Visual communication desensitization (VCD©): a novel two-phased approach to interviewing traumatized individuals in investigative contexts

Abstract

Investigators are increasingly faced with the challenge of obtaining evidence from witnesses who are distressed or traumatized by war, genocide, human trafficking and/or sexual violence. However, interviewing best practice does not explicitly address interviewees’ psychological needs, and therapeutic approaches do not incorporate techniques to obtain accurate, detailed information. This article charts the development of the visual communication desensitization (VCD©) interview procedure, designed to secure accurate accounts rapidly while reducing distress. Originally developed to aid a traumatized victim of violent sexual assault who was unable to provide evidence in court, it comprises a two-part cognitive-behavioural approach to eliciting information from cooperative, traumatized witnesses, consisting of a ‘narrative-graph’ information-gathering component and a dovetailed therapeutic component. The laboratory findings and applied practice observations suggest that the VCD© interview procedure is beneficial for assisting traumatized interviewees in providing accounts, and thus could potentially support capacity-building in humanitarian response and international investigation contexts.

Introduction

Due to widespread geopolitical instability, the number of men, women and children experiencing distress and trauma as a consequence of war, genocide, human trafficking and/or sexual violence is increasing. At the end of 2014, the number of people forcibly displaced by conflict and human rights violations stood at an all-time high of 59.2 million, many of whom remain interned in refugee camps near their conflict zones. More recently, many others have taken significant risks to seek asylum and refuge in Europe (UNHCR, 2015). Displaced people from Middle Eastern and Central African conflict regions are also likely to have endured the additional endemic, but less visible, problem of gender-based violence (United Nations Security Council, 2015; World Health Organization, 2013). Further, in 2012 it was estimated that approximately 4.5 million people – primarily women and children – had been forced into lives of sexual exploitation (International Labor Organization, 2012), and it is likely that sex trafficking numbers have increased markedly with the more recent influx of refugees from the Middle East into Europe (International Organization for Migration, 2015). Domestic crime statistics also suggest that there has been an increase in violent crime, particularly against women, since 2009 (Walby, Towers, & Francis, 2016). Such humanitarian and gender-based abuses have a significant effect on the psychological well-being of victims, inflicting a long-term impact on rehabilitation and recovery at both the individual and the societal level.

In critical response contexts, it is often the case that specialist resources to respond to the psychological needs of victims and obtain witness accounts take second place, unsurprisingly, to the more urgent need to preserve life and ensure basic safety (World Health Organization, 2007). However, receiving authorities and other agencies are faced with two equally pressing concerns in terms of responding to victims: (a) the need to conduct investigative evidence gathering (what has happened and who is responsible?) and (b) the need to provide humanitarian–therapeutic support (how can we help the individual?). Both concerns rely on gathering accurate information from people's memories of traumatic events, an undertaking that at present is not necessarily well informed by good interviewing practices, adequate techniques for managing trauma, or the application of standardized approaches. These deficits may lead to incomplete or unreliable accounts while also failing to address the psychological needs of traumatized individuals. Worse still, poor interviewing practices may increase and prolong the trauma of interviewees while also causing distress to any intermediaries involved, as well as to the interviewers themselves (e.g. Sandick, 2012). Finally, inadequate contemporaneous accounts (i.e. incomplete, erroneous, contaminated) obtained using poor interviewing methods are likely to impede the future delivery of justice.

Unfortunately, investigative interviewing approaches developed in civilian policing contexts may fail to meet the requirements of many victims, witnesses, investigators and non-governmental organizations (NGOs) in broader response contexts. For instance, gold-standard police interviewing procedures, such as the cognitive interview (Fisher & Geiselman, 1992), are not designed to address the psychological needs of traumatized interviewees who may be experiencing high levels of distress. Therefore, a major challenge for investigators (e.g. the police, the International Criminal Court) and other groups (e.g. asylum assessors, front-line or receiving NGOs) working to build evidence for reporting or tribunal purposes lies in eliciting detailed and reliable accounts from traumatized individuals (Sandick, 2012). Successfully obtaining such accounts can be especially difficult when many victims have witnessed or participated in conflicts that have occurred under rapidly shifting circumstances, an issue which is often further compounded by potential informants having informally repeated their accounts and shared their experiences with an array of individuals and agencies (Herlihy, Scragg, & Turner, 2002). Indeed, Bergsmo (2011) highlights the challenges faced in international investigation cases, noting in particular the difficulties of ‘old evidence’ and ‘contaminated’ witness testimony. Too often, accounts of incidents and abuses become distorted and contaminated, as a function of delay, poor interviewing techniques and a lack of understanding of how to ethically elicit accounts from reticent, confused and/or traumatized victims (Hope, 2013). Additionally, charges of sexual violence perpetrated again vulnerable populations has been limited, in part because of reticence, disregard and embarrassment when it comes to investigating matters pertaining to sexual activity (Aranburu, 2010). These observations resonate with calls for improved interview procedures and training for those conducting such investigations (O'Brien & Kebbell, 2014; Sandick, 2012).

In these contexts, victim and witness memories can be affected by distress and trauma (Hamilton, 2014). For example, memories of traumatic events can be fragmented or compounded so that sequence and perception of time become distorted. Symptoms of post-traumatic stress disorder (PTSD) can also include amnesia, difficulty recalling important aspects of an event and involuntary distressful memories or ‘reliving’ events (Brewin, 2007, 2011). These and other factors can contribute to the difficulty of gathering coherent, time-lined information from traumatized persons that is suitable to use for legal purposes. Assessments and interventions designed to assist traumatized persons in coming to terms with their experiences (e.g. narrative exposure therapy, cognitive processing therapy) focus, quite appropriately, on therapeutic benefits rather than investigative and evidentiary concerns (Bass et al., 2011). Difficulties can occur, however, when investigative and clinical interviewing goals conflict. For example, therapeutic interviewing processes may risk distorting a victim or complainant's memory of events, by supporting or reinforcing statements as though they are fact and by utilizing procedures such as visualization or other suggestive techniques that risk creating false memories (Goff & Roediger, 1998; Laney & Loftus, 2008; Strange, Garry, & Sutherland, 2003), thereby reducing the prospect of obtaining accurate information.

While there is a sizeable literature on investigative interviewing in domestic policing contexts (Vrij, Hope, & Fisher, 2014) and an emerging literature on front-line psychological interventions in challenging contexts (e.g. Bass et al., 2011), there does not appear to be an interview method that is explicitly devised to serve a therapeutic function for the interviewee in order to facilitate the communication of memories of traumatic experiences (through visual as well as verbal procedures) in a way that also attends to the interviewer's need to obtain a reliable account of events.

This article presents a case study and a laboratory-based test of an innovative interview procedure that is suitable, in the first instance, for gathering information from traumatized victims and witnesses. This procedure has been developed to overcome the tension that exists between eliciting information for investigative purposes (in which the focus is necessarily on the accuracy and quantity of the information elicited) and gathering information in the therapeutic context (in which reports are accepted as reflecting personal experiences and the focus is more on managing the effects of trauma).

The visual communication desensitization (VCD©) interview procedure

The visual communication desensitization (VCD©) interview procedure comprises a two-part cognitive-behavioural approach to eliciting information from a cooperative but traumatized victim or witness. Part 1 of the procedure comprises a ‘narrative graph’ information-gathering component and Part 2 comprises a dove-tailed therapeutic component. The development of both parts of the VCD© interview procedure has been informed by the memory literature, the PTSD literature and clinical experience in eliciting evidential accounts from traumatized individuals. The development of the procedure also took into account that intrusive thoughts and images associated with traumatic memories are often visual, and yet assessments of an individual's traumatic state typically rely on verbal procedures such as clinical interviews and questionnaires (e.g., Foa, Ehlers, Clark, Tolin, & Orsillo, 1999). Further, prompting the articulation of traumatic memories may pose a communication hurdle for those individuals who for cultural and/or religious reasons feel that their experiences are unspeakable (Dudgeon, 2012), who are communicating in a second language and/or who are afraid of the consequences of disclosing abuse (International Rehabilitation Council for Torture Victims, 2009). As victims of trauma can present with symptoms that interfere with communication, the key initial aim of the VCD© interview procedure is to reduce distress levels in order to encourage traumatized interviewees to provide clearer and more complete information. Consistent with contemporary approaches to the attenuation of PTSD symptoms via visuospatial tasks (Holmes, James, Kilford, & Deeprose, 2010) and emotion-regulation training (Bryant et al., 2013), the VCD© interview procedure aims to assist traumatized individuals by providing self-directed visual feedback through the use of a ‘narrative graph’. This procedure is intended to (a) enable interviewees to indicate and control the levels of distress experienced; (b) prompt further recall; and (c) facilitate the production of a time-lined account.

The first section of this article presents a case study describing how the VCD© interview procedure was originally devised as an innovative and direct response to levels of distress that were grossly interfering with the ability of a traumatized woman (referred to here using the pseudonym ‘Kay’) to recount events for a criminal court, including difficulties recalling aspects of the events and the order in which they transpired. Hence, the key initial aim of the VCD© interview procedure was to reduce Kay's distress levels in order to help her provide clearer and more complete information without contaminating her report. The second section reports the findings of an empirical study designed to determine whether or not using the narrative-graphing technique to elicit an account has any systematic or detrimental effects on the accuracy of reporting by mock witness participants.

The case study: ‘Kay’

The interviews reported in this case study were conducted by the first author, who is a licensed New Zealand psychologist with clinical experience of working with traumatized adults and children involved in legal systems, including NGO services for institutionalized children.

Kay was a woman in her twenties who first came to a clinic for help with memory problems. During initial attempts to interview her, she exhibited observable signs of psychological trauma, including shaking, gazing off and ceasing to speak. Her records showed that she had recently been discharged from hospital following head and spinal injuries. Prior to this hospital admission she had been interviewed by the New Zealand Police who, investigating a crime which included abduction and rape, wanted her to testify as a complainant. The investigators requested that Kay receive clinical assistance to achieve this but, understandably, did not want her account to be contaminated by any therapeutic process, so it was decided that her initial statement to them should not be disclosed to the therapist-interviewer.

In the initial session with the therapist-interviewer, Kay was given a graph-like form, the vertical axis of which was labelled ‘client sense of distress’ and scaled from 0 to 100, and the horizontal axis of which was open-ended and labelled ‘client directed timeline’. This ‘narrative graph’ constitutes Part 1 of the VCD© procedure. Kay was asked to think back to the traumatic events that occurred and use a pencil to draw a line on the narrative graph which fluctuated (i.e. went up and down) according to her degree of distress whilst recalling these events. This line had to begin at a less distressful point just before the remembered distressful events and end at a less distressful point just after them. Kay was free to decide the beginning and end points of each remembered traumatic event. The therapist-interviewer told Kay that she could choose to verbalize her remembered events as she progressed, or to remain silent if she preferred. She could stop at any time for as long as she wanted but was encouraged to get to the ‘end’, as she saw it, of her remembered traumatic events. If she stopped, she had to try to keep her pencil in the same place and choose either to verbally describe her visualized events pertaining to that point in time (referred to herein as a ‘pause point’) or to put a short vertical line or ‘hash mark’ or word(s) on her line at that pause point as a reminder before carrying on. She could put as many hash marks or words as she liked on her line and resume drawing it within a given session as many times as she wanted, either resuming from the last pause point or restarting from the beginning of the timeline. Initially, if Kay was silent or inarticulate, the therapist-interviewer marked the graph with a hash point when Kay's pencil paused and, using only open questions, asked Kay what she was recalling. Kay's words or phrases were then either written next to those hash marks or listed on a separate sheet of paper. It is important to note that, throughout the interview, the therapist-interviewer worked with this visual representation to access further details solely through the use of open questions and non-leading prompts.

Each session ended when Kay or the interviewer determined. Sessions were repeated using a new (blank) narrative graph in order to allow Kay to try to reach the end of her remembered series of traumatic events and/or become accustomed to communicating the timeline of these events. All sessions had the same starting point. As the sessions progressed, pause points were treated as ‘anchors’ and more information was requested in terms of what happened just before and just after these points. Figure 1 illustrates Kay's narrative graphs over the six sessions in which she charted her degree of distress and narrative of the traumatic events.

Figure 1.

Completion of the Part 1 narrative graph over six VCD© interview procedure sessions (case study).

As can be seen from Figure 1, Kay did not reach the ‘end point’ of her account until the fourth session. In addition, her distress levels followed a similar path across all sessions, although the amount and timing of the information reported varied across sessions. Kay's report was initially comprised of single words or phrases that were written down by the interviewer, but in the last two sessions progressed to sentences that Kay wrote down herself. Her account was consistent across the sessions. During Sessions 4, 5 and 6 there were gaps in her recall that were labelled ‘P’ for ‘Passed out’. Part 1 of the VCD© interview procedure (the narrative graph) enabled Kay to provide information at her own pace while giving her visual anchors to prompt further recall and placing the remembered events immediately in sequence within a time frame.

This information – captured on the narrative graphs, along with the written notes of what she reported – then provided the basis for the therapeutic process used in Part 2 of the VCD© interview procedure. This therapeutic process started with dividing the levels of distress on her final graph into equal bands from 0 to 100, with the least distressful events falling within to 0 to 10 range and the most distressful events falling within the 90 to 100 range (Figure 2).

Figure 2.

Example of the desensitization process in Part 2 of the VCD© interview procedure (case study).

Kay was introduced to cognitive procedures typical of systematic desensitization as developed by Wolpe (1961), initially comprising a graduated physical relaxation procedure that incorporated visualizing a calming scene of her choice in order to combat tension, along with some elements of rational self-talk taken from cognitive behavioural therapy (CBT). She was then asked to relax, visualize the events from her narrative that fell within the lowest distress range (0–10) and replace these with both the visualized calming scene and self-talk. This continued until she could visualize the target narrative within this distress range and remain calm. This process was repeated across several sessions in a stepwise fashion up the hierarchy of distress ranges, using events from across Kay's trauma timeline within each step. Again, the therapist-interviewer used only open questions to elicit information, and Kay proceeded up the hierarchy at her own pace.

Discussion of the case study

Through the use of Parts 1 and 2 of the VCD© interview procedure, Kay successfully met her goal to speak about and clearly convey events – wherever they fell across the timeline – within each of the distress ranges without dissociating or becoming unable to verbalize her experience. Her report during sessions was subsequently found to be consistent with her original statement to the police. She was able to provide relevant details when asked first within therapy sessions and then at trial, leading to the successful prosecution of her assailants.

Although the use of the VCD© interview procedure in clinical practice has so far proved successful in a number of settings at enabling traumatized individuals to reduce their distress levels and provide a coherent account of their experiences (see the General Discussion section below), the procedure had not been subjected to empirical testing in order to determine whether or not any aspect of the narrative graphing procedure systematically impairs, impedes or otherwise contaminates recall. Interviewing best practice typically emphasizes the importance of using open questions and fostering unimpeded free recall (Vrij et al., 2014). Given that the narrative graph adds an additional output element to assist the interviewee in representing, and subsequently attenuating, distress levels, it was deemed important to determine whether or not this additional component has any detrimental effect on either the accuracy or the quantity of the reported information. For sound ethical reasons it is not generally permissible to generate high levels of distress in experimental participants; therefore, the accuracy and quantity of information reported under laboratory conditions was assessed. Using a standard mock witness paradigm, the information provided by mock witnesses who gave their account using the narrative graph (VCD© Part 1) was compared with those who gave their account in a verbal free-recall format.

Method

Design and participants

A total of 50 adults (29 females, 21 males) were recruited, using contacts and social media, in Tblisi, Georgia, by research assistants trained by the first author in using the VCD© interview procedure. The inclusion criteria consisted of being under 55 years of age and wearing corrective lenses as necessary. The average age of the participants was 28.85 years (range: 20–51). After witnessing the stimulus event, participants were randomly allocated to one of two experimental conditions: the narrative-graph condition (n = 28) or the free-recall condition (n = 22).

Materials

The stimulus event was a short film segment (2 min) depicting a professionally filmed, staged scenario in which a male perpetrator breaks into a house and begins to verbally threaten the three females in the house (one of whom is apparently his ex-partner). The women escape from the house and manage to impede and evade the perpetrator. Although no physical violence is depicted, pilot evaluations of the stimulus event indicated that the event is engaging and reflects a mildly distressing scenario due to the implied threat to the female characters. Further, the pilot participants were able to plot a representation of their recall of the scenario that reflected their perception of the implied threat (i.e. increased distress ratings when the threat was greatest). The participants were warned in advance about the content of the stimulus event via appropriate informed consent procedures.

Procedure

The participants were instructed to imagine themselves as witnesses to the incident shown and then watched the stimulus event individually on a laptop computer screen. After a short delay, the participants were reminded of their witness role and asked to recall as much as they could about the event they had just seen, from beginning to end. In both conditions, the participants received the same overall recall instructions requiring them to report as much information as they could about the witnessed incident, including detailed descriptions of people, actions, objects and settings. No time restrictions were imposed on reporting in either condition.

The participants in both conditions were given around 1 minute to absorb the overall recall instructions before receiving the instructions specific to the condition to which they had been allocated.

The participants in the narrative-graph condition were given a VCD© narrative-graph form – a blank simple graph with its x-axis labelled ‘client directed timeline’ and its y-axis labelled ‘client's level of distress’ and marked from 0 to 100 in intervals of 10 (as per the graphs depicted in Figure 1). They were instructed to use the horizontal axis of the graph to help them sequence their memories of the witnessed event and use the vertical axis to reflect how they felt as they remembered the event, with marks higher on the vertical axis reflecting feelings of greater distress. They were also instructed that they could add words or other marks onto the graph as they completed it if they felt it would help them remember details. After drawing their narrative graph, participants were then asked to asked to provide a verbal account of what they had seen (keeping their narrative graph in view). On completion of the verbal account, the participants were then asked to refocus on the narrative graph and invited to provide any further details that they could remember.

In the free-recall condition, the participants were simply asked to provide a verbal account of what they had seen. On completion of the verbal account, they were then asked to refocus and invited to provide any further details that they could remember.

As the VCD© interview procedure is designed to be used over multiple sessions if necessary, all participants were interviewed on three occasions in order to compare their performance in repeated interviews between the two conditions. The first interview took place shortly after viewing the stimulus, the second interview took place after a delay of one week, and the third interview took place a further week later. The interview instructions remained the same across all three interviews and the participants remained in the same interviewing condition throughout the process.

Coding

Recall data were coded for accuracy using a coding template consistent with Gabbert, Hope, and Fisher (2009). An item was deemed correct if it was present in the event and described correctly, and deemed incorrect if it was present in the event but described incorrectly or was not present in the event. Sequencing errors were coded by assessing the order in which action information was reported. A sequencing error was recorded where action information was not reported in the order in which the actions in the stimulus event occurred. Coding strictly adhered to the coding template and the researchers discussed and resolved the coding of any ambiguous details. Subjective responses (such as ‘he was ugly’) were not coded.

Results and discussion

The main focus of interest is on the performance in the first interview, which took place shortly after the event was encoded. Specifically, the aim is to assess whether or not eliciting information using the narrative-graph procedure has any systematic or detrimental effect on the accuracy of the participants’ reporting.

The accuracy rate was calculated by dividing the total number of correct details reported by the number of total details reported (both correct and incorrect). Overall, the accuracy rate for the reported information is high and, although it is slightly higher in the free-recall condition (M = 98%) than the narrative-graph condition (M = 96%), this difference is not significant, t(48) = 1.81, p = .08, d = .63. For the first interview there is no significant difference in the number of correct details reported between the free-recall (M = 55.68) and narrative-graph (M = 45.75) conditions, t (48) = 1.95, p = .06, d = .57. Similarly, there is no significant difference in the number of incorrect details reported between the free-recall (M = 1.36) and narrative-graph (M = 2.00) conditions, t(48) = −1.76, p = .25, d = .34. There is also no significant difference in the number of sequencing errors between the free-recall (M = 0.14) and narrative-graph (M = 0.18) conditions, t(48) = −0.39, p = .69, d = .11. This pattern of results is replicated across the second and third interviews, with no significant differences emerging between the experimental groups for any of the dependent measures (Table 1).

Table 1.

Descriptives and analyses for Interviews 2 and 3.

		Free-recall condition		Narrative-graph condition
		M	SD	M	SD	t	d
Interview 2	Accuracy rate	0.97	0.03	0.96	0.04	0.96	.28
	Correct details	53.09	11.67	43.82	20.69	1.88	.55
	Incorrect details	1.73	2.02	1.93	2.01	−0.35	.10
	Sequencing errors	0.46	0.67	0.68	1.06	−0.39	.25
Interview 3	Accuracy rate	0.97	0.03	0.96	0.04	1.12	.28
	Correct details	52.32	11.44	43.82	21.99	1.64	.48
	Incorrect details	1.73	1.86	2.11	2.38	−0.62	.12
	Sequencing errors	0.33	0.58	0.29	0.46	0.32	.08

These results suggest that the narrative graph – Part 1 of the VCD© interview procedure – does not significantly impair the quality or quantity of the information provided about a witnessed event (relative to free recall) across repeated interviews. Although the means indicate that fewer correct details were reported in the narrative-graph procedure across the sessions, the differences are not statistically significant. It is also worth bearing in mind that the goal of this method is not necessarily to elicit more information than existing methods but rather to interview traumatized individuals in a way that addresses both investigative and therapeutic concerns.

General discussion

The VCD© interview procedure is based on solid principles underpinning both interviewing best practice and the current understanding of the effects of trauma. The VCD© interview format was originally developed to reduce and manage the trauma of a potential complainant so that she could communicate her memories of traumatic events, uncontaminated by the interview process, both coherently and consistently during cross-examination. It has since been developed in the course of consultation with investigative interviewing practitioners and clinicians internationally.

The VCD© narrative-graph interview procedure incorporates sound investigative interviewing practices, as the interviewee provides her or his account with minimal prompts from the interviewer. Instead of a question-and-answer format, it incorporates a timeline element that should serve to enhance recall of both actions and the sequence of events (Hope, Mullis, & Gabbert, 2013). Carefully conducted interviews that facilitate repeated opportunities to recall events can elicit highly accurate details (Fisher, Brewer, & Mitchell, 2009; La Rooy & Lamb, 2008), and the VCD© interview procedure aids interviewees in providing further details without the risk of interviewer contamination. Systematic testing of the VCD© reporting format across repeated recall sessions suggests that the procedure itself is as effective as open-ended free recall and does not significantly impair interviewees’ memories and reporting of witnessed events.

In addition to facilitating the provision of an account, the VCD© interview procedure attempts to quantify levels of distress and monitor changes in emotional response. The framing of remembered events within a visual timeline, in combination with the repeated nature of Part 1 either within a session or across sessions, acts as a means of externalizing traumatic memories and has initial therapeutic benefits for the interviewee. By incorporating distress-reducing elements, the VCD© interview procedure targets a key challenge faced in many international investigative and humanitarian response settings. For example, in a report for the United Nations High Commissioner for Refugees (UNHCR), Gorlick (2002) highlights this challenge, noting that the ‘memory of a person under stress or fearful of expressing themselves to a person in authority can play a crucial role in an applicant's inability to provide testimony which is consistent and coherent’ (p. 14).

A potential benefit of the VCD© interview procedure relates to the vulnerabilities of certain victims and witnesses in the context of their experiences and the timing of their disclosure of sensitive information. The findings of a recent review suggest that while an open-question approach in verbal interviews may facilitate accurate reports for certain vulnerable samples, it does not necessarily increase the amount of information reported (Bowles & Sharman, 2014). For example, Lindholm, Cederborg, and Alm (2014) observed that some adolescent girls who have been exploited in the sex trade and subjected to violence will ‘withhold information’ if interviewed with open questions ‘soon after the exploitation has ceased’ (p. 12). The VCD© interview procedure enables the interviewee to initially communicate her or his distress levels across a timeline and signal different levels of vulnerability, which in turn allows the interviewer to tailor the interview and ask open questions – beginning with low-distress recollections that might be easier for the interviewee to disclose, thus providing a pivotal entry point to subsequently obtaining relevant but embarrassing, culturally sensitive and/or security-related information. Of course, further research is needed to establish the situations in which it might be most beneficial – for both interviewee and interviewer – to conduct the VCD© interview procedure or more traditional open-ended interviews of this nature. Finally, the narrative-graph format of the VCD© interview procedure results in a permanent record that allows rapid visual analyses of repeated accounts and levels of distress that are useful both to the investigator who may wish to examine or track contemporaneously dated accounts over time and to the therapist who may wish to examine efficacy of support.

Applications of the VCD© interview procedure

The VCD© interview procedure has been found in clinical practice to assist a small number of traumatized individuals across a variety of different contexts (e.g. individuals who have experienced childhood sexual abuse; internally displaced persons who have experienced conflict-related trauma; indigenous remand prisoners who have reported being traumatized by their offending). Combined client and clinician feedback suggests that the VCD© interview procedure provides a means for self-directed visual feedback that (a) provides an indication of the degree of distress, (b) helps the client control feelings of distress, (c) prompts further recall and (d) enables the production of a more coherent/time-lined account. It may be that for some people the opportunity to communicate at their own pace by means of a visual narrative that incorporates self-prompts can facilitate the provision of useful information and evidence without recourse to lengthy verbal interviews. For example, the narrative graph (VCD© Part 1) has the potential to be privately and relatively rapidly self-administered by non-complainant rape victims, then stored along with the semen samples taken at the time and made available for subsequent comparison with other victims’ reports and semen samples in order to gather cumulative evidence in suspected serial rape cases. Similar usage may be possible for the reporting of repeated abuse or violence.

Training workshops on the VCD© interview procedure have been conducted with small groups of practitioners in the United States (US) and in Georgia, comprising psychologists, psychiatrists, community leaders and translators with varying degrees of experience of working with traumatized clients with histories of abduction, torture and/or sexual assault. In these workshops, training on the use of the VCD© interview procedure was administered in two ways. First, a one-day ‘first responder’ workshop was provided, in which the attendees were trained (including through the use of practice interviews) how to administer the narrative graph. The attendees then used the graph in vivo as an initial interview only, and learned how to make a professional referral to a therapist. Evaluation of this first responder training workshop suggests that the VCD© can be implemented by first responders with minimal training to (a) gather information about traumatic event(s) with no contamination of the original account, (b) evaluate any communication difficulties present during the assessment, (c) illustrate the impact of trauma on the interviewee via the graphed levels of distress and (d) reduce translator time and translation errors. Second, in the course of a two-day workshop for therapeutic professionals, attendees learned how to dovetail the narrative-graph method with the systematic desensitization component of Part 2 of the VCD© interview procedure, and how to apply and monitor the effects of three therapeutic sessions with their clients. The subsequent feedback on clinical utility from the attending para-therapists (n = 16) and clinicians (n = 12) suggests that the VCD© interview procedure is useful for professionals working with highly traumatized clients as both an initial interview procedure and a form of therapy. Furthermore, the feedback received from translators working with medical and trauma specialists indicates that the non-verbal elements of the VCD© interview procedure increase interviewee privacy, as well as being more sensitive to interviewee cultural norms and intermediary distress.

Conclusion

The VCD© interview procedure has the potential to assist four end users: the traumatized individual, the first responder, the (para-)therapist and the legal investigator. First, it assists persons who are unable to articulate intensely personal, traumatizing events by reducing their level of distress and helping them to communicate to another person what has happened. Second, the narrative graph (VCD© Part 1) may be suitable for first responders in large-scale humanitarian response contexts who need to quickly elicit coherent accounts from traumatized individuals – including victims of gender-based violence – while assessing their need for referral to specialist services. Third, Part 2 of the VCD© interview procedure facilitates, via the graphed distress levels, a systematic desensitization process which further alleviates the impact of trauma and can be used as the basis for longer-term therapy or as an adjunct to CBT. Finally, when implemented correctly, the narrative graph (VCD© Part 1) provides a permanent record that is potentially useful to legal investigators, both as evidence gathered close to the time of the traumatic event and as a measure of reporting consistency when compared with subsequent evidence obtained in the same way.

Usefully, the stepwise and specific nature of the VCD© interview procedure – wherein Part 1 dovetails into Part 2 – lends itself to the provision of short-course training for local professionals, ranging from officials charged with conducting criminal investigations through to those concerned with the delivery of critical social services, thus facilitating the rapid evaluation and referral of individuals to relevant professionals for the purposes of both aiding investigations and providing psychological assistance.

In light of its use to date and the early-stage testing reported herein, it appears that the VCD© interview procedure has the potential to address immediate therapeutic and referral challenges, as well as longer-term capacity-building and international-investigation goals. Of course, one important limitation of this initial evaluation is that it is not possible to systematically assess the distress-reducing component of the VCD© interview procedure in a laboratory test. Future work should comprise a larger field trial or, ideally, a randomized controlled trial with diverse traumatized populations in order to document the clinical performance of the narrative graph in combination with its therapeutic component (e.g. following Weiss et al., 2015).

Further development is also needed to establish the most effective training manuals, referral processes and programmes for first responders and therapeutic responders working with large samples of traumatized individuals, including children and adolescents.

Acknowledgments

The authors would like to extend grateful thanks to the researchers for this project, Tamta Saamishvili, Tamar Mumladze and Nino Mikaia.

Declaration of conflicts of interest

Jane Mary Castelfranc-Allen has declared no conflicts of interest.

Lorraine Hope has declared no conflicts of interest.

Ethical approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional research committee of Tbilisi State University, Georgia and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Informed consent

Informed consent was obtained from all individual participants included in the study.