Healing through giving testimony: An empirical study with Sri Lankan torture survivors

Abstract

Sri Lanka has recently emerged from a three decade long civil war between government forces and the Liberation Tigers of Tamil Eelam. Behind the actual arena of conflict, forms of organised violence were often perpetrated on ordinary Sri Lankans who came into contact with law enforcement officials and other state authorities. The effects of these encounters on mental health, well-being, and community participation can be severe and long-lasting. Considering the generally poor availability of mental health services in many low-income countries, brief efficient interventions are required to enhance the lives of individuals and their families affected by torture, trauma, or displacement. In this context, the present study evaluated the effectiveness of testimonial therapy in ameliorating the distress of Sri Lankan survivors of torture and ill-treatment. The results indicated that over a 2- to 3-month period, psychosocial functioning was significantly enhanced in the therapy group compared to the waitlist control group. The general benefits of testimonial therapy, the ease with which it can be incorporated into ongoing human rights activities, and its application by trained nonprofessionals encourage greater use of the approach.

Keywords

human rights ill-treatment psychosocial functioning Sri Lanka testimonial therapy torture

Sri Lanka recently emerged from a three decade long civil war with massive human rights violations alleged by all sides of the conflict. However, behind the actual arena of conflict, torture and other forms of organised violence were perpetrated by law enforcement officials and other state authorities on ordinary Sri Lankans. The effects of these encounters on mental health, well-being, and community participation can be severe and long-lasting. Considering the generally poor availability of mental health services in many low-income countries, brief efficient interventions are required to enhance the lives of individuals and their families affected by torture, trauma, or displacement. The present study evaluated the effectiveness of one such brief intervention, testimonial therapy, in addressing the distress of Sri Lankan survivors of torture and ill-treatment.

Since 1992, Jana Sansadaya (People’s Forum), a nongovernmental organisation in Sri Lanka, has assisted survivors of torture and other rights violations to seek justice. Survivors are offered financial and legal support, given temporary protection, or referred to specialist medical treatment. Over the years, organisational activities have expanded to raise rights awareness and restore a rights-conscious society via solidarity group meetings, street campaigns, seminars, and debates. However, seeking justice within the country’s legal system is a lengthy and arduous process that may drag on for 5 to 10 years, in spite of a requirement that human rights violation cases need to be filed within 1 month of the violation (B. Fernando, 2010; B. Fernando & Puvimanasinghe, 2004).

Consequently, torture survivors have to endure not only the distress resulting from their initial traumatic experience, but also the effect of added daily stressors such as constant anxiety about possible reprisal, possible loss of livelihood, and social stigma arising from mental problems or inability to work (G. A. Fernando, 2008; G. A. Fernando, Miller, & Berger, 2010). Western-style mental health interventions administered by tertiary-trained professionals are expensive and consequently scarce. There is an urgent need for brief, effective, culturally appropriate, psychosocial interventions that can be administered by trained nonprofessionals. Testimonial therapy (TT) offers such a possibility (Agger, Igreja, Kiehle, & Polatin, 2012). However, objective empirical evaluations have been difficult to conduct and information about exactly what aspects of the intervention are effective, for whom the intervention is most effective, and what culturally appropriate measures are most useful, remains limited (Agger et al., 2012; Ginzburg & Neria, 2011; Igreja, Kleijn, Schreuder, van Dijk, & Verschuur, 2004; McFarlane & Kaplan, 2012; Patel et al., 2007; Saraceno et al., 2007).

Testimonial therapy

Testimonial therapy is an individual psychotherapy for survivors of torture and other types of violence that has the narration of the survivors’ traumatic experiences as its central component (Agger et al., 2012). The trauma stories are recorded, jointly edited by the therapist and survivor, and compiled into a document, which may be used by the survivor to raise public and international awareness about human rights violations (Igreja et al., 2004; Weine, Kulenovic, Pavkovic, & Gibbons, 1998). TT was originally developed by two Chilean psychologists documenting the experiences of political detainees during the military dictatorship in that country (Cienfuegos & Monelli, 1983). They discovered that telling the trauma story and reexperiencing the suffering in a safe and trusting environment had positive therapeutic value to the majority of their patients and resulted in reduced psychopathological symptoms such as anxiety or acute depression.

TT and variations of it have been successfully adapted to diverse cultural and political contexts. For example, 20 Bosnian refugees in the US afforded six sessions of TT (Weine et al., 1998) showed significant improvement in posttraumatic stress disorder (PTSD) diagnosis rates, PTSD symptom severity, depression, and global functioning. Forty-three Sudanese refugees living in a Ugandan refugee settlement showed significant decreases in the frequency and severity of PTSD symptoms after receiving four sessions of a variation of TT described as narrative exposure therapy (NET; Neuner, Schauer, Klaschik, Karunakara, & Elbert, 2004). NET was superior to supportive counselling and psycho-education in ameliorating PTSD symptoms. A study with 18 elderly survivors of political repression in Romania also found five sessions of NET to significantly reduce PTSD symptoms and depression (Bichescu, Neuner, Schauer, & Elbert, 2007). Several case studies with distressed asylum seekers and refugees have also described the benefits of TT in alleviating psychological suffering (Agger et al., 2012; Agger & Jensen, 1990; van Dijk, Schoutrop, & Spinhoven, 2003). Survivors of civil war in rural Mozambique showed significant improvement in posttraumatic stress symptoms following a one session intervention of TT (Igreja et al., 2004). Here, the nonintervention group showed improvements similar to the therapy group which were attributed to uncontrolled interaction effects between therapy and nontherapy participants in a close rural community. Despite these shortcomings, the authors suggested that the testimony method was a feasible intervention in regions where mental health resources were scarce and there were few opportunities to assist traumatised people (Igreja et al., 2004).

There appear to be no empirical studies specifically investigating the effectiveness of TT with a Sri Lankan population. The adaptation of TT to different countries or cultural groups requires sensitivity to local customs, spirituality, and political contexts (Agger et al., 2012; McFarlane & Kaplan, 2012). For example, the bulk of TT research has focused on assessing the PTSD symptom clusters of intrusion, avoidance/dissociation, and hyperarousal as measured by, for instance, the PTSD Symptom Scale (Bichescu et al., 2007; Igreja et al., 2004; Weine et al., 1998). However, the adequacy of the PTSD category to capture the complexity of traumatic distress in persons facing human rights (HR) violations in collectivist societies has been questioned (Bracken, Giller, & Summerfield, 1995; Isakson & Jurkovic, 2013; McFarlane & Kaplan, 2012). For example, McFarlane and Kaplan (2012) argued that empirical evaluations of interventions need to also consider contextual aspects of functioning and impairment if the full effect of an intervention is to be appreciated. The “psychosocial” construct, comprising mental and physical health, social relations, religion, culture, and values, has been suggested to more fully represent traumatic distress in collectivist societies (Agger et al., 2012; Ginzburg & Neria, 2011; Isakson & Jurkovic, 2013; McFarlane & Kaplan, 2012; Strang & Ager, 2003). Hence, particular effort has been made in the current study to select instruments measuring constructs more appropriate to the population under investigation.

In 2009, the testimonial method was introduced to HR organisations providing legal assistance to torture survivors in Sri Lanka (Perera, Puvimanasinghe, & Agger, 2009). A manual was compiled and a group of HR activists and community workers were trained in taking testimonies, as well as in monitoring and evaluation. Features of TT designed for Sri Lanka, and as described in the manual, included: (a) a brief format (three to four sessions) including psycho-education, monitoring, and evaluation; (b) use of nonprofessional therapists; (c) a testimonial ceremony; (d) a short mindfulness meditation; and (e) a training course for therapists. The testimonial ceremony involves the presentation of the testimony document to participants in the presence of significant others and is considered a crucial element of this brief-format TT. The ceremony symbolises the transition of the survivors’ private pain into the public, legal, and political arena (Agger et al., 2012; Ginzburg & Neria, 2011) and links the healing process at the individual level to the social memory of the community.

Theoretical underpinnings

Although the mechanisms of action of TT are not conclusively known, a plausible argument is as follows. Traumatic events such as being tortured are extraordinary occurrences that are perceived as life-threatening or likely to cause severe injury (Schauer, Neuner, & Elbert, 2005). Memories of these events may be encoded and stored differently in the brain than normal day-to-day events. That is, memory of normal life episodes involves the integration of spatial, temporal, and contextual information of an experience together with its affective, cognitive, and sensory-motor components, and is stored as personal narratives (van der Kolk & Fisler, 1995). However, during trauma, extreme emotional arousal may interfere with the process of integration, resulting in memories of traumatic events being stored as isolated fragments such as visual images or terrifying fear, while the person’s ability to give a coherent story or recall the chronological order of events is also hampered (van der Kolk & Fisler, 1995; van der Kolk, Hopper, & Osterman, 2001). Terror, pain, fear, confusion, and anguish may be reexperienced. Gradually, individuals learn to avoid environmental cues that trigger reexperiencing and may consequently withdraw from social interaction and participation in social activities (Schauer et al., 2005; van Brakel et al., 2006).

According to cognitive behavioural theory, gradual exposure to painful memories in safe and trusting surroundings desensitises the person to the traumatic memories, thus reducing reexperiencing and avoidance (van Dijk et al., 2003). The testimonial process facilitates desensitisation and assists the integration of dissociated memory fragments by encouraging participants to narrate their traumatic experiences in chronological sequence together with accompanying emotions and cognitions. Rereading and editing testimonies further enhances desensitisation (van Dijk et al., 2003; Weine et al., 1998). Giving testimony is also an opportunity to reappraise and adjust thought patterns detrimental to psychological well-being.

The effective mechanism of TT may be related to the use of expressive writing as a therapeutic technique (Frattaroli, 2006; Pennebaker, 1993). Frattaroli (2006) concluded from a meta-analysis of 146 randomised studies of experimental disclosure regarding traumatic events that repeated exposure was the most likely effective mechanism for the effectiveness of writing interventions. Greater dosage of exposure was more effective than less exposure, and those studies that used participants with a history of trauma showed greater effect sizes for subjective impact than those studies that did not specifically include participants with trauma history. TT includes a verbal description of the person’s traumatic event, a written and rewritten version of the event, and a ceremonial expression of the event, suggesting that exposure is maximised in several modalities in a short period of time.

More recently, Agger et al. (2012) emphasised the role of spirituality and mindfulness in helping trauma survivors face disturbing and often fragmented sensations, images, and feelings, and integrate these through a healing process. Automatic physiological responses to the experience and reexperiencing of trauma also need to be extinguished. The narrative and ceremony components of TT allow acceptance and a sense of purposeful identity to be developed, while the calm, respectful atmosphere allows automatic anxiety responses to dissipate.

Challenges of implementing therapies in low-income, non-western countries

Intervention studies conducted in low-income, non-western settings face unique challenges. Cultural characteristics determine aspects of communication and interaction, how emotion is expressed, and commitment and compliance to protocols. The target population often has limited literacy. Difficulties are exacerbated when studies are conducted at the grassroots level where there is a paucity of mental health staff and material resources, and few, if any, clinical measurement tools in local languages. In addition, minimising contamination effects on a no-treatment control group without interfering with normal customs and social expression is particularly difficult (Agger, Raghuvanshi, Khan, Polatin, & Laursen, 2009; Igreja et al., 2004). Consequently, many studies are plagued by problems related to small sample sizes, participant dropout, control group contamination, and incomplete follow-up assessments (Kazdin, 2003). However, such research needs to be conducted: Empirically informed debate will enhance the effectiveness of brief affordable interventions in conflict zones around the world. The present study, with many of the same challenges, sought to maximise information about the impact and appropriateness of TT, as well as identify conditions which facilitate or hinder its effective application.

Method

Participants

The study included 26 Sri Lankan survivors of torture and ill-treatment. They were selected by HR activists and community workers from a list of people who sought legal redress at Jana Sansadaya or one of its regional offices and had indicated trauma-related symptoms such as an inability to give a coherent story, incessant crying, or other emotional outbursts. No participant approached for the study declined participation. Exclusion criteria were applied to individuals showing signs of severe depression or other severe symptoms, or who were not motivated to give their testimony (Perera et al., 2009). All participants were 18 years or over, Buddhist, and conversant in the Sinhala language. All had experienced violence personally or had significant others experience of violence within the previous 5 years. Participants were not paid for their participation but their travelling expenses were reimbursed. All survivors were served tea and food according to local custom and participants attending the testimony ceremonies partook in a communal meal afterwards. Table 1 contains the sociodemographic characteristics of participants in the two groups.

Table 1.

Sociodemographic characteristics of participants (N = 26).

Characteristic	Control N (%)*	Therapy N (%)	Total N (%)
Gender
Male	8 (61.5)	8 (61.5)	16 (61.5)
Female	5 (38.5)	5 (38.5)	10 (38.5)
Status
Primary	9 (69.2)	9 (69.2)	18 (69.2)
Secondary	4 (30.8)	4 (30.8)	8 (30.8)
Nature of violation
Torture	10 (76.9)	10 (76.9)	20 (76.9)
Ill-treatment	2 (15.4)	2 (15.4)	4 (15.4)
Psychological	1 (7.7)	1 (7.7)	2 (7.7)
Perpetrator
Police	10 (76.9)	9 (69.2)	19 (73.1)
Military	1 (7.7)	1 (7.7)	2 (7.7)
School	1 (7.7)	2 (15.4)	3 (11.5)
Prison	1 (7.7)	1 (7.7)	2 (7.7)
Age (M, SD)	39.4 (14.71)	41.5 (12.14)	40.5 (13.26)
Education
0-5 years	1 (7.7)	1 (7.7)	2 (7.7)
6-9 years	4 (30.8)	6 (46.2)	10 (38.5)
10+ years	8 (61.5)	6 (46.2)	14 (53.8)

*Note. Number and percentage are given for all characteristics except age, for which the mean and standard deviation are provided.

Measures

Sociodemographic information

Several items tapping aspects of the torture event or personal characteristics were developed, including: participant status (“primary torture survivors” had personally experienced torture or ill-treatment; “secondary survivors” had close relatives who were the primary survivors and often suffered many of the same psychological distress and social adversity symptoms as primary torture survivors) and nature of violation (actual physical torture; ill-treatment not amounting to torture, e.g., punching and kicking; and psychological torture only). Importance of religion (not important, somewhat, quite, very), and age, gender, and education levels were also recorded.

Sri Lanka Index of Psychosocial Status (SLIPSS-A)

The SLIPSS-A was specifically designed as a general measure of psychosocial functioning in rural Sri Lankan, Sinhalese-speaking adults affected by trauma (G. A. Fernando, 2008). The scale consists of 26 items rated on a 5-point scale from 0 (never) to 4 (6–7 days per week) assessing themes related to anger, loss of hope, fear, bad memories, lack of self-care, and not performing duties. Lower scores indicate better psychosocial functioning.

The SLIPSS-A was developed in the colloquial Sinhala language using ethnographic methods mostly related to the suffering experienced in the 2004 tsunami (G. A. Fernando, 2008). The final instrument was subsequently completed by 170 Sri Lankans, mostly of rural and lower to lower to middle socioeconomic status. Approximately 50% reported exposure to at least one traumatic event such as the tsunami or war. The SLIPSS-A reliably distinguished between people exposed to tsunami-related traumatic events and those who had not been exposed after controlling for gender and sample type. In addition, the scale correlated highly negatively with a one item measure of life satisfaction and highly positively with a Sinhalese translation of the PTSD Checklist Civilian Version (Weathers, Litz, Huska, & Keane, 1994). Internal reliability was reported as Cronbach’s alpha = .92.

Participation Scale (P-scale)

The P-scale is an 18-item, interview-based instrument measuring social participation in people affected by leprosy, disability, or other stigmatising conditions (van Brakel et al., 2006). Example items include: “Do you visit other people in the community as often as other people do?” and “Are you as socially active as your peers are? (e.g., in religious/community affairs).” Respondents can initially respond yes (0) or no; and if the answer was “no,” rate the difficulty of the problem 1 (no problem) to 5 (large problem). Higher scores on the P-scale reflect more restricted social participation.

The scale was developed simultaneously in seven languages in Nepal, Brazil, and India (with the master scale remaining in English) to maximise cross-cultural application. Cronbach’s alpha was reported as .92 and interinterviewer reliability (r = .80) and intrainterviewer stability (r = .83) were also satisfactory. The P-scale significantly correlated (r = .44, p < .005, n = 227) with expert ratings of participation restriction. Severity of participation restriction was categorised based on the distribution of scale scores in the control and clinical populations (< 13 = no significant restriction, 13–22 = mild; 23–33 = moderate; 34–53 = severe, and > 53 =extreme). Although the P-scale was not specifically developed for torture survivors, it was designed for use in rehabilitation, stigma reduction, and social integration programs where lack of social participation is indicative of poorer mental health.

World Health Organization Five Well-being Index (WHO-5)

The WHO-5 was used to measure emotional well-being including positive mood, vitality, and general interests (WHO, 1998). It contains five items rated on a 6-point scale from 5 (all of the time) to 0 (at no time). Total scores range from 0 to 25, with higher scores indicating better emotional well-being. Example items include, “How often have you felt cheerful and in good spirits?” and “How often have you felt active and vigorous?” The instrument has been recommended as a screening tool for depression with the cut-off score set at 13 (Henkel et al., 2003; WHO, 1998).

The WHO-5 has been translated into more than 25 languages with satisfactory internal consistency for several Asian language versions, for instance, Japanese (alpha = .89; Awata et al., 2007), Thai (alpha = .87; Saipanish, Lotrakul, & Sumrithe, 2009), and Malay (alpha = .86; Momtaz, Hamid, Yahaya, & Ibrahim, 2010). Additionally, in Bangladesh 86% of torture survivors (n = 225) scored less than 13 points on the WHO-5 (Wang, Haque, Masum, Biswas, & Modvig, 2009) while 23 Indian torture survivors had a pretherapy score mean of 7.7 (Agger et al., 2009).

Forward- and back-translation

No translation was required for the SLIPSS-A as it was written and developed in Sinhala. The WHO-5 and P-scales were translated in several stages: (a) Forward translations involved translating each scale item from English to Sinhala with the aid of instructions in the P-scale manual and the English–Sinhala dictionary (Malalasekera, 1997); (b) Sinhala-worded items were independently back-translated into English by two bilingual professionals who were blind to the original translations; (c) the final versions for each item were decided through group discussion and consultation. (Instruments, information, and consent forms used in the study in Sinhalese language are available on request).

Qualitative data collection and analysis

All testimonies were recorded (with consent), orthographically transcribed, and then qualitatively analysed by the first author to identify prominent themes in the completed testimonies (i.e., thematic analysis; Braun & Clarke, 2006). Interviews after the ceremonies and follow-up assessments 2 to 3 months later were conducted to ascertain the thoughts and feelings of the survivors about their experiences with TT.

Procedure

Research commenced after receiving approval from the University of New England Ethics Committee. The study was carried out during April to June 2010 by the therapist (first author of this study) who was assisted by an HR activist. Both had received training in taking testimonies. Initially, study objectives and procedures were explained and schedules drawn up to ensure compatibility with organisational workload and the availability of resources. The first author contributed to the development of the manual for this version of TT and hence adherence to treatment protocols was assured.

Potential participants were sent letters containing a brief description of the study and were invited to attend an interview at the organisation’s headquarters or a regional office. The protocol included five sessions as follows: The first session with all participants covered four aspects, including, (a) introducing the therapist and establishing the therapist’s working relationship with the organisation, with the goal of encouraging participants to extend to the therapist the trust and confidence they had placed in the organisation; (b) rapport-building and alleviating distrust and fear; (c) informing participants of the study rationale and the testimonial process; and (d) confidentiality issues and written consent. This was followed by the collection of sociodemographic details and pretherapy assessment. Survivors who were assigned to the treatment group were invited to give testimony on a mutually convenient date 1–2 weeks later. Control group participants were informed that they would be contacted again and offered testimony at a future date. Assignment to condition was decided by using the characteristics of gender, participant status, and nature of violation, to match pairs of participants as closely as possible. Participants were then randomly assigned to either the treatment group to receive TT, or to the waiting-list type control condition.

The second session was conducted according to the protocol in the manual (Perera et al., 2009). Each survivor was individually interviewed to give details of the main event causing distress. If there was more than one event, the survivor was assisted to separate overlapping stories. During narration, he or she was encouraged to reveal attending emotions, cognitions, and bodily sensations at each stage of the traumatic event, as well as experiences after the event.

Testimonies were audio recorded with the consent of each participant. No one refused. Each session lasted about 60 to 90 minutes. Session 2 was carried out over 2 days for two survivors who had difficulty in constructing a coherent narrative or avoided essential details. This additional time was to avoid their distress becoming overwhelming. The session concluded with the therapist and survivor joining in a short mindfulness meditation as described in the manual (Perera et al., 2009). During the third session, the audio-taped testimony which had been transcribed by the therapist was read out to the survivor in a loud voice. The survivor was requested to edit or add to the testimony in any way he or she desired. This session also ended with a short mindfulness meditation. One testimony was edited over the telephone and another read out at the survivor’s home due to unexpected illness. In the fourth session, the typed testimony, compiled into an attractive document, was presented to the survivor at the testimonial ceremony attended by family members, other torture survivors, HR activists, and media personnel. Consequently, six ceremonies were organised in the towns of Colombo, Galle, Kalutara, Ratnapura, and Nuwara Eliya. Testimony ceremonies were conducted approximately 4–6 weeks after the taking of testimonies.

A Sri Lankan custom at the beginning of any ceremonial event is the lighting of the oil lamp by dignitaries present. At the testimony delivery ceremonies, the traditional oil lamp was lit by the torture survivors, who were then welcomed with betel leaves (a mark of respect) and seated at the head table to signify honour and glory. Two to four survivors were honoured at each event. The testimony-taker or HR activist read out the testimonies to the audience signifying the symbolic transition of the survivor’s private pain into the public realm (Agger et al., 2009; Perera et al., 2009). The document was publicly signed by the therapist and survivor and delivered to the survivor by the event’s chief guest amidst applause and the flashing of cameras. Presentation was followed by speeches commending the courage and resilience of the survivor and a pledge to support his or her quest for justice. Sometimes a discussion ensued around the survivors’ story and the event concluded with a communal meal. Follow-up assessment (Session 5) was carried out 4–6 weeks after the testimonial ceremony.

Statistical procedures

The approach reported here for the evaluation of the intervention is a two-group, two-occasion mixed design ANOVA. There is still dispute as to whether this approach or the multiple regression approach predicting residualised change scores is best in these situations (e.g., Allison, 1990). However, here the findings were the same using both approaches hence we decided to report the mixed ANOVA approach. Effect size calculations are taken from Morris (2008). Reliable change index (RCI) scores and clinical significance (Jacobson & Truax, 1991) are also reported to identify which participants showed improvement, decline, or no change. Traditional ANOVA analyses identify group mean changes but do not assess change at the individual level (Atkins, Bedics, McGlinchey, & Beauchaine, 2005; Deane, Kelly, Crowe, Coulson, & Lyons, 2013). RCI analyses provide information on which individuals have improved (or declined) beyond chance or beyond the inherent error of the measurement instrument, and clinical significance refers to information on meaningful clinical change from a dysfunctional range to a functional range (Jacobson, Roberts, Berns, & McGlinchey, 1999; Jacobson & Truax, 1991; Martinovich, Saunders, & Howard, 1996). Hence, the effectiveness of therapeutic intervention can be evaluated on two dimensions. Pearson correlations were used throughout and significance was set at alpha = .05.

Separate analyses compared the control and treatment groups over time on each of the three measures. Analyses were also run using nature of torture, status of torture, importance of religion, and education, as covariates, but no effect was found for each of these, so the findings are reported for no covariate. Data were checked for distributional properties using the Shapiro-Wilk’s test before the main analyses. There were no significant departures from normality nor were there problems with variance heterogeneity. Two measures (P-scale and WHO-5) were unavailable for one person at follow-up. These points were substituted with pretest scores according to intent-to-treat analysis as described by Kazdin (2003).

Results

The relationship between the initial characteristics of the sample and our main measures is shown in Table 2. In particular, the measures correlated significantly with each other in the expected direction. The only significant relationship with key demographic characteristics was a significant negative correlation between importance of religion and the WHO-5 (both at pretest and follow-up) which indicated emotional well-being was less for those indicating higher importance of religion. Cross-tabulation analyses found that males were significantly more likely than females to have experienced physical torture, compared to psychological, or ill-treatment, χ² (2) = 7.0, p = .03, and to be primary survivors, compared to secondary, χ² (1) = 11.7, p = .001.

Table 2.

Correlations of pretest measures with sociodemographic characteristics.

	SLIPSS-A	P-scale	WHO-5
P-Scale pre	.62**
WHO-5 pre	−.68**	−.50**
Age	.29	.04	−.08
Gender^a	.30	.29	−.05
Status^b	.17	−.07	−.05
Religious importance	.31	.15	−.53**

Note. ^amale = 0, female = 1; ^bprimary = 1, secondary = 2

*p < .05. **p < .01.

Table 3 presents the means and standard deviations for the main analyses. Comparison of mean pretest scores for the two groups found no significant differences, for SLIPSS-A, t(24) = −1.08, p = .29; P-scale, t(24) = 0.52, p = .61; WHO-5, t(24) = −0.26, p = .80. There were also no age or gender differences between the two groups.

Table 3.

Comparison of therapy and control groups on outcome measures (N = 26)

	Control (n = 13)	Therapy (n = 13)
	Mean (SD)	Mean (SD)
SLIPSS-A
Pretest	29.23 (18.13)	38.15 (23.67)
Follow-up	29.77 (21.81)	25.54 (16.26)
Follow-up minus pretest	.54 (8.78)	−12.62 (15.26)
P-scale
Pretest	24.23 (18.82)	21.00 (12.34)
Follow-up	22.85 (21.04)	17.38 (15.55)
Follow-up minus pretest	−1.38 (13.12)	−3.62 (8.48)
WHO-5
Pretest	11.46 (4.67)	12.00 (5.87)
Follow-up	10.31 (6.02)	11.38 (6.97)
Follow-up minus pretest	−1.15 (5.11)	−.62 (6.69)

For SLIPSS-A, a significant interaction was found, F(1, 24) = 7.26, p = .01, effect = .60, indicating that survivors who received testimonial therapy improved over time relative to the control group. A significant main effect for time was also found, F(1, 24) = 6.12, p = .02, indicating a general improvement in psychosocial functioning over time regardless of group membership. The between group main effect was not significant, F(1, 24) = 0.10, p = .80.

For the P-scale, there was no significant interaction, F(1, 24) = 0.27, p = .61, effect = .14, nor were there significant main effects: time: F(1, 24) = 1.33, p = .26; group: F(1, 24) = 0.46, p = .50. Similarly for the WHO-5 measure, there was no significant interaction, F(1, 24) = 0.05, p = .82, effect = .09, nor main effects: time: F(1, 24) = 0.57, p = .46; group: F(1, 24) = 0.16, p = .69.

Reliable Change Index (RCI) and clinical significance

For the SLIPSS-A, the standard error of difference (SE_diff) was calculated from the combined pretest SD for the control and therapy groups (SD_pre = 21.16) and the measure’s internal reliability (alpha = .92; G. A. Fernando, 2008). The resulting SE_diff (8.35) was multiplied by 1.96 to obtain an RCI of 16.37 (Jacobson & Truax, 1991). Means and standard deviations for normative (M_nor = 43; SD_nor = 11.3) and traumatised populations (M_clin = 51.5; SD_clin = 17.6) were obtained from G. A. Fernando (2008) to calculate the cut-off point for clinical significance as 46.32. Accordingly, six participants receiving TT exceeded the RCI of 16.37 and were considered to show a statistically reliable increase in psychosocial functioning. Three of these participants also showed clinical improvement as the cut-off point of 46.32 was “crossed over” from pretest to follow-up and one participant showed clinical improvement but not statistical improvement, giving seven of 13 showing demonstrable improvement on the SLIPSS-A (and none showing decline).

Using similar calculations for the P-scale, SE_diff was calculated as 6.19 where SD_pre = 5.68 and Cronbach’s alpha = .92 (van Brakel et al., 2006). An RCI of 12.13 resulted in three survivors obtaining statistically reliable improvement in participation restriction. Clinical significance for the P-scale could not be calculated directly as data for normative and dysfunctional populations have not been reported. However, if the cut-off score of 13 between people with no significant participation restriction and mild participation restriction was taken (van Brakel et al., 2006), two of these three survivors also achieved clinically significant improvement giving 23% improvement (and no declines).

For the WHO-5, the RCI was 5.57 (SD_pre = 5.6) and the cut-off point was 13 (WHO, 1998), while Cronbach’s alpha was taken as .87 (Saipanish et al., 2009). Based on these parameters, four participants obtained both statistical and clinical significant improvement while four participants significantly deteriorated from pretest to follow-up.

The therapy group showed 14 improvements and four declines while the control group showed six improvements and six declines, however, this was not significant, χ² (1) = 2.5. If the number of cases in each group that showed improvement on at least one measure are considered (i.e., 77% of survivors treated with TT and 38% in the no therapy condition), a significant result is obtained, χ² (1) = 3.9.

Qualitative findings

The 13 testimonies delivered by the therapy group were recorded, transcribed, and analysed. Combined with information derived from the interviews, four themes were identified: coherence of story, expression of emotion, coping strategies, and future aspirations. Brief descriptions of the themes are as follows:

Coherence of the story

The majority of survivors provided coherent accounts of their experiences of torture and ill-treatment, possibly because they had told their stories many times during the legal process. Stories became more incoherent and fragmented when describing ancillary incidents of distress such as the death of a son or the suicide of a sibling. Two testimonies were taken over 2 days (Session 2) because of the difficulty encountered by these participants in giving testimony. An 18-year-old who was initially reluctant to talk about the severe torture endured said he was afraid that “talking about it might be mentally overwhelming” and impede his progress in life.

Expression of emotion

Although initially testimonies were devoid of emotional content, gradually survivors were guided to share their feelings and thoughts. The majority said they experienced dukha (“sadness”) when being tortured or ill-treated, or when they saw a loved one suffer. Most men expressed intense anger extending to thoughts of maragena merenna (“to kill others and kill oneself”) when being tortured, but also for the resulting injustice, humiliation, and loss of reputation. Many remained angry and revengeful for years after the event. Entering the legal process appeared to have moderated the intensity or immediacy of seeking revenge. As a 34-year-old fisherman explained, “I won’t do anything to endanger my case, but if I don’t get justice through the courts, I will seek it myself.” Other salient emotions expressed by both genders included fear of reprisals, disillusionment with the legal system, disgust, shame, and regret as well as a colloquial expression of feeling, manasika weteema (“to mentally fall”).

Coping strategies

Testimonies contained various strategies used by participants to cope with their trauma. For example almost all survivors were grateful for the legal assistance and support afforded by Jana Sansadaya to end the torture, to obtain bail, to complain, and to pursue their legal battles. The support they received from staff members, who had stood by them in their desperation and had given them hope, helped them to carry on living despite the traumatic incidents. Some found strength in the support of parents, spouses, or children, or in religious practices and precepts. One 43-year-old school teacher said that the adverse incident itself (an assault on her sick child) and its consequences had given her “manasika shaktiya” (mental strength) she did not know she had before the incident.

Future aspirations

While seeking justice through the legal system was the salient future goal for many survivors, their conceptualisation of justice varied. For some, justice was limited to receiving monetary compensation for injuries suffered and livelihoods lost. For others, monetary redress was inadequate or irrelevant. Justice to them meant punishing the perpetrators and/or being proclaimed innocent of crimes they had not committed so they could restore their reputation. Other future ambitions included succeeding in business, building a home, and educating their children.

Ceremonies, “the turning point.”

All survivors who were interviewed by HR workers after the follow-up assessment said that they felt positively towards the testimonial ceremony, a novel event in their lives. Their stories had been told before—to lawyers, courts, and tribunals. However, at the ceremony, for the first time, their trauma story was read out in public and they were honoured for enduring their suffering. Some said the experience strengthened and motivated them to continue struggling for justice. Others said they kept their testimony document safely at home and shared it with friends and neighbours with pride. They used the testimony to motivate people in their community to seek justice as opposed to being silent victims of injustice. Conversely, one 55-year-old man said that, despite it all, it was hard to forget that his problems still continued, yet the ceremony had given him a glimmer of hope to see justice one day.

Discussion

The present study aimed to evaluate the efficacy of testimonial therapy for Sri Lankan torture survivors using a randomised controlled study. Taken together, the findings suggest TT may be an effective therapy for survivors of torture and ill-treatment in Sri Lanka. In total, 77% of survivors treated with TT obtained statistical or clinical improvement on one or more outcome measures compared to 38% in the no-therapy condition. While this finding is generally consistent with the conclusion of Agger et al. (2012; Agger et al., 2009), our use of a no-therapy control, statistical analysis, and individual outcome methodology lends additional weight to the efficacy of TT. Follow-up at 2-3 months found the most significant improvements occurred in the general psychosocial functioning of participants as assessed with a measure specifically designed to assess the effects of trauma in the Sri Lankan rural population (G. A. Fernando, 2008). The beneficial effect of TT was not found in the group means for social participation or emotional well-being despite significant initial correlations amongst the three measures.

Despite the general support for TT as a viable therapy for torture survivors seeking justice, inconsistencies in therapeutic change need closer examination. While four participants receiving treatment showed declines in emotional well-being as measured by the WHO-5, all of them showed improvements on at least one of the other measures. However, the outcome measures had psychometric limitations. Even though the WHO-5 has been translated and used in many other countries, there were problems finding equivalent terms in Sinhalese for concepts such as “cheerful,” “rested,” and “good spirits.” In particular, the item “How often have you felt … in good spirits” has different connotations in a religious rural Sri Lankan population.

An interesting aspect of the use of the WHO-5 was the significant negative relationship with religious importance indicating that those highest in the importance of religion reported the least emotional well-being as measured by the WHO-5. G. A. Fernando (2008) found that during the development of the SLIPSS-A, many Sri Lankans showed high life satisfaction scores despite reporting suffering from trauma. She argued that this might relate to “resilience” in Sri Lankan Buddhism and proposed that “life satisfaction” might be more related to religious beliefs and practices and less to one’s misfortunes. Along with the psychometric issues, this suggests WHO-5 may not be particularly useful as a measure of general emotional well-being in a Sri Lankan population.

Another limitation of the study was the likelihood of contamination of the control group through informal interaction and communication among treatment group members. Igreja et al. (2004) reported on the effectiveness of a single session TT intervention with Mozambican war survivors and found that the control group improved to a similar degree as the intervention group. This was attributed to uncontrolled interaction and communication among community members. In the present study, participants assigned to the wait-list group lived in the same community, participated in HR activities, and occasionally attended the ceremonies along with the participants in the intervention group.

Another challenge in assessing the impact of the intervention was the impact of intervening events or daily stressors after the ceremony and before the follow-up assessment. Miller and Rasmussen (2010) and G. A. Fernando et al. (2010), among others, have found that daily stressors play a significant part in mediating the distressing effects of trauma and violence. Consequently, the therapeutic effects of therapy could be enhanced or diminished by events that occur prior to follow-up assessment (Kazdin, 2003). In the present study, this was shown in follow-up interviews with participants for whom losing a court case, police arrest of a family member, or physical illness had occurred. Of course, control group participants would be expected to be exposed to similar events but the small sample size may not have allowed adequate control. Hopefully, future studies can build on our experience with larger samples, more robust measures, and longer term follow-up.

Benefits of testimony therapy in Sri Lanka

TT can be easily incorporated into existing HR work to enhance the therapeutic value of services provided to torture survivors and their families. By providing a copy of the testimony document to lawyers, case workers, and others, TT minimises the need for repetition of survivors’ trauma stories, potentially enhancing the therapeutic value of the legal process. However, care must be taken to ensure there are no significant discrepancies between survivors’ emotional testimonies and their legal evidence which could be used in a court of law to discredit the survivor’s story.

An objective of the local HR movement is to transform silent victims of rights violations into resilient survivors who collectively demand justice. The testimonial ceremony allows for this transformation when survivors’ private and subjective pain is reframed into the public, legal, and political domain (Agger et al., 2009; Perera et al., 2009). Being publicly honoured also restores self-esteem and mitigates social stigma by promoting social acceptance and understanding regarding the torture survivor’s plight. It can bolster survivors’ commitment and resolve to continue seeking justice, and this was a salient theme in the qualitative analyses. Even when survivors fail to obtain the justice they are seeking, TT may offer them a socially constructive way to channel their intense anger and frustration, which could otherwise spiral into revenge-seeking, domestic violence, or suicidal behaviour (Agger & Jensen, 1996; Cienfugos & Monelli, 1983).

There are resource limitations and numerous barriers to mental healthcare access in low-income countries such as Sri Lanka (Patel et al., 2007; Saraceno et al., 2007). Survivors of human rights violations may be reluctant to avail themselves of existing services due to the stigma of mental illness as well as fear and suspicion. Our version of TT seems especially useful in this context because it is designed for use by trained nonspecialists who have already gained the trust and confidence of their clientele. The entry point through the justice process prevents any stigmatising label of mental illness (Clancy & Hamber, 2008).

Conclusion

The present study evaluated the efficacy of TT for survivors of torture and ill-treatment in Sri Lanka, using a randomised controlled trial. Overall, findings suggest that TT can be a useful therapy for survivors of torture and ill-treatment in Sri Lanka. Small sample size, possible control group contamination, and unavailability of optimum measures with known psychometric properties, impeded the study. Despite these limitations, TT appears to be a viable therapy for torture survivors motivated to give testimony and committed to seeking justice. With further adaptation and training, TT has the potential to benefit not only torture survivors, but other trauma survivors in Sri Lanka and elsewhere.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Acknowledgements

We wish to thank the survivors of torture and ill-treatment who participated in the study, as well as the staff of Jana Sansadaya, especially Amitha Priyanthi and Harshi C. Perera, for their invaluable contribution to this project.

Teresa Puvimanasinghe, PhD, BA (Hons), Attorney at Law, is a research associate at the University of South Australia and counsellor to survivors of torture and trauma at STTARS, Adelaide, South Australia. Her research interests include: meaning making, coping and thriving after severe trauma, refugee and migrant mental health, and acculturation. Previously she worked as a human rights lawyer and advocate in Sri Lanka and Hong Kong. She is the editor/co-editor of books such as: An Exceptional Collapse of the Rule of Law: Told Through Stories by Families of the Disappeared in Sri Lanka and An X-Ray of the Sri Lankan Policing System and Torture of the Poor.

Ian Price, PhD, is a senior lecturer at the University of New England, Armidale, Australia. His teaching and research interests centre on psychological testing, addictive behaviour, and mental health. He is a registered psychologist with other research interests in neuroscience, animal behaviour, and statistics.

References

Agger

Igreja

Kiehle

Polatin

(2012) Testimony therapies in Asia: Integrating spirituality in testimonial therapy for torture survivors in India, Sri Lanka, Cambodia, and the Philippines. Transcultural Psychiatry 49: 568–589.

Agger

Jensen

S. B.

(1990) Testimony as ritual and evidence in psychotherapy for political refugees. Journal of Traumatic Stress 3(1): 115–130.

Agger

Jensen

S. B.

(1996) Trauma and healing under state terrorism, London, UK: Zed Books.

Agger

Raghuvanshi

Khan

S. S.

Polatin

Laursen

L. K.

(2009) Testimonial therapy: A pilot project to improve psychological wellbeing among survivors of torture in India. Torture 19(3): 204–217.

Allison

P. D.

(1990) Change scores as dependent variables in regression analysis. Sociological Methodology 20: 93–114.

Atkins

D. C.

Bedics

J. D.

McGlinchey

J. B.

Beauchaine

T. P.

(2005) Assessing clinical significance: Does it matter which method to use? Journal of Consulting and Clinical Psychology 73: 982–989.

Awata

Bech

Yoshida

Masashi

Susumu

Motoyasu

Oka

(2007) Reliability and validity of the Japanese version of the World Health Organization-Five Well-Being Index in the context of detecting depression in diabetic patients. Psychiatry and Clinical Neurosciences 61: 112–119.

Bichescu

Neuner

Schauer

Elbert

(2007) Narrative exposure therapy for political imprisonment-related chronic posttraumatic stress disorder and depression. Behaviour Research and Therapy 45: 2212–2220.

Bracken

P. J.

Giller

J. E.

Summerfield

(1995) Psychological responses to war and atrocity: The limitations of current concepts. Social Science & Medicine 40: 1073–1082.

10.

Braun

Clarke

(2006) Using thematic analysis in psychology. Qualitative Research in Psychology 3: 77–101.

11.

Cienfuegos

A. J.

Monelli

(1983) The testimony of political repression as a therapeutic instrument. American Journal of Orthopsychiatry 53(1): 43–51.

12.

Clancy

A. C.

Hamber

(2008, September) Trauma, peace-building, and development: An overview of key positions and critical questions. Paper presented at the conference of the International Conflict Research Institute (INCORE) and International Development Research Centre (IDRC), New Delhi, India.

13.

Deane

F. P.

Kelly

P. J.

Crowe

T. P.

Coulson

J. C.

Lyons

G. C. B.

(2013) Clinical and reliable change in an Australian residential substance use program using the Addiction Severity Index. Journal of Addictive Diseases 32: 194–205.

14.

Fernando

(2010) Sri Lanka: Impunity, criminal justice and human rights, Hong Kong, China: Asian Human Rights Commission.

15.

Fernando

Puvimanasinghe

(2004) An X-ray of the Sri Lankan policing system and torture of the poor, Hong Kong, China: Asian Human Rights Commission.

16.

Fernando

G. A.

(2008) Assessing mental health and psychosocial status in communities exposed to traumatic events: Sri Lanka as an example. American Journal of Orthopsychiatry 78: 229–239.

17.

Fernando

G. A.

Miller

Berger

(2010) Growing pains: The impact of disaster related and daily stressor on the psychological and psychosocial functioning of youth in Sri Lanka. Child Development 81: 1192–1210.

18.

Frattaroli

(2006) Experimental disclosure and its moderators: A meta-analysis. Psychological Bulletin 132: 823–865.

19.

Ginzburg

Neria

(2011) Mental health interventions for survivors of torture: Current status and future directions. Journal of Psychology 219: 187–189.

20.

Henkel

Mergl

Kohnen

Maier

Moller

H. J.

Hegerl

(2003) Identifying depression in primary care: A comparison of different methods in a prospective cohort study. British Medical Journal 326(7382): 200–201.

21.

Igreja

Kleijn

W. C.

Schreuder

B. J. N.

van Dijk

J. A.

Verschuur

(2004) Testimony method to ameliorate post-traumatic stress symptoms. British Journal of Psychiatry 184: 251–257.

22.

Isakson

B. L.

Jurkovic

G. J.

(2013) Healing after torture: The role of moving on. Qualitative Health Research 23: 749–761.

23.

Jacobson

N. S.

Roberts

L. J.

Berns

S. B.

McGlinchey

J. B.

(1999) Methods for defining and determining the clinical significance of treatment effects: Description, application, and alternatives. Journal of Consulting and Clinical Psychology 67: 300–307.

24.

Jacobson

N. S.

Truax

(1991) Clinical significance: A statistical approach to defining meaningful change in psychotherapy research. Journal of Consulting and Clinical Psychology 59: 12–19.

25.

Kazdin

A. E.

(2003) Research design in clinical psychology, (4th ed. Boston, MA: Allyn & Bacon.

26.

Lang

P. J.

(1979) A bio-informational theory of emotional imagery. Psychophysiology 16: 195–512.

27.

Malalasekera

G. P.

(1997) English–Sinhalese dictionary, (3rd ed. Colombo, Sri Lanka: M. D. Gunasena & Co.

28.

Martinovich

Saunders

Howard

K. I.

(1996) Some comments on “Assessing clinical significance.”. Psychotherapy Research 6: 124–132.

29.

McFarlane

C. A.

Kaplan

(2012) Evidence-based psychological interventions for adult survivors of torture and trauma: A 30-year review. Transcultural Psychiatry 49: 539–567.

30.

Miller

K. E.

Rasmussen

(2010) War exposure, daily stressor, and mental health in conflict and post-conflict settings: Bridging the divide between trauma-focused and psychosocial frameworks. Social Science & Medicine 70: 7–16.

31.

Momtaz

Y. A.

Hamid

T. A.

Yahaya

Ibrahim

(2010) Effects of chronic comorbidity on psychological well-being among older person in northern peninsular Malaysia. Applied Research Quality Life 5: 133–146.

32.

Morris

S. B.

(2008) Estimating effect sizes from pretest–posttest–control group designs. Organizational Research Methods 11: 364–386.

33.

Neuner

Schauer

Klaschik

Karunakara

Elbert

(2004) A comparison of narrative exposure therapy, supportive counselling, and psychoeducation for treating posttraumatic stress disorder in an African refugee settlement. Journal of Consulting and Clinical Psychology 72(4): 579–587.

34.

Patel

Araya

Chatterjee

Chisholm

Cohen

De Silva

van Ommeren

(2007) Treatment and prevention of mental disorders in low-income and middle-income countries. The Lancet 370: 991–1005.

35.

Pennebaker

J. W.

(1993) Putting stress into words: Health, linguistic, and therapeutic implications. Behavior, Research & Therapy 31: 539–548.

36.

Perera

Puvimanasinghe

Agger

(2009) Giving voice to the voiceless: A manual for community workers and human rights activists in Sri Lanka, Copenhagen, Denmark: Rehabilitation and Research Centre for Torture Victims.

37.

Saipanish

Lotrakul

Sumrithe

(2009) Reliability and validity of the Thai version of the WHO-Five Well-Being Index in primary care patients. Psychiatry and Clinical Neurosciences 63: 141–146.

38.

Saraceno

van Ommeren

Batniji

Cohen

Gureje

Mahoney

Underhill

(2007) Barriers to improvement of mental health services in low-income and middle-income countries. The Lancet 370: 1164–1174.

39.

Schauer

Neuner

Elbert

(2005) Narrative exposure therapy: A short-term intervention for traumatic stress disorders after war, terror, or torture, Cambridge, MA: Hogrefe & Huber.

40.

Strang

A. B.

Ager

(2003) Psychosocial interventions: Some key issues facing practitioners. Intervention 1(3): 2–12.

41.

Van Brakel

W. H.

Anderson

A. M.

Mutatkar

R. K.

Bakirtzief

Nicholls

P. G.

Raju

M. S.

Das-Pattanayak

R. K.

(2006) The Participation Scale: Measuring a key concept in public health. Disability and Rehabilitation 28(4): 193–203.

42.

Van der Kolk

Fisler

(1995) Dissociation and the fragmentary nature of traumatic memories: Overview and exploratory study. British Journal of Psychotherapy 12(3): 352–361.

43.

Van der Kolk

Hopper

J. W.

Osterman

J. E.

(2001) Exploring the nature of traumatic memories: Overview and exploratory study. British Journal of Psychotherapy 12: 352–361.

44.

Van Dijk

J. A.

Schoutrop

M. J. A.

Spinhoven

(2003) Testimony therapy: Treatment method for traumatized victims of organized violence. American Journal of Psychotherapy 57(3): 361–373.

45.

Wang

S.-J.

Haque

M. A.

Masum

S.-u. D.

Biswas

Modvig

(2009) Household exposure to violence and human rights violations in western Bangladesh (II): History of torture and other traumatic experience of violence and functional assessment of victims. BMC International Health and Human Rights 9: 31. Retrieved from http://www.biomedcentral.com/1472-698X/9/31.

46.

Weathers

F. W.

Litz

B. T.

Huska

J. A.

Keane

T. M.

(1994) PCL-C for DSM-IV, Boston, MA: National Center for PTSD, Behavioral Science Division.

47.

Weine

S. M.

Kulenovic

A. D.

Pavkovic

Gibbons

(1998) Testimony psychotherapy in Bosnian refugees: A pilot study. American Journal of Psychiatry 155: 1720–1726.

48.

World Health Organization: Regional Office for Europe (1998) Well-being measures in primary health care: The DepCare Project. (Report on a WHO meeting, Stockholm, Sweden, 12–13 February 1998). Copenhagen, Denmark: World Health Organization, Regional Office for Europe.