street shootings. Melbourne: University ofMelbourne, Department of Psychology.

Creamer, M., Burgess, P., & McFarlane, A. C.(2001). Post Traumatic Stress Disorder: Findingsfrom the Australian National Survey of MentalHealth and Well-being. Psychological Medi-cine, 31, 1237–1247.

Cremniter, D., Crocq, L., Louville, P., Batista, G.,Grande, C., Lambert, Y., et al. (1997). Posttrau-matic reactions of hostages after an aircraft hijack-ing. Journal of Nervous and Mental Disease, 185,344–346.

Crisis Management International. (2003). “BouncingBack” website prototype. Retrieved on November11, 2006 from http://www.bouncing-back.com

Davey, G. C. L. (1993). Trauma revaluation, condi-tioning, and anxiety disorders. BehaviourChange, 10, 131–140.

Deahl, M. P., Gillham, A. B., Thomas, J., Dearle,M. M., & Strinivasan, M. (1994). Psychologicalsequelae following the Gulf war: Factors associ-


ated with subsequent morbidity and the effective-ness of psychological debriefing. British Journalof Psychiatry, 165, 60–65.

Derogatis, L. R. (1992). SCL-90: Administration,scoring and procedures manual-II. Baltimore:Clinical Psychometric Research.

Devilly, G. J. (2002). Clinical intervention, support-ive counseling and therapeutic methods: A clarifi-cation and direction for restorative treatment. In-ternational Review of Victimology, 9, 1–14.

Devilly, G. J. (2005). The power therapies and pos-sible threats to the science of psychology & psy-chiatry. Australian & New Zealand Journal ofPsychiatry, 39, 437–445.

Devilly, G. J. & Annab, R. (in press). A randomisedcontrolled trial of group debriefing. Journal ofBehavior Therapy & Experimental Psychiatry.

Devilly, G. J., & Borkovec, T. D. (2000). Psycho-metric properties of The Credibility/ExpectancyQuestionnaire. Journal of Behavior Therapy andExperimental Psychiatry, 31, 73–86.

Devilly, G. J., & Cotton, P. (2003). Psychologicaldebriefing and the workplace: Defining a concept,controversies and guidelines for intervention. Aus-tralian Psychologist, 38, 144–150.

Devilly, G. J., & Spence, S. H. (1999). The relativeefficacy and treatment distress of EMDR and acognitive behavior Trauma Treatment Protocol inthe amelioration of Posttraumatic Stress Disorder.Journal of Anxiety Disorders, 13, 131–157.

Devilly, G. J., Varker, T., Hansen, K., & Gist, R. (inpress). An analogue study of the effects of psycho-logical debriefing on eyewitness memory. Behav-iour Research & Therapy.

Doctor, R. S., Cutris, D., & Isaacs, G. (1994). Psy-chiatric morbidity in policemen and the effect ofbrief psychotherapeutic intervention: A pilotstudy. Stress Medicine, 10, 151–157.

Everly, G. S., & Boyle, S. (1997). A meta-analysis ofthe Critical Incident Stress Debriefing (CISD). Pa-per presented at the Fourth World Congress onStress, Trauma and Coping in the Emergency Ser-vices Professions, Baltimore, MD.

Everly, G. S., Flannery, R. B., & Mitchell, J. T.(2000). Critical incident stress management(CISM): A review of the literature. Aggression andViolent Behavior, 5, 23–40.

Everly, G. S., & Mitchell, J. T. (1997). CriticalIncident Stress Management (CISM): A new eraand standard of care in crisis intervention. EllicottCity, MD: Chevron.

Foa, E. B., Dancu, C. V., Hembree, E., Jaycox, L. H.,Meadows, E. A., & Street, G. P. (1999). Theefficacy of Exposure Therapy, Stress InoculationTraining and their combination in amelioratingPTSD for female victims of assault. Journal ofConsulting and Clinical Psychology, 67, 194–200.

Foa, E. B., Hearst-Ikeda, D., & Perry, K. J. (1995).Evaluation of a brief cognitive-behavioral programfor the prevention of chronic PTSD in recent as-sault victims. Journal of Consulting and ClinicalPsychology, 63, 948–955.

Foa, E. B., & Kozac, M. J. (1986). Emotional pro-cessing of fear: Exposure to corrective informa-tion. Psychological Bulletin, 99, 20–35.

Foa, E. B., & Meadows, E. A. (1998). Psychosocialtreatments for Posttraumatic Stress Disorder: Acritical review. Annual Review of Psychology, 48,449–480.

Foa, E. B., Rothbaum, B. O., Riggs, D. S., & Mur-doch, T. B. (1991). Treatment of PosttraumaticStress Disorder in rape victims: A comparisonbetween cognitive-behavioral procedures andcounseling. Journal of Consulting and ClinicalPsychology, 59(5), 715–723.

Ford, J. D., Shaw, D., Sennhauser, S., Greaves, D.,Thacker, B., Chandler, P., et al. (1993). Psycho-social debriefing after operation desert storm: Mar-ital and family assessment and intervention. Jour-nal of Social Issues, 49, 73–102.

Fullerton, C. S., Ursano, R. J., Vance, K., & Wang, L.(2000). Debriefing following trauma. PsychiatricQuarterly, 71, 259–276.

Galea, S., Ahern, J., Resnick, H., Kilpatrick, D.,Bucuvalas, M., Gold, J., et al. (2002). Psycholog-ical sequelae of the September 11 terrorist attacksin New York City. New England Journal of Med-icine, 346, 982–987.

George, J. M. (1989). Mood and absence. Journal ofApplied Psychology, 74, 317–324.

George, J. M. (1996). Trait and state affect. In K. R.Murphy (Ed.), Individual differences and behav-iour in organisations, (pp. 145–171). San Fran-cisco: Jossey-Bass.

Gilbert, D. T., & Silvera, D. H. (1996). Overhelping.Journal of Personality and Social Psychology, 70,678–690.

Gist, R. (2002). What have they done to my song?Social science, social movements, and the debrief-ing debates. Cognitive and Behavioral Practice, 9,273–279.

Gist, R., & Devilly, G. J. (2002). Post-trauma de-briefing: The road too frequently travelled. TheLancet, 360, 741–742.

Gist, R., Lohr, J. M., Kenardy, J., Bergmann, L.,Meldrum, L., Redburn, B. G., et al. (1997). Re-searchers speak on CISM. Journal of EmergencyMedical Services, 22(5), 27–28.

Gist, R., & Lubin, B. (1999). Epilogue. In R. Gist &B. Lubin (Eds.), Response to disaster: Psychoso-cial, community, and ecological approaches (pp.347–352). Philadelphia: Brunner/Mazel.

Gist, R., Lubin, B., & Redburn, B. G. (1999). Psy-chosocial, ecological, and community perspectiveson disaster response. In R. Gist & B. Lubin (Eds.),


Response to disaster: Psychosocial, community,and ecological approaches (pp. 1–27). Philadel-phia: Brunner/Mazel.

Gist, R., & Woodall, S. J. (1995). Occupational stressin contemporary fire service. Occupational Medi-cine: State of the Art Reviews, 10, 763–787.

Gist, R., & Woodall, S. J. (1999). There are no simplesolutions to complex problems: The rise and fall ofcritical incident stress debriefing as a response tooccupational stress in the fire service. In R. Gist &B. Lubin (Eds.), Response to disaster: Psychoso-cial, community, and ecological approaches (pp.211–235). Philadelphia: Brunner/Mazel.

Gist, R., Woodall, S. J., & Magenheimer, L. K.(1999). And then you do the Hokey-Pokey and youturn yourself around. . . In R. Gist & B. Lubin(Eds.), Response to disaster: Psychosocial, com-munity, and ecological approaches (pp. 269–292).Philadelphia: Brunner/Mazel.

Gump, B. B., & Kulik, J. A. (1997). Stress, affilia-tion, and emotional contagion. Journal of Person-ality and Social Psychology, 72, 305–319.

Hammarberg, M. (1992). Penn Inventory for Post-traumatic Stress Disorder: Psychometric proper-ties. Psychological Assessment, 4, 67–76.

Hart, P. M., & Cotton, P. (2003). Conventional wis-dom is often misleading: Police stress within anorganisational health framework. In M. F. Dollard,A. H. Winefield, & H. R. Winefield (Eds.), Occu-pational stress in the service professions, (pp.103–140). London: Taylor and Francis.

Hart, P. M., Wearing, A. J., & Headey, B. (1994).Perceived quality of life, personality, and workexperiences: Construct validation of the PoliceDaily Hassles and Uplifts Scales. Criminal Justiceand Behavior, 21, 283–311.

Harvey, A. G., & Bryant, R. A. (1998). The relation-ship between acute stress disorder and Posttrau-matic Stress Disorder: A prospective evaluation ofmotor vehicle accident survivors. Journal of Con-sulting and Clinical Psychology, 66, 507–512.

Helzer, J. E., Robins, L. N., & McEvoy, L. (1987).Post-traumatic stress disorder in the general pop-ulation: Findings of the Epidemiologic CatchmentArea survey. New England Journal of Medicine,317, 1630–1634.

Hobbs, G., & Adshead, G. (1997). Preventive psy-chological intervention for road crash victims. InM. Mitchell (Ed.), The aftermath of road acci-dents: Psychological, social and legal perspectives(pp. 159–171). London: Routledge.

Hobbs, M., Mayou, R., Harrison, B., & Worlock, P.(1996). A randomised controlled trial of psycho-logical debriefing for victims of road traffic acci-dents. British Medical Journal, 313, 1438–1439.

Hoffman, R. R. (2001). The relation of optimism andself-mastery to responses to unfavorable contin-gency changes. Dissertation Abstracts Interna-

tional: Section B: The Sciences and Engineering,61(9), 5055B.

Hovens, J. E., & Van de Weerd, M. (1998). Posttrau-matic Stress Disorder in debriefed survivors of abus accident. Psychological Reports, 82, 1075–1081.

International Critical Incident Stress Foundation.(2003). Education and Training. Retrieved onMay 12, 2003 from http://www.icisf.org/Training/classes/index.asp

Jenkins, S. R. (1996). Social support and debriefingefficacy among emergency medical workers after amass shooting incident. Journal of Social Behaviorand Personality, 11, 477–492.

Jones, N., Roberts, P., & Greenberg, N. (2003). Peer-group risk assessment: A post-traumatic manage-ment strategy for hierarchical organizations. Occu-pational Medicine, 53, 469–475.

Kadet, A. (2002). Good grief! Smart Money, 11,108–114.

Kaniasty, K., & Norris, F. (1993). A test of thesupport deterioration model in the context of nat-ural disaster. Journal of Personality and SocialPsychology, 64, 395–408.

Kaniasty, K., Norris, F. H., & Murrell, S. A. (1990).Perceived and received social support followingnatural disaster. Journal of Applied Social Psy-chology, 20, 85–114.

Kenardy, J. (2000). The current status of psycholog-ical debriefing: It may do more harm than good.British Medical Journal, 321, 1032–1033.

Kenardy, J., & Taylor, C. B. (1999). Expected versusunexpected panic attacks: A naturalistic prospec-tive study. Journal of Anxiety Disorders, 13, 435–445.

Kenardy, J. A., Webster, R. A., Lewin, T. J., Carr,V. J., Hazell, P. L., & Carter, G. L. (1996). Stressdebriefing and patterns of recovery following anatural disaster. Journal of Traumatic Stress, 9,37–49.

Kessler, R. C., Sonnega, A., Bromet, E., Hughes, M.,& Nelson, C. B. (1995). Posttraumatic Stress Dis-order in the national comorbidity survey. Archivesof General Psychiatry, 52, 1048–1060.

Kessler, R. C., Zhao, S., Katz, S. J., Kouzis, A. C.,Frank, R. G., Edlund, M., et al. (1999). Past-yearuse of outpatient services for psychiatric problemsin the National Comorbidity Survey. AmericanJournal of Psychiatry, 156, 115–123.

Keuler, D. J., & Safer, M. A. (2001). Memory bias inthe assessment and recall of pre-exam anxiety:How anxious was I? Applied Cognitive Psychol-ogy, 12, S127–S137.

Kierein, N. M., & Gold, M. A. (2000). Pygmalion inwork organizations: A meta-analysis. Journal ofOrganizational Behavior, 21, 913–928.

Kulik, J. A., Mahler, H. I. M., & Moore, P. J. (1996).Social comparison and affiliation under threat: Ef-


fects on recovery from major surgery. Journal ofPersonality and Social Psychology, 71, 967–979.

Lee, C., Slade, P., & Lygo, V. (1996). The influenceof psychological debriefing on emotional adapta-tion in women following early miscarriage: A pre-liminary study. British Journal of Medical Psy-chology, 69, 47–58.

Litz, B. T., Adler, A. B., Castro, C. A., Wright, K., &Thomas, J., & Suvak, M. K. (2004, November). Acontrolled trial of Group Debriefing. In M. Fried-man (Chair), Military psychiatry, then and now.Plenary session conducted at the 20th AnnualMeeting of the International Society for TraumaticStress Studies, New Orleans, LA.

Litz, B. T., Gray, M. J., Bryant, R. A., & Adler, A. B.(2002). Early intervention for trauma: Current sta-tus and future directions. Clinical Psychology: Sci-ence and Practice, 9, 112–134.

Lohr, J. M., Hooke, W., Gist, R., & Tolin, D. F.(2003). Novel and controversial treatments fortrauma-related stress disorders. In S. O. Lilienfeld,S. J. Lynn, & J. M. Lohr (Eds.), Science andpseudoscience in clinical psychology (pp. 243 –272). New York: Guilford Press.

Lopez, G., Piffaut, G., & Seguin, A. (1992). Psycho-logical treatment of victims of rape. PsychologieMedicale, 24, 286–288.

Macnab, A. J., Russell, J. A., Lowe, J. P., & Gagnon,F. (1998). Critical incident stress intervention afterloss of an air ambulance: Two-year follow up.Prehospital and Disaster Medicine, 14, 8–12.

Madon, S., Smith, A., Jussim, L., Russell, D. W.,Eccles, J., Palumbo, P., et al. (2001). Am I as yousee me or do you see me as I am? Self-fulfillingprophecies and self-verification. Personality andSocial Psychology Bulletin, 27, 1214–1224.

Major, B., Cozzarelli, C., Sciacchitano, A. M., Coo-per, M. L., & Testa, M. (1990). Perceived socialsupport, self-efficacy, and adjustment to abortion.Journal of Personality and Social Psychology, 59,452–463.

Matthews, L. R. (1998). Effect of staff debriefing onposttraumatic stress symptoms after assaults bycommunity housing residents. Psychiatric Ser-vices, 49, 207–212.

Mayou, R. A., Ehlers, A., & Hobbs, M. (2000).Psychological debriefing for road traffic accidentvictims: Three-year follow-up of a randomizedcontrolled trial. British Journal of Psychiatry, 176,589–593.

McCarroll, J. E., Fullerton, C. S., Ursano, R. J., &Hermsen, J. M. (1996). Posttraumatic stress symp-toms following forensic dental identification: Mt.Carmel, Waco, TX. American Journal of Psychi-atry, 153, 778–782.

McCarroll, J. E., Ursano, R. J., & Fullerton, C. S.(1993). Symptoms of Posttraumatic Stress Disor-

der following recovery of war dead. AmericanJournal of Psychiatry, 150, 1875–1877.

McCarroll, J. E., Ursano, R. J., Fullerton, C. S., &Lundy, A. (1993). Traumatic stress of a wartimemortuary: Anticipation of exposure to mass death.Journal of Nervous and Mental Disease, 181, 545–551.

McCarroll, J. E., Ursano, R. J., Ventis, W. L., &Fullerton, C. S. (1993). Anticipation of handlingthe dead: Effects of gender and experience. BritishJournal of Clinical Psychology, 32, 466–468.

McCrae, R. R. (1984). Situational determinants ofcoping responses: Loss, threat, and challenge.Journal of Personality and Social Psychology, 46,919–928.

McFarlane, A. C. (1988). The longitudinal course ofposttraumatic morbidity: The range of outcomesand their predictors. Journal of Nervous and Men-tal Disease, 176, 30–39.

McFarlane, A. C., Bookless, C., & Air, T. (2001).Posttraumatic Stress Disorder in a general psychi-atric inpatient population. Journal of TraumaticStress, 14, 633–645.

McNally, R. J., Bryant, R., & Ehlers, A. (2003). Doesearly psychological intervention promote recoveryfrom post traumatic stress. Psychological Sciencein The Public Interest, 4, 45–79.

McNally, R. J., Luedke, D. L., Besyner, J. K., Peter-son, R. A., Bohm, K., & Lips, O. J. (1987). Sen-sitivity to stress-relevant stimuli in Post-TraumaticStress Disorder. Journal of Anxiety Disorders, 1,105–116.

Mitchell, J. T. (1983). When disaster strikes . . . Thecritical incident stress debriefing process. Journalof Emergency Services, 8, 36–39.

Mitchell, J. T. (1992). Protecting your people fromcritical incident stress. Fire Chief, 36, 61–67.

Mitchell, J. T. (2003). Crisis intervention and CISM:A research summary. Retrieved on May 12, 2003,from http://www.icisf.org/articles/cism_research_summary.pdf

Mitchell, J. T., & Everly, G. S. (1997). The scientificevidence for critical incident stress management.Journal of Emergency Medical Services, 22(1),86–93.

Mitchell, J. T., & Everly, G. S. (2000). Critical inci-dent stress management and critical incident stressdebriefings: Evolutions, effects and outcomes. InB. Raphael & J. P. Wilson (Eds.), Psychologicaldebriefing: Theory, practice and evidence. NewYork: Cambridge University Press.

NATO: North Atlantic Treaty Organization. (2002).NATO-Russia advanced research workshop on so-cial and psychological consequences of chemical,biological and radiological terrorism. Retrieved onAugust, 2006 from http://www.nato.int/docu/update/2002/03-march/e0325a.htm


Norris, F. H. (2001, September). 50,000 disaster vic-tims speak: An empirical review of the empiricalliterature, 1981–2001. Unpublished manuscript,The National Center for PTSD and The Center forMental Health Services (SAMHSA).

Norris, F. H., Kaniasty, K., & Thompson, M. P.(1997). The psychological consequences of crime.In R.C. Davis, A. J. Lurigo, & W. G. Skogan(Eds.), Victims of crime (2nd ed.). Thousand Oaks,CA: Sage.

Nurmi, L. A. (1997). Experienced stress and value ofCISD among Finnish police officers (DVI) andemergency personnel in the Estonia ferry disaster.Paper presented at the Fourth World Congress onStress, Trauma and Coping in the Emergency Ser-vices Professions, Baltimore, MD.

Nurmi, L. A. (1999). The sinking of the Estonia: Theeffects of Critical Incident Stress Debriefing(CISD) on rescuers. International Journal ofEmergency Mental Health, 1, 23–31.

Parry, G. (2001). Evidence based clinical practiceguidelines for treatment choice in psychologicaltherapies and counselling. London: UK Depart-ment of Health.

Raphael, B. (1986). When disaster strikes: A hand-book for caring professionals. London: Hutchin-son.

Redburn, B. G. (1992). Disaster and rescue: Workereffects and coping strategies. Dissertation Ab-stracts International, 54(01-B), 447.

Redburn, B. G., Gensheimer, L. K., & Gist, R. (1993,June). Disaster aftermath: Social support amongresilient rescue workers. Paper presented at theFourth Biennial Conference on Community Re-search and Action, Society for Community Re-search and Action (Division 27, American Psycho-logical Association), Williamsburg, VA.

Ritchie, E. C. (2001). Draft consensus document ofthe mass violence and early intervention work-shop. Washington, DC: U.S. Department of De-fense.

Robinson, R., & Mitchell, J. T. (1993). Evaluation ofpsychological debriefings. Journal of TraumaticStress, 6, 367–382.

Rose, S., Brewin, C. R., Andrews, B., & Kirk., M.(1999). A randomized controlled trial of individualpsychological debriefing for victims of violentcrime. Psychological Medicine, 29, 793–799.

Rose, S., Wessely, S., & Bisson, J. (2001). Briefpsychological interventions (“debriefing”) for trau-ma-related symptoms and prevention of Post Trau-matic Stress Disorder: Review. The Cochrane Li-brary, Issue 2.

Rubonis, A. V., & Bickman, L. (1991). Psychologi-cal impairment in the wake of disaster: The disas-ter-psychopathology relationship. PsychologicalBulletin, 109, 384–399.

Salzer, M. S., & Bickman, L. (1999). The short andlong term psychological impact of disasters: Im-plications for mental health interventions and pol-icy. In R. Gist & B. Lubin (Eds.), Response todisaster: Psychosocial, community, and ecologicalapproaches (pp. 63–82). Washington, DC: Taylor& Francis.

Shalev, A. Y., Peri, T., Rogel-Fuchs, Y., Ursano,R. J., & Marlowe, D. (1998). Historical groupdebriefing after combat exposure. Military Medi-cine, 163, 494–498.

Sijbrandij, M. (2002, November). A closer look atdebriefing: Emotional ventilation vs. psychoeduca-tion. In M. Oliff & R. Bryany (Chairs), Earlyinterventions: New contributions to outcome re-search. Symposium conducted at the 18th AnnualMeeting of the International Society for TraumaticStress Studies, Baltimore, MD.

Solomon, Z., Mikulincer, M., & Benbenishty, R.(1989). Locus of control and combat-related post-traumatic stress disorder: The intervening role ofbattle intensity, threat appraisal and coping. BritishJournal of Clinical Psychology, 28, 131–144.

Staab, J. P., Fullerton, C. S., & Ursano, R. (1999). Acritical look at PTSD: Constructs, concepts, epide-miology, and implications. In R. Gist, & B. Lubin(Eds.), Response to disaster: Psychosocial, com-munity, and ecological approaches. Philadelphia,PA: Brunner/Mazel.

Stallard, P., Velleman, R., & Baldwin, S. (2000).Prospective study of post-traumatic stress disorderin children involved in road traffic accidents. Brit-ish Medical Journal, 7173, 1619–1623.

Taylor, S. E. (1983). Adjustment to threateningevents: A theory of cognitive adaptation. AmericanPsychologist, 38, 1161–1173.

Taylor, S. E. (1991). Asymmetrical effects of posi-tive and negative events: The mobilization-minimization hypothesis. Psychological Bulletin,110, 67–85.

Taylor, S. E., & Brown, J. D. (1988). Illusion andwell-being: A social psychological perspective onmental health. Psychological Bulletin, 103, 193–210.

Taylor, S. E., & Lobel, M. (1989). Social comparisonactivity under threat: Downward evaluation andupward contacts. Psychological Review, 96, 569–575.

Turner, S. W., Thompson, J., & Rosser, R. M. (1995).The Kings Cross fire: Psychological reactions.Journal of Traumatic Stress, 8, 419–427.

Turpin, G., Downs, M., & Mason, S. (2005). Effec-tiveness of providing self-help information follow-ing acute traumatic injury: Randomised controlledtrial. British Journal of Psychiatry, 187, 76–82.

van Emmerik, A. A. P., Kamphuis, J. H., Hulsbosch,A. M., & Emmelkamp, P. M. G. (2002). Single


session debriefing after psychological trauma: Ameta-analysis. The Lancet, 360, 766–771.

Viney, L. L., Clarke, A. M., Bunn, T. A., & Ben-jamin, Y. N. (1985). An evaluation of three crisisintervention programmes for general hospital pa-tients. British Journal of Medical Psychology, 58,75–86.

Wee, D. F., Mills, D. M., & Koehler, G. (1993).Stress response of emergency medical servicespersonnel following the Los Angeles civil distur-bance. Paper presented at the Third World Con-gress on Stress, Trauma and Coping in the Emer-gency Services Professions, Baltimore, MD.

Wee, D. F., Mills, D. M., & Koehler, G. (1999). Theeffects of Critical Incident Stress Debriefing(CISD) on emergency medical services personnelfollowing the Los Angeles civil disturbance. Inter-national Journal of Emergency Mental Health, 1,33–37.

Weiss, D., & Marmar, C. (1997). The Impact ofEvent Scale -Revised. In J. Wilson & T. Keane(Eds.), Assessing psychological trauma and PTSD(pp. 399–411). New York: Guilford Press.

Wiedemann, G., Pauli, P., & Dengler, W. (2001). Apriori expectancy bias in patients with panic dis-order. Journal of Anxiety Disorders, 15, 401–412.

Woodall, S. J. (1994). Personal, organizational, andagency development: The psychological dimen-sion. A closer examination of critical incidentstress management. (Available from the NationalFire Academy Executive Fire Officer Program,Learning Resource Center, National EmergencyTraining Center, Emmitsburg, MD)

Yates, S., Axom, D., Bickman, L., & Howe, G.(1989). Factors influencing help seeking for men-tal health problems after disasters. In R. Gist & B.Lubin (Eds.), Psychosocial aspects of disaster (pp.163–189). New York: Wiley and Sons.

Yates, S., Axom, D., & Tiedeman, K. (1999). Thehelp-seeking process for distress after disasters. InR. Gist & B. Lubin (Eds.), Response to disaster:Psychosocial, community, and ecological ap-proaches (pp. 133–165). Philadelphia: Brunner/Mazel.

Zlotnick, C., Warshaw, M., Shea, M. T., & Keller,M. B. (1997). Trauma and chronic depressionamong patients with anxiety disorders. Journal ofConsulting and Clinical Psychology, 65, 333–336.

Received August 1, 2005Revision received December 6, 2005

Accepted February 6, 2006 �

Acknowledgment of Ad Hoc Reviewers

The Editor and Editorial Board wish to thank the following persons who served asreviewers during the preparation of volume 10.

Woo-kyoung AhnIrving AlexanderJamie ArndtIlan AviramRainer BanseNicole BaurenbaumGordon BearMarijana BeneshRichard CarlsonStephen ColarelliKathryn DindiaMichael DougherDana DunnAlan ElmsGordon GallupJamie GoldenbergNick Haslam

William HirstRobert HoganHarry HuntBlair JohnsonCarl N. JohnsonDerek JohnstonArvid KappasAlan KazdinMark Koltko-RiveraJustin KrugerBrian LittleTania LombrozoDario MaestripiereDouglas MedinMichele MiozzoDonald OwingsKathleen Page

Jennifer PalsSally PanalpDel PauhusAaron PincusJ. T. PtacekJames ReamDaniel ReisbergMichele RobinsonArolda RodriquesDavid RubinCatherine SalmonPaul SilviaJanet StoppardAnn WeberBernard WeinerCarsten WroschLisa Zunshine


Ready! Fire! Aim! The Status of Psychological Debriefing and ...

Documents

Transcript of Ready! Fire! Aim! The Status of Psychological Debriefing and ...