The Efficacy of Imagery Rescripting on PTSD, Depression and Anxiety
Disorders: A Meta-Analysis
Master Thesis
Faculty of Social and Behavioural Sciences
Department of Psychology - Programme group Clinical Psychology Reem Dannaoui
Supervisor: Dr. N. Morina October 2 , 2015
1
Abstract
This meta-analysis evaluates the efficacy of imagery rescripting (IR) on Posttraumatic Stress Disorder (PTSD), depression and anxiety disorders. Relevant clinical trials were collected from the databases PsychInfo, Web of Science and Medline if at least five patients were treated with IR and if at least 50% of the applied treatment consisted of IR. Twelve studies met inclusion criteria and were analyzed using the Comprehensive Meta-Analysis software package. Three of the 12 included studies were randomized controlled trials (RCTs). IR appeared to be effective in the treatment of PTSD, depression and anxiety
disorders. The benefits of IR were sustained for an average of 14 weeksfollowing treatment completion. The analysis including three RCTs revealed that IR outperformed active control conditions (Hedges’ g = 0.72), however these results must be seen as preliminary
considering the limited number of included trials. Notably, the effect of IR was already visible after one session. The found results indicate that IR may be effective in the treatment of PTSD, depression and anxiety disorders. Future research should focus on conducting research with children, since none of the included nor screened research has done so. Finally, it is of importance that future research looks into the effectiveness of IR compared to alternative treatments in RCTs.
2
Contents
1. Introduction………3 2. Method………..6 2.1 Search………...6 2.2 Inclusion criteria………6 2.3 Coding………7 2.4 Analysis……….7 3. Results………8 3.1 Search results ………8 3.2 Sample characteristics ………. 9 3.3 Effect sizes………. 10 4. Discussion……….14 Acknowledgements………..17 References………18 Appendix………213
1. Introduction
Mental images are part of our mental experience. When mental images occur
spontaneously and automatically, they are perceived as intrusive images (Rusch et al., 2000). Intrusive images may be experienced as repetitive, uncontrollable and distressing (Brewin et al., 2010). Research indicates that imagery may elicit stronger emotional responses than do comparable verbal cognitions (Holmes et al., 2008). Furthermore, distressing intrusive images are often reported as a symptom for many psychological disorders, among them Posttraumatic Stress Disorder (PTSD), other anxiety disorders, and depression (Rusch et al., 2000).
Already since the inception of cognitive therapy, the importance of mental imagery was emphasized in understanding emotional distress (Holmes et al., 2007). According to
Edwards (2007) imagery, including forms of rescripting has been used as a therapeutic technique for over 20.000 years. However, the treatment of imagery has not been emphasized in scientific research until recently.
One of the available treatments that target imagery is imagery rescripting (IR). IR is a therapeutic technique that is used to change the meaning of emotional memories and images (Arntz, 2012). The individual is asked to imagine the memory or image of an aversive event as vividly as possible and is then asked to change the course of the events in a more desirable direction. This allows the individual to rescript the traumatic or negative
experience, imagine having control over the situation and to act according to his needs. The patient is also given the opportunity to discover and express inhibited emotions related to the event (Arntz et al., 2007). According to Arntz et al. (2007) IR tends to be a less stressful treatment for the therapist, compared to exposure without rescripting. The therapists
4 reported feeling less emotionally distressed and experienced less helplessness during the IR treatment.
It has been suggested that IR causes change in the fear memory (Arntz, 2012). According to former research if a fear memory is reactivated, it gets in a labile state in which it can be modified or even erased (Forcato et al., 2007). IR could be a way of modifying the fear memory in order for it to be reconsolidated with a different meaning, so it no longer elicits the same emotional responses (Anrtz, 2012). The study of Hagenaars and Arntz (2011) concluded that IR does not lead to forgetting the aversive event; it may even lead to a better knowledge of the event. These results support the theory that IR does not replace or erase the memory of the event, but rather works through changing its meaning (Arntz, 2012). There are many unsettled controversies about the actual mechanism of IR. Future research needs to investigate the underlying mechanism of imagery rescripting, contributing to the most effective application of this treatment.
Research on IR has grown the past decade. IR was initially developed to treat patients suffering from PTSD and intrusive images as a result of childhood sexual abuse (Whitaker et al., 2010). IR has been applied in treating patients suffering from PTSD (Arntz et al., 2007), depression (Wheatley et al., 2007) and anxiety disorders (Wild et al., 2008). Former research previously emphasized the positive effect IR has on treating disorders with symptoms of intrusive images (Arntz et al., 2007; Wheatley et al., 2007; Wild et al., 2008).
While several trials and reviews have been conducted, a meta-analysis on the results that would offer a quantitative review of the existing publications is missing. By combining existing publications and its results statistically by means of a meta-analysis, a conclusion can be reached about the efficacy of imagery rescripting on PTSD, depression and anxiety disorders.
5 The intent of this meta-analysis is to evaluate the efficacy of imagery rescripting on PTSD, depression and anxiety disorders. This will determine: 1) the IR treatment effect from pre- to post-treatment, 2) the long-term treatment outcome of IR, and 3) the comparison of IR to control conditions. On average IR lasts between one and 15 sessions. The conducted
analysis will inform us about the importance of the number of sessions. Finally, the effect of IR on secondary outcome measures will be determined.
6
2. Method
2.1 Search
Studies were identified using four search strategies. First, the bibliographic databases PsychInfo, Web of Science and Medline were searched using the following command: imagery rescripting OR ImRs OR IRT OR updating memory OR imagery modification OR cognitive restructuring OR imaginal reliving. Second, reference lists from articles that met inclusion criteria for the meta-analysis were examined. Third, missing data from articles were retrieved by contacting the authors. Fourth, registers of controlled trials were searched. The following registers were searched: Australian New Zealand Clinical Trials Registry, Chinese Clinical Trial Register (ChiCTR), Clinical Trials (service of NIH), Clinical Trials Registry- India (CTRI), German Clinical Trials Register (DRKS), ISRCTN Register, Nederlands Trial Register (NTR) and UMIN Clinical Trials Registry. For a final search aiming at covering all existing publications the databases Research Gate and Google Scholar were searched.
2.2 Inclusion criteria
Studies were included in the meta-analysis if they applied IR as a means of reducing post-traumatic symptoms, or symptoms of depression or anxiety. Studies were excluded from the meta-analysis if (a) less than 50% of the applied treatment consisted of IR and (b) the IR condition consisted of fewer than five participants. Whereas we assessed whether the intervention was conducted with clinical populations or not, we have included both, trials with clinical as well as non-clinical participants. Similarly, both, controlled and uncontrolled trials were included. Furthermore, no age or language restrictions were imposed at the
7 search and filtering stage. Missing data in the selected studies were retrieved by contacting the authors directly.
2.3 Coding
A code protocol was developed for the purpose of coding and extracting the needed information from each study included in the meta-analysis. The code protocol was
developed by the author, with the help of supervisor Dr. Morina. Dr. Arntz and Dr. Lancee, both researchers at the department of psychology at the University of Amsterdam,
collaborated in the development of this code protocol. The following information was extracted from each study included in the meta-analysis: (a) study identification, i.e., author(s), title, year of publication, journal, country; (b) sample characteristics, i.e., sample size, percentage female, age (mean and SD), type of population, inclusion criteria, diagnosis, medication, other treatment; (c) treatment methodology, i.e., study objective, type of intervention, number of treatment sessions, duration of treatment sessions, intervention format, treatment manual, inclusion of other interventions; (d) methodological
characteristics, i.e., study design, control condition, follow-up, intend to treat, treatment completers, drop out, executed statistical analysis; (e) measures, i.e., instruments
diagnosing disorder, primary outcome measures, secondary outcome measures; (f) data for calculating effect sizes, i.e., mean and standard deviations for each measure at pre-, post-treatment and follow-up. Studies were coded independently by the author and supervisor to ensure data accuracy.
8 2.4 Analysis
The Comprehensive Meta-Analysis software package (Borenstein et al., 2009) was used to calculate the effect sizes. Effect sizes were calculated by using Hedges’ g, a variation of Cohen’s d that corrects for bias due to small sample sizes (Hedges & Olkin, 1984). The magnitude of Hedges’ g can be interpreted by using the following recommendation: An effect size of 0.2–0.49 is considered small, 0.50–0.79 medium, and equal and larger than 0.80 large (Glass, 1976). The following analysis were computed: 1) Pre and post treatment analysis, 2) Pre and follow-up analysis, 3) Between analysis, 4) Number of session analysis, 5) Disorder specific analysis and 6) Analysis on the secondary outcome measures.
Secondary outcome measures were only analyzed if they were reported in at least three studies. While we have collected the mean and standard deviations for each measure at pre-, post-treatment and follow-up, for the analysis we have substituted the standard deviations at post-treatment and follow-up with the standard deviation at pre-treatment (Morris, 2008).
3. Results
3.1 Search results
The bibliographic search identified 238 studies. Reference list and register search identified five studies, bringing the total number of identified studies to 243 (Fig. 1). Based on screening the abstracts 212 studies were excluded. A common reason for exclusion was not meeting the criteria: the screened studies did not research imagery rescripting or included disorders other than the ones reviewed. Since we expected that most studies
9 researched the efficacy of IR on PTSD, depression and anxiety disorders, we decided to only include these disorders in the meta-analysis. The studies on the efficacy of IR on nightmares were excluded, since several meta-analysis on this subject have been published recently (Casement and Swanson, 2012). Other reasons for exclusion were having fewer than five participants included in the study. The 31 trials assessing the efficiency of imagery
rescripting were retrieved for further assessment. Among these 31 trials, 14 were excluded due to the fact that IR was part of a treatment package, resulting in being less than 50% of the total treatment. From the 17 studies that met inclusion criteria, five studies were
excluded. Two studies (Hunt, 2006; Hunt, 2007) were excluded because of missing data. We did not receive a reply to our email requesting raw data of the mentioned studies from the author. Three studies (Hagenaars et al., 2012; Hagenaars & Arntz, 2012; Dibbets & Arntz, 2012) were excluded, because they were experimental. The three studies were analogue studies using aversive films on a non-clinical population in order to induce anxiety, intrusive thoughts and images. For this reason, it is questionable to what extent the results of these studies can be generalized to a clinical population. Therefore, the experimental studies were not included in the statistical analysis. This resulted in twelve studies that met all inclusion criteria for the meta-analysis. Coding was conducted independently by the author and supervisor and any discordances were resolved by discussion.
3.2 Sample characteristics
Table 1 presents descriptive characteristics for each study included in the meta-analysis. Study sample sizes ranged from six to 34. The total number of participants included in the meta-analysis was 170 (mean age = 35.29, SD= 8.64). All of the included studies were limited to adult samples. The included studies were clinical and all participants met one of the
10 following primary diagnosis: PTSD (41.67%), social phobia (41.67% ), depression (Brewin et al., 2009) and OCD (Veale et al., 2015). All the studies included in the analysis were conducted in an individual format. The sessions of IR ranged from one to 16 and lasted for an average of 74 minutes. Out of the total of twelve studies, seven included a control condition. In the study of Arntz et al. (2013) the participants were compared to themselves in the exploration phase. Frets et al. (2014) also compared the participants to themselves in the no treatment baseline phase. The control condition in the study of Jung and Steil. (2013) consisted of a wait list control group. Lee and Kwon (2013) compared the intervention group to an active control condition with support interviews, while in the study of Nilsson et al. (2012) the control consisted of a reading task. Veale et al. (2015) and Wild et al. (2008) both compared participants to themselves in a non-directive therapy session. Three studies, representing 65 participants, were randomized controlled trials. Three studies conducted a follow-up after only one week (Nilson et al., 2012; Wild et al., 2007; Wild et al., 2008). The majority of the studies
conducted a follow-up between four and 12 weeks. The study of Frets et al. (2014)
conducted the follow-up at six months and Brewin et al. (2009) conducted the follow-up 12 months after treatment completion.
Primary outcome measures as well as secondary outcome measures (depression,
encapsulated beliefs, image distress, image vividness, memory distress) were reported and included in the analysis (see Table 1).
11 3.3 Effect sizes
Pre- versus Post-treatment analysis
In an analysis comparing pre- and post-treatment, which included 12 studies and 170 participants, IR proved to be effective at post-treatment (Hedges’ g = 1.07, standard error, SE = 0.13, 95% CL: 0.81 – 1.32). The calculated effect size is large.
Pre-treatment versus Follow-up analysis
In an analysis comparing pre-treatment to follow-up, including 10 studies and 142 participants, IR proved to be effective at follow-up (Hedges’ g = 1.68, standard error, SE = 0.17, 95% CL: 1.35 – 2.00). The average follow-up period was 14 weeks. The calculated effect size is large.
Between analysis
In an analysis comparing control conditions to IR treatment conditions, including three studies and 65 participants, the IR treatment condition outperformed the control conditions (Hedges’ g = 0.72, standard error, SE = 0.25, 95% CL: 0.23 – 1.20). The calculated effect size is considered medium. IR proved to be more effective than the control conditions including the wait list condition, support interview condition and reading task condition.
Number of session analysis
To examine the effect of the number of sessions on the efficacy of IR, we have performed an analysis comparing studies including one or two sessions and studies including more than two sessions. In the meta-analysis a total number of seven studies included one or two sessions, while five studies included more than two sessions. The results portray that both
12 one or two sessions (Hedges’ g = 0.83, standard error, SE = 0.13, 95% CL: 0.58 – 1.08) and more than two sessions (Hedges’ g = 1.53, standard error, SE = 0.21, 95% CL: 1.11 – 1.94) are effective.
Disorder specific analysis
We have computed an analysis on the effect of IR on a variety of disorders. Five studies included PTSD as the primary diagnosis. IR proved to be effective in reducing the PTSD symptoms at post-treatment (Hedges’ g = 1.76, standard error, SE = 0.41, 95% CL: 0.95 – 2.56) as well as at follow-up (Hedges’ g = 2.06, standard error, SE = 0.45, 95% CL: 1.17 – 2.94). Five studies included social phobia as the primary diagnosis. One can conclude from the analysis that IR is effective in reducing the social phobia symptoms (Hedges’ g = 0.81, standard error, SE = 0.16, 95% CL: 0.49 – 1.12). The effects are maintained at follow-up (Hedges’ g = 1.49, standard error, SE = 0.44, 95% CL: 0.62 – 2.36). Two studies researched the effectivity of IR on reducing the Feeling of Being Contaminated (FBC). IR proved to be effective both at post-treatment (Hedges’ g = 0.85, standard error, SE = 0.23, 95% CL: 0.40 – 1.30) and at follow-up (Hedges’ g = 1.68, standard error, SE = 0.31, 95% CL: 1.06 – 2.29).
The study by Brewin et al. (2009) researched the effectivity of IR on depression and concluded that IR is effective in reducing symptoms of depression (Hedges’ g = 1.61, standard error, SE = 0.46, 95% CL: 0.70 – 2.51). These effects were maintained at follow-up (Hedges’ g = 1.90, standard error, SE = 0.51, 95% CL: 0.89 – 2.90). Only one study researched the effect of IR on OCD. Veale et al. (2015) concluded that IR is an effective treatment for the reduction of OCD symptoms (Hedges’ g = 1.12, standard error, SE = 0.35, 95% CL: 0.43 – 1.81), with results maintaining at follow-up (Hedges’ g = 1.80, standard error, SE = 0.46, 95% CL: 0.91 – 2.69). The calculated effect sizes in all conducted analysis are large.
13 Analysis on the secondary outcome measures
To examine the efficacy of IR on secondary outcome measures we have conducted the
following analysis. The analyzed secondary outcome measures are depression, encapsulated beliefs, imagery distress, imagery vividness and memory distress.
Four studies included depression as a secondary outcome measure. IR proved to be
effective in the treatment of the symptoms of depression (Hedges’ g = 0.71, standard error, SE = 0.28, 95% CL: 0.17 – 1.25), with positive results being maintained at follow-up (Hedges’ g = 0.78, standard error, SE = 0.29, 95% CL: 0.21 – 1.34). The effect sizes are considered medium.
Three studies included encapsulated beliefs as a secondary outcome measure. IR proved to be effective in the treatment of encapsulated beliefs (Hedges’ g = 2.30, standard error, SE = 0.34, 95% CL: 1.64 – 2.97). The large effect size was maintained at follow-up (Hedges’ g = 2.27, standard error, SE = 0.68, 95% CL: 0.94 – 3.61).
Three studies included imagery distress as a secondary outcome measure. IR proved to be effective in the treatment of imagery distress (Hedges’ g = 3.20, standard error, SE = 1.26, 95% CL: 0.74 – 5.67). Also at follow-up IR proved to be effective (Hedges’ g = 2.08, standard error, SE = 0.87, 95% CL: 0.38 – 3.78).
Three studies included imagery vividness as a secondary outcome measure. IR proved to be effective in the treatment of imagery vividness (Hedges’ g = 1.68, standard error, SE = 0.68, 95% CL: 0.34 – 3.01). The large effect size was maintained at follow-up (Hedges’ g = 1.70, standard error, SE = 0.72, 95% CL: 0.29 – 3.12).
Four studies included memory distress vividness as a secondary outcome measure. IR proved to be effective in the treatment of memory distress (Hedges’ g = 1.91, standard
14 error, SE = 0.51, 95% CL: 0.92 – 2.90). Three out of the four studies conducted a follow-up. The large effect size was maintained at follow-up (Hedges’ g = 2.76, standard error, SE = 1.09, 95% CL: 0.63 – 4.89).
4. Discussion
To research the efficiency of IR on PTSD, depression and anxiety disorders, we have conducted this meta-analysis and examined 12 studies including 170 participants. The results from this meta-analysis indicate that IR may be effective in the treatment of PTSD, depression and anxiety disorders (Hedges’ g = 1.07). Analysis of long term treatment outcomes indicated that the benefits of IR were sustained for an average of 14 weeks following treatment completion (Hedges’ g = 1.68). The meta-analysis revealed that IR outperformed the control conditions (Hedges’ g = 0.72), including the wait list condition, support interview condition and reading task condition. Further, findings indicate that IR can be effective after only one or two sessions (Hedges’ g = 0.83). The analysis on the disorders revealed that IR is an effective treatment for PTSD (Hedges’ g = 1.76), social phobia (Hedges’ g = 0.81) and FBC symptoms (Hedges’ g = 0.85). Besides being effective on the primary outcome measures, IR also proved to have a positive effect on the secondary outcome measures, such as depression, encapsulated beliefs, imagery distress, imagery vividness and memory distress. However, it is important to note that the mentioned results are
preliminary. The results should be interpreted with great sensitivity, considering the lack of RCTs and the small sample sizes in this meta-analysis. Future research is required to be able to confirm and expand the aforementioned results.
In eight out of the 12 included studies IR was standardized and based on the treatment manual of Arntz and Weertman (1999). Since not all studies used this standardized form of
15 IR, there is possibly a variation in the IR conduction over the different trials. The lack of theoretical knowledge on the mechanism of IR, and therefore the lack of information on the exact way it should be conducted, leads to a possible variation between the trials. Former research revealed that IR possibly works through modifying the fear memory in order for it to be reconsolidated with a different meaning, so it no longer elicits the same emotional responses (Anrtz, 2012). However, explicit information on how IR works is lacking. Future research should focus on the mechanisms of IR, in order to be able to develop a distinct and unambiguous manner and protocol to conduct IR in the most effective way.
Another form of variation between the trials is the combination of several treatments. The study of Wild et al. (2008) did not exclusively include IR, but also integrated other forms of treatment such as cognitive restructuring. IR was at least 50% in this trial, making it eligible to be included in the analysis. Nevertheless, conducting several treatments within the same trial group leads to carry over effects. This is disadvantageous, since it is not possible to distinguish to what extent the treatments effects were caused solely by IR. In this regard, there is still room for improvement in current research. Future research should focus more on IR as a stand-alone treatment, rather than being part of a treatment package.
Another limitation of the reviewed studies is that only two out of the 12 examined studies included an active control group. The control groups consisted of a reading task condition and a support interview condition. None of the studies compared IR to another treatment. In order to be able to make a well-considered choice between interventions, it is of great importance that future research looks into the effectiveness of IR compared to alternative treatments. Another caveat is that only three of the 12 included studies in the meta-analysis were RCTs. Studies that are not randomized or controlled have more potential for bias and tend to show greater treatment effects (Thornton & Lee, 2000). On
16 the same note, the follow-up period in the current studies varied between one week and one year (Brewin et al., 2009), with an average of 12 weeks. The studies of Wild et al. (2007) and Nilson et al. (2012) reported a follow-up after one week. In the meta-analysis these measurements were labeled as post-treatment, since post data was missing and the follow-up period length was short. In general, the follow-follow-up period in the included studies is
relatively short and should be extended in order to be able to reach a conclusion on the long term effects of IR. While the results from this meta-analysis are promising, from a
methodological point of view the current research is still limited.
Furthermore, publication bias, a type of selection bias that occurs when the publication of research depends on the nature and direction of results, can possibly lead to false-positive associations. In this meta-analysis we have attempted to limit the effect of
publication bias by researching published as well as unpublished studies in registers. There have been various methods proposed in order to determine the presence of publication bias and even correct for it. However, these methods have their limitations. It is therefore of importance to be aware of the existence of this bias and subsequently moderate the strength of conclusions accordingly (Thornton & Lee, 2000).
It is notable that none of the included nor screened studies researched the efficacy of IR on children. It is of relevance that future research looks into this field. If IR proves to be effective in the treatment of children as well, it should be applied as a stand-alone
treatment or incorporated in existing treatments. Another area that has not been explored extensively is the effect of IR on depression. Only one of the included studies (Brewin et al., 2009) looked at the effect of IR on depression symptoms and found it to be effective
17 an effective treatment. The majority of current research is applied to PTSD and social
phobia. Future research should be extended to other disorders, including depression. Imagery rescripting may be an effective treatment for PTSD, depression and anxiety disorders. The found effects proved to be maintained at the long term. Compared to active control treatments as supportive interviews, IR outperformed the former. The results of this meta-analysis should be interpreted carefully, bearing in mind the limited number of
included sample sizes and RCTs. Compared to other exposure based treatments, IR tends to be a more favorable treatment among therapists, since therapists reported feeling less emotionally distressed and experienced less helplessness during the IR treatment (Arntz et al., 2007). Notably, trials including only one session of IR managed to be effective (Nilsson et al., 2012; Veale et al., 2015). With future research focusing on comparing treatments to imagery rescripting in RCTs and researching the underlying mechanisms, the development of a promising treatment is in action.
Acknowledgement
I would like to thank my supervisor Dr. Nexhmedin Morina for his help and supervision during the conduction of this meta-analysis and the development of the code protocol. Furthermore, I would like to gratefully acknowledge Dr. Arnoud Arntz and Dr. Jaap Lancee for their help and thoughtful comments during the process.
18
References
Arntz, A., & Weertman, A. (1999). Treatment of childhood memories: Theory and practice. Behaviour Research and Therapy, 37, 715–740.
Arntz, A., Tiesema, M., Kindt, M. (2007). Treatment of PTSD: A comparison of imaginal exposure with and without imagery rescripting. Journal of Behavior Therapy and Experimental Psychiatry, 38, 345-370.
Arntz, A. (2012). Imagery rescripting as a therapeutic technique: review of clinical trials, basic studies, and research agenda. Journal of Experimental Psychopathology, 3, 198–208.
Borenstein, M., Hedges, L. V., Higgins, J. P. T., Rothstein, H. R. (2009) Software, in Introduction to meta-analysis. Chichester, UK: John Wiley & Sons. doi: 10.1002/9780470743386.ch44. Brewin, C. R., Gregory, J. D., Lipton, M., Burgess, N. (2010). Intrusive images in psychological
disorders. Characteristics, neural mechanisms, and treatment implications. Psychological Review, 117, 210–232.
Casement, M. D. & Swanson, L. M. (2012). A meta-analysis of imagery rehearsal for post-trauma nightmares: Effects on nightmare frequency, sleep quality, and posttraumatic stress. Clinical Psychology Review, 32, 566–574.
Edwards, D. J. A. (2007). Restructuring implicational meaning through memory based imagery: Some historical notes. Journal of Behavior Therapy and Experimental Psychiatry, 38, 306-316.
Foa, E. B., Riggs, D. S., Dancu, C. V., Rothbaum, B. O. (1993). Reliability and validity of a brief instrument for assessing posttraumatic stress disorder. Journal of Traumatic Stress, 6, 459- 473.
Forcato, C., Burgos, V. L., Argibay, P. F., Molina, V. A., Pedreira, M. E., Maldonado, H. (2007). Learning and Memory, 14, 295-303.
19
Frets, P. G., Kevenaar, C., van der Heiden, C. (2014). Imagery rescripting as a stand-alone treatment for patients with social phobia: A case series. Journal of Behavior Therapy and Experimental Psychiatry, 45, 160-169.
Glass, G. V. (1976). Primary, secondary and meta-analysis of research. Educational Researcher, 10, 3– 8.
Hagenaars, M.A. & Arntz, A. (2011). Reduced intrusion development after post-trauma imagery Rescripting: An experimental study. Journal of Behavior Therapy and Experimental Psychiatry, 43, 808-814
Holmes, E. A., Arntz, A., Smucker, M.R. (2007). Imagery rescripting in cognitive behavior therapy: Images, treatments techniques and outcomes. Journal of Behavior Therapy and Experimental Psychiatry, 38, 297-305.
Holmes, E. A., & Bourne, C. (2008). Inducing and Modulating Intrusive Emotional Memories: A review of the trauma film paradigm. Acta Psychologica, 127, 553-566.
Rachman, S., & De Silva, P. (1978). Abnormal and normal obsessions. Behavior Research and Therapy, 16, 233-248.
Rusch, M. D., Grunert, B. K., Mendelsohn, R. A., Smucker, M. R. (2000). Imagery Rescripting for Recurrent, Distressing Images. Cognitive and Behavioral Practice, 7, 173-182.
Smucker, M. R., Dancu, C., Foa, E. B., Niederee, J. L. (1995). Imagery rescripting: A new treatment for survivors of childhood sexual abuse suffering from posttraumatic stress. Journal of Cognitive Psychotherapy, 9, 3-17.
Thornton, A., Lee, P. (2000). Publication bias in meta-analysis: its causes and consequences. Journal of Clinical Epidemiology, 53, 207-216.
Whitaker, K., Brewin, C. R., Watson, M. (2010). Imagery Rescripting for Psychological Disorder following Cancer: A case study. British Journal of Health Psychology, 15, 41-50.
20
Wild, J., Hackmann, A., Clark, D. M. (2007). When the present visits the past: updating traumatic memories in social phobia. Journal of Behavior Therapy and Experimental Psychiatry, 38, 386- 401.
Wild, J., Hackmann, A., Clark, D. M. (2008). Rescripting early memories linked to negative images in social phobia: A pilot study. Behavior Therapy, 39, 47-56.
21 Appendix Fig. 1 Flow Chart Scr e e n in g El ig ib ili ty
Records identified through database searching (n = 238 ) PsychInfo = 98 Medline = 52 Web of Science = 88
Additional records identified through other sources
(n = 5 ) Records screened (n = 243 ) Records excluded (n = 212 ) Reasons:
Did not research imagery rescripting or included disorders other than the
ones reviewed
Full-text articles assessed for eligibility
(n = 31 )
Full-text articles excluded (n = 14)
Reasons:
< 5 patients were treated with IR < 50% of the applied treatment
consisted of IR
Studies meeting inclusion criteria (n = 17 ) Studies included in quantitative synthesis (meta-analysis) (n = 12 ) Id e n tifi cat ion In cl u d e d Records excluded (n = 5 ) Reasons: Experimental studies (n = 3) Missing data (n = 2)
21
Table 1. Overview of included trials
Study and
type of treatment (number of sessions) N used in pre-post analyses % female participant Age M (SD)
Primary outcome Additional outcome variables Follow-up Diagnosis (%) Arntz et al., 2013 ImRs (10 sessions) Control 10 20% 39.9 (12.24) PSS BDI 12 weeks PTSD (100%) Brewin et al., 2009 IR (8 sessions)
10 80% 41.3 (n.r.) BDI 12 months Depression (100%)
Frets et al., 2014 IR (11.2 sessions)
Control
6 50% 32.83 (n.r.) SPS 6 months Social Phobia
(100%)
Jung & Steil, 2013 CRIM (2 sessions) Control WL 28 100% 37 (n.r.) FBC BDI-II 4 weeks PTSD (100%) Kindt et al., 2007 IR (10 sessions) 25 80% 33 (10.1) PSS-SR 4 weeks PTSD (100%)
Lee & kwon, 2013 IR (3 sessions) Control (3 sessions)
23
13
10
52.17 23.92 (3.35) LSAS Encapsulated beliefs Image distress Image vividness Memory distress
12 weeks Social Phobia (100%)
23
Note: ImRs: Imagery Rescripting; PSS: The Posttraumatic Symptom Scale; BDI: Beck Depression Inventory; IR: Imagery Rescripting; CRIM: Cognitive Restructuring and Image
Modification; WL: Waitlist Control; FBC: Feeling of Being Contaminated; PSS-SR: The PTSD Symptom Scale Self-Report; LSAS: Liebowitz Social Anxiety Scale; SIAS: Social Interaction Anxiety Scale; CAPS: Clinician Administered PTSD Scale; Y-BOCS: Yale-Brown Obsessive-Compulsive Scale; SPWSS: Social Phobia Weekly Summary Scale; n.r.: not reported
Nilsson et al., 2012 IR (1 session)
Control
14 43% 33.5 (12.9) SIAS Image distress Image vividness
1 week Social Phobia (100%)
Raabe et al., 2015
IR (16 sessions)
8 75% 34.4 (8.2) CAPS BDI-II 12 weeks PTSD (100%)
Steil et al., 2011 CRIM (2 sessions) 9 100% 43.78 (8.98) FBC 6 weeks PTSD (100%) Veale et al., 2015 IR (1 session) Control
12 41.67% 40 (n.r.) Y-BOCS BDI 12 weeks OCD (100%)
Wild et al., 2007 IR (1 session) 14 35.71% 28.64 (3.75) SPWSS Encapsulated beliefs Image distress Image vividness Memory distress
1 week Social Phobia (100%)
Wild et al., 2008 IR (1 session)
Control
11 63.64% 35.18 (9.36) Liebowitz Anxiety Encapsulated beliefs Memory distress Image distress Image vividness
1 week Social Phobia (100%)
23
Table 2. IR Pre- versus Post-treatment analysis
Table 3. IR Pre- treatment versus Follow-up analysis
Subgroup Time point Comparison Study name Outcome Statistics for each study Hedges's g and 95% CI
Hedges's Standard Lower Upper
g error limit limit
IR Pre-Post within Arntz et al. (2013) 2,499 PTSD 0,629 1,266 3,732 IR Pre-Post within Brewin et al. (2009) 1,606 Depression 0,461 0,702 2,510 IR Pre-Post within Frets et al. (2014) 1,268 Social Phobia 0,502 0,284 2,252 IR Pre-Post within Jung & Steil. 2013 0,991 Contamination fear 0,314 0,376 1,605 IR Pre-Post within Kindt et al. (2007) 1,226 PTSD 0,260 0,717 1,736 IR Pre-Post within Lee & Kwon, 2013 0,823 Social Phobia 0,306 0,223 1,422 IR Pre-Post within Nilsson et al. (2012) 0,379 Social Phobia 0,344 -0,295 1,053 IR Pre-Post within Raabe et al. (2015) 2,022 PTSD 0,595 0,856 3,189 IR Pre-Post within Steil et al. (2011) 0,682 Contamination fear 0,341 0,013 1,351 IR Pre-Post within Veale et al. (2015) 1,121 OCD 0,353 0,430 1,813 IR Pre-Post within Wild et al. (2007) 0,766 Social Phobia 0,368 0,045 1,488 IR Pre-Post within Wild et al. (2008) 1,042 Social Phobia 0,356 0,344 1,740 1,066 0,132 0,808 1,324
-4,00 -2,00 0,00 2,00 4,00
Favours A Favours B
Meta Analysis
Meta Analysis
Subgroup Time point Comparison Study name Outcome Statistics for each study Hedges's g and 95% CI
Hedges's Standard Lower Upper
g error limit limit
IR Pre-FU within Arntz et al. (2013) 2,501 PT SD 0,629 1,267 3,734
IR Pre-FU within Brewin et al. (2009) 1,896 Depression 0,513 0,890 2,902
IR Pre-FU within Frets et al. (2014) 1,378 Social Phobia 0,526 0,348 2,409
IR Pre-FU within Jung & Steil. 2013 1,718 Contamination fear 0,411 0,913 2,524
IR Pre-FU within Kindt et al. (2007) 1,475 PT SD 0,285 0,917 2,033
IR Pre-FU within Lee & Kwon, 2013 2,427 Social Phobia 0,563 1,322 3,531
IR Pre-FU within Raabe et al. (2015) 2,805 PT SD 0,768 1,299 4,311
IR Pre-FU within Steil et al. (2011) 1,619 Contamination fear 0,486 0,666 2,571
IR Pre-FU within Veale et al. (2015) 1,797 OCD 0,455 0,906 2,688
IR Pre-FU within Wild et al. (2008) 0,911 Social Phobia 0,339 0,246 1,576
1,677 0,166 1,352 2,001
-4,00 -2,00 0,00 2,00 4,00
Fav ours A Fav ours B
Meta Analysis
25
Code protocol: The Efficacy of Imagery Rescripting on PTSD, Depression and Anxiety Disorders: A meta-analysis
ID study (Given number): Initials rater:
A. Study identification Author: Title: Publication year: Journal: Country: B. Sample characteristics Sample size:
Gender □ female …….. % □ not reported
Age □ M: …….. □ SD: …….. □ not reported
Type of population □ Clinical: ……..……..…….. □ Non-clinical □ Other: ……..……..……..
Inclusion criteria sample: ……..……..……..
Exclusion criteria sample: ……..……..……..
Diagnoses
□ Primary (Axis-1) diagnoses (%): □ PTSD ………..% □ Depression …..……%
□ Anxiety Disorder (specify) : ………% □ Other: ………%
Comorbidity
□ Secondary Axis-1 diagnosis(disorder, %) : □ Yes, ……….%
26
□ Axis-2 diagnoses(disorder, %): □ Yes, ………. % □ No
□ not reported □ No diagnosis assessed or reported
□ Participants with high scores of ………. included
Medication □ Yes □ No □ not reported Before/during intervention?: ……..……..……..
Type of medication: ……..……..……..
Other treatment □ none
□ Ongoing treatment (type & duration): ……….……..……..……… □ Treated before study (type & duration): ……….……..……..……..
C. Treatment methodology
Study objective: ……….
Type of intervention: ………. Specify IR type:
□ With cognitive restructuring component (English version) □ Without cognitive restructuring component
□ Other: ………. □ Unknown
Number of sessions: ……..……..……..
Duration of sessions: ……..……..……..
Intervention format □ Individual sessions
□ Group sessions
□ Combination of both
Standardization of treatment (Treatment manual) □ Yes □ No □ not reported
Inclusion of other (than IR) intervention method □ Yes, ……..……..……..
Percentage of IR total treatment : …….. % □ No
27
D. Methodological characteristics
Study design □ RCT □ Non-randomized controlled study □ Uncontrolled trial □ Multiple Baseline design
□ Case series □ Other: ……..……..……..
Control condition □ Active : ……… □ Inactive : ……… □ None
Follow-up (period in weeks): □ Yes, ………. □ No
Intend to treat (N): ………
Treatment completers (N): ………
Drop out (N): ………
Executed statistical analysis: ……….……….
E. Measures
Measures of disorder
□ Instrument diagnosing PTSD: ………..
□ Instrument diagnosing Depression: ………. □ Instrument diagnosing Anxiety disorder : ……….
Specify anxiety disorder: □ Social anxiety disorder
□ Other : ………..……..……..
Measures of outcome
Primary outcome measure: ……….
28
F. Data for calculating effect sizes
Experimental condition Control condition Comorbidity Notes PAM SAM …. Pre-treatment M(SD) Post-treatment FU 1 FU 2 PAM SAM …. Pre-treatment Post-treatment FU1 FU 2 PAM SAM …. Pre-treatment Post-treatment FU 1 FU 2
