Counselor and Participant Perspectives of Trauma-Focused Cognitive Behavioral Therapy for Children in Zambia: A Qualitative Study

Abstract

Objective

This study examined Zambian counselors, children, and caregivers' perceptions of an evidence-based treatment (EBT) for trauma (Trauma-Focused Cognitive Behavioral Therapy, TF-CBT) utilized in Zambia to address mental health problems in children.

Method

Semi-structured interviews were conducted with local counselors trained in TF-CBT (N=19; 90% of those trained; 12 Female) and children/caregivers who had received TF-CBT in a small feasibility study (N=18; 86% of the children and N=16; 76% of the caregivers) who completed TF-CBT (Total completed; N=21). Each client was asked six open-ended questions, and domain analysis was used to explore the data.

Results

Counselors were positive about the program, liked the structure and flexibility, reported positive changes in their clients, and discussed the cultural adaptation around activities and language. Counselors stated the training was too short, and the supervision was necessary. Challenges included client engagement and attendance, availability of location, funding, and a lack of community understanding of “therapy.” Children and caregivers stated multiple positive changes they attributed to TF-CBT, such as better family communication, reduction of problem behaviors, and ability to speak about the trauma. They recommended continuing the program.

Conclusion

This study brings a critical examination of providers' and clients' perspectives of the implementation of an EBT for children in a low-resource setting. Clinical implications include changing implementation methods based on responses. Research implications include future study directions such as an effectiveness trial of TF-CBT and an examination of implementation factors.

Children in low and middle-income countries (LMIC) often experience multiple traumatic events as a result of extreme poverty, child labor, political violence, internal displacement, and human trafficking (Masinda & Muhesi, 2004; UNICEF, 2009; Whetten, Ostemann, Whetten, O'Donnell & Theilman, 2011). In Zambia, there are an estimated 993,000 orphaned children, many of whom have multiple stressors such as nursing dying parents, poverty, being forced to move, separated from siblings, withdrawn from school, and exposed to abuse (UNICEF, 2008). A qualitative study in Zambia supported that the local community also perceived abuse, neglect, and violence as significant problems (Murray et al., 2006). Among children, exposure to trauma can have adverse short and long term consequences and, without treatment, can seriously impact affective, cognitive, behavioral, physical, and social functioning (Chartier, Walker & Naimark, 2007; Chen et al., 2010; Kilpatrick et al., 2003; Masinda & Muhesi, 2004;). The magnitude of traumatic events and the ensuing mental health problems are especially problematic in LMIC, where quality and effective mental health services are scarce (Jordans et al., 2010; Saxena, Thornicroft, Knapp & Whiteford, 2007).

In the field of global mental health, there has been increasing support for the use of evidence-based treatments (EBT) in LMIC. In a systematic review of research conducted in LMIC, Schunemann and colleagues advocated for the use of EBTs to apply more effective treatments globally (Schunemann, Fretheim & Oxman, 2006). Other researchers assert that with limited resources it is essential in LMIC to focus on evidence-based interventions as they have proven to be effective in other contexts and offer the best possibility of impact (Pearson & Jordan, 2010; Patel et al., 2007). The World Health Organization (WHO) recently encouraged the use of EBT in their Mental Health Gap Action Programme (mhGAP), which aims at scaling up services for mental, neurological, and substance use disorders for countries, especially LMIC (WHO, 2010). The growing scientific literature also demonstrates that EBTs can be effectively and feasibly implemented in LMIC (Bolton et al., 2007; Bolton et al., 2003; Bolton, Bass & Murray, 2011; Patel, Chowdhary, Rahman &Verdeli, 2011; Rahman, Malik, Sikander, Roberts & Creed, 2008). Finally, there is increasing literature suggesting that EBTs can be culturally adapted and that the core principles of EBTs remain intact (Kaysen et al., 2013; Patel et al., 2011; Verdeli et al., 2003; Verdeli et al., 2008).

Still, there is caution around the use and implementation of EBTs cross-culturally. In their review of the use of EBTs with survivors of child abuse and neglect, Chaffin and Friedrich (2004) argue that some EBTs may be inflexible and restrictive of therapist creativity. Research has also suggested that EBTs developed in high-income countries (HIC) are often developed for the general population and then applied to specific cultures without consideration of differences in culture and worldview (Bigfoot & Schmidt, 2010; Gone, 2009). There is general agreement that collaboration between community members and researchers is essential for EBTs to be culturally appropriate while maintaining the core components of the treatment (Bernal & Rodriguez, 2009; Bos, Schaalma, & Pryor, 2008; Gone, 2009; Pearson & Jordan, 2010; Wessells, 2009).

As a debate ensues on using EBTs in LMIC, the global literature lacks an understanding of local provider and client perspectives – a topic usually examined under dissemination and implementation (D&I) science literature. The D&I science of EBTs for children and youth mental health has received increased attention in the United States, yet a significant gap in our understanding of this area remains (Aarons, 2005; Hoagwood, Burns, Kiser, Ringeisen & Schoenwald, 2001). Implementing an EBT in “real-world” settings is a challenging process – and arguably even more so in LMIC where the health systems infrastructures are immature. Researchers emphasize that D&I research is critical for improving the quality of treatment services as well as their sustainability (Burns, 2003; Hoagwood et al., 2001) and urge the examination of constructs such as acceptability, feasibility, fidelity, and uptake (Proctor & Rosen, 2008; Proctor et al., 2009). Implementation of an EBT in LMIC with lay counselors includes new processes such as different training and supervision, fidelity monitoring, and protocol adherence (Elliot & Mihalic, 2004). Receiving an EBT is likely an entirely new experience for children and families in most LMIC. Understanding local perspectives could help address barriers to implementation, guide future programmatic endeavors, and advance sustainability efforts (Glenton, Lewin & Scheel, 2011). To our knowledge, no studies exist that have investigated the frontline provider perspectives on using an EBT in a LMIC, or the local perspective of clients in LMIC who have received an EBT.

The literature encourages the use of mixed-methods, particularly when examining implementation outcomes (Palinkas et al., 2011; Robins et al., 2008), which was the overall study approach. This manuscript focuses on the qualitative methods and results. Qualitative methods are particularly helpful to explore and obtain depth of understanding as to the reasons for success or failure in the implementation of an EBT and to identify strategies for facilitating implementation (Palinkas, et al, 2011; Teddlie & Tashakkori 2003). We were interested in exploring whether an EBT for childhood trauma was feasible and acceptable in Lusaka, Zambia.

The objective of this qualitative study was to examine Zambian counselors, children, and caregivers' perceptions of an EBT (Trauma-Focused Cognitive Behavioral Therapy - TF-CBT) utilized in Zambia to address mental health problems in children who have experienced trauma. Semi-structured interviews were conducted with local counselors trained in TF-CBT and children/caregivers who had received TF-CBT. The foci of the open-ended questions were acceptability, barriers, facilitators, the impact of the TF-CBT program, and recommendations for the future.

Background Feasibility Study

This qualitative study on client and counselor perspectives was part of a larger feasibility study. Following the Design, Implementation, Monitoring, and Evaluation (DIME) approach (JHU, 2012), a community based qualitative study was first completed to gather the local perspectives of the problems of women and children affected by HIV in this region (Murray et al., 2006). The results of this study suggested a high incidence of child trauma including sexual abuse, physical abuse, domestic violence, and related trauma symptomatology (e.g., “crying,” “thinking too much,” “alone and withdrawn,” “fearful that it will happen again,”), as well as grief. Based on these results, mental health assessment tools for youth were adapted and validated. Results showed that the adapted measure for trauma symptomatology, the Post-traumatic Stress Disorder Reaction Index (PTSD-RI; Frederick, Pynoos & Nader, 1992), had good local reliability and validity (Murray et al., 2011).

Trauma-Focused Cognitive Behavioral Therapy (TF-CBT) was identified for implementation through a community participatory research process (Murray et al., submitted for publication). Briefly, community participatory research was utilized and followed a process to: 1) examine the results from the qualitative study, 2) document existing services that were available, 3) identify services that seemed to be missing, and 4) review the literature for different treatment options. Some of the reasons TF-CBT was chosen to trial in Lusaka by local stakeholders included: 1) its ability to address children across the developmental spectrum, aged preschool to 18 years, 2) the extensive evidence on its effectively reducing a wide range of symptoms, and 3) the inclusion of traumatic grief components within the model given the high incidence of death related to HIV.

TF-CBT includes 9 individual components for children ages 4-18 and for their parents or caregivers/family systems (if available). The intervention consists of 12-16 1-hour sessions, although is flexible and individualized to meet the needs of each child/youth and family. TF-CBT (Cohen, Mannarino & Deblinger, 2006; www.musc.edu/tfcbt) has several randomized controlled trials supporting its efficacy in treating many outcomes, including child behavior problems, safety skills, social competence, depression, PTSD, traumatic grief, child shame, and parental reactions to child trauma (Cohen & Mannarino, 2004; Cohen, Berliner & Mannarino, 2010; Cohen, Deblinger, Mannarino & Steer, 2004; Cohen et al., 2006; King et al., 2000). Studies have also shown that TF-CBT is adaptable across different cultures (BigFoot & Schmidt, 2010; Huey & Polo, 2008). The literature on cross-cultural adaptation of TF-CBT (e.g., BigFoot & Schmidt, 2010), as well as the wider literature on EBT (Patel et al., 2011; Verdeli et al., 2003; Verdeli et al., 2008) suggest that modifications are primarily on methods of implementation (rather than core concepts), simplifying text, and reducing jargon. One component of TF-CBT that was hypothesized to require some adaptations in LMIC is parenting skills. This was based on previous qualitative studies and the local Zambian authors' knowledge of the culture, suggesting that parenting practices are often different from high-income countries. For example, children as young as age six in Zambia are often significant caregivers for their younger siblings, carrying them throughout the day on their backs.

For the feasibility study, 22 Zambians (13 females and 9 males) were trained by the lead author. Given the limited number of mental health professionals, the task-shifting model was used (Verdeli et al., 2003; WHO, 2010), which utilizes lay counselors without formal mental health training. Counselors' backgrounds included students from local schools, staff of local and international non-governmental organizations (NGOs), local “counselors,” and workers from the University Teaching Hospital and the local psychiatric hospital. Experience and educational background in teaching or psychology varied within the group; however, only three had any formal clinical training. A phased, apprenticeship model of training and ongoing supervision (Murray et al., 2011) was used, which included two five-day live trainings, local practice groups for role-playing with about six counselors per group, and close supervision for cases. Three local supervisors were chosen from within the group based on strong skill uptake, interest, time, and leadership and/or teaching skill. All were female; 1 was a lecturer at the University of Zambia, 1 was a counselor at a local centre, and 1 worked at a NGO doing non-mental health work. Supervision was conducted by local supervisors for at least two hours per week in small groups; in addition, local supervisors spoke to a TF-CBT trainer weekly (usually by phone or Skype). Additional support from the TF-CBT trainer was provided via email as needed. Nineteen counselors agreed to take a pilot case: three counselors did not continue due to demands at their current positions.

Children and caregivers were referred to the feasibility study from a center that serves youth who have experienced sexual violence - the “One-Stop Centre” (Chomba et al., 2010). Existing services at the Centre included medical exams, HIV testing, post-exposure prophylaxis, and legal interviews. Inclusion criteria for the clients was a score of 39 or higher on the modified PTSD-RI and indication of a traumatic event. Forty cases started TF-CBT, and 15 of these were lost or did not return within the first 3 sessions. Of the 25 that started the Trauma Narrative component (approximately session 6-7), 4 did not finish the treatment. The sample (N=21) was comprised of girls (100%) with a mean age of 12.76 (SD= 1.75). The sample endorsed being exposed to a number of traumatic events, and 100% had been sexually abused. Paired sample t-tests were conducted to examine the mean differences of trauma symptoms at baseline and follow-up (N=18). Results revealed a significant difference in the means between baseline, (M= 72.17, SD= 37.79) and follow up (M=50.50, SD= 43.47) PTSD-RI scores, t₍₁₇₎= 2.23, p<.05 (CI 1.21, 42.13).

The Zambian counselors, children, and caregivers who participated in this feasibility study were the respondents of the qualitative study. This manuscript describes examining the perceptions of TF-CBT utilized in Zambia to address mental health problems in children who have experienced trauma.

Method

Qualitative Study

Site and clients

All TF-CBT counselors participating in pilot cases (N=19), along with three graduate students from Boston University (BU), were trained by study authors (LKM and SS) in qualitative interviewing methods and research ethics. The 3 students were obtaining their Masters in Public Health and were participating in an internship based in Zambia for the summer. TF-CBT counselors conducted interviews for the clients (children and caregivers; separately). Interviews were assigned so that the interviewer was not the counselor who provided TF-CBT to the client, and was matched to assure they spoke the same language (e.g., Nyanja, Bemba or Tonga). The BU graduate students conducted the interviews with the local TF-CBT counselors working in pairs, with one being the lead interviewer and the other being the primary recorder. After each interview, the students compared and consolidated their notes.

Interviewers asked a series of six open-ended questions: 1) Tell me about your experience with TF-CBT; 2) Tell me about the challenges of the TF-CBT program; 3) What did you like about the program?; 4) What did you dislike about the program?; 5) Describe any changes in the clients/family/child/self since starting the treatment; and 6) Tell me about any recommendations for the program. Interviewers used open ended, non-leading probes such as “tell me more about that,” or “explain/describe that” to elicit additional information about responses. The questions were developed to be very simple and broad to promote comprehension and allow translation into the three local tribal languages.

All children/caregivers who completed TF-CBT (N=21) were approached, as well as all counselors who were initially trained in TF-CBT (N=22). Initial qualitative interviews lasted approximately one hour with each respondent. Written translated interviews were submitted to the study team for review. Subsequently, a second interview took place for further clarification and probing on the respondents' initial responses. In addition to the written notes, all interviews were tape recorded to aid in reviewing notes and translating the interview. If there was a question on the written notes, the tapes were used to clarify.

Analysis

Analysts consisted of two faculty with extensive experience in Zambia, two graduate-level students, and two local Zambians. The Zambian analysts worked in both the local languages and English, and at least one of them reviewed all question responses for cultural and semantic interpretation. All analysts were trained by the lead author. U.S.-based coders were already familiar with qualitative analysis from graduate level courses.

Our data analytic method followed a grounded theory approach (Glaser & Strauss, 1967; Strauss & Corbin, 1998). First, analysts reviewed all data to develop a broad understanding of content as it related to the project's specific aims. Second, analysts wrote down everything that was relevant to the implementation of TF-CBT and which interviewee said it, noted by the ID number. These segments of text ranged from a phrase to several paragraphs - as much as necessary to explain a concept. Where two or more interviewees made the same statement using the same language, the statement was not recorded twice, but rather both IDs were recorded next to the statement. Once this was complete, the analysts reviewed the lists for each question, looking for statements that may have the same meaning even if the wording was different. Analysts discussed which wording was most likely to be clearest to most of the study population. The other wordings with the same meaning were then removed from the list and their IDs were listed next to the selected wording. This step is particularly important when working in different cultures with tribal languages and where we rely heavily on our local analysts. Finally, analysts organized these lists into: a) “cover terms” (e.g., “challenges”) based on a priori themes from the interview, or emergent themes, and b) “included terms,” or ideas that fall under the cover term.

Analysts recorded data by the ID of the informant so upon completion researchers knew how many and which respondents spoke about a particular term. All of the interviews were reviewed multiple times for a general understanding of their content, which provided greater familiarity with the data by multiple people. Where differences were found between analysts, the team conducting the analyses reviewed the full original responses from informants (on paper and/or via audio recording), and made a determination through discussion and consensus (Hill et al., 2005).

We chose not to use coding or analytic software in this study. The analysis process was quite simple and consisted primarily of summarization. In addition, although we were not able to do analysis with a large group of local speakers as preferred, we did have a few Zambian analysts. Since we were working between three tribal languages and English, we wanted to reduce possible error from translations. We prefer that analytic decisions are made based on the original language and by representatives of the local community, reflecting their vocabulary and word use.

Results

Zambian Counselors' Responses

Nineteen counselors were interviewed (12 females); all but one did two separate interviews of approximately one hour each. Three were out of the country at the time of interviews. Counselor responses across all questions fell into the broad categories of a) TF-CBT is a good program b) Challenges of TF-CBT, c) Perceived changes in children and families, d) Cultural adaptations, e) Training and supervision, and f) Suggestions for improvement. The most frequently mentioned ‘included terms’ (four or more counselors stated similar responses) are presented in Table 1.

Table 1. Zambian Counselor Perspectives (N=19).

COVER TERM	Included term summary	Number reporting
TF-CBT is a good program
	The skills counselors learn from TF-CBT are useful in their own life	9
	Structure of the program was useful to the counselor; easy to follow.	7
	The benefits of the program extend beyond patients – to the parents and counselors themselves	7
	TF-CBT builds the relationship between caregiver and child	6
	Involving the caregiver as the support system for the child is an important strength of the model	6
	TF-CBT is empowering	5
	TF-CBT is flexible - you can adapt it to the client	5
	Benefits for the clinicians to be involved in this	5
	The program is practical – both the exercises and the skills taught	4
	People appreciate the program	4
	Helps children get over/come to terms with thoughts and feelings	4
	Good program for sexual abuse	4
Challenges
	Poor attendance at and/or commitment to therapy sessions	15
	TF-CBT duration is too long	9
	Challenges for first-time counselors	6
	Community misconceptions and/or lack of awareness of TF-CBT	5
	Challenging to explain TF-CBT in the local languages	4
	Difficult to talk about sex in our culture	4
Perceived changes in children and families
	The child is more open (to talk about problems/trauma with the caregiver and others)	17
	Overall positive change/growth (attitudes, more trust, sleeps better, learns about feelings and thoughts, goes to school, good hygiene, self-esteem, self-confidence, less angry, more relaxed)	14
	Improved child/caregiver relationship.	11
	Increased support from caregivers	7
	Development of positive/helpful thoughts	7
	Caregiver learns parenting skills	5
	Child socializes more/better	5
	TF-CBT causes change and growth in the child and caregiver	5
	Clients make their own decisions	4
Cultural Adaptations
	There is a need to adjust the activities from the TF-CBT manual to fit the local context	10
	Important to address issues/difficulties regarding use of local languages	9
	It is taboo to talk about sex	8
	Parenting skills were challenging for parents	6
	TF-CBT is working well in Zambia culture	4
	There are differences in parent/child relationship here	4
Training & Supervision
	The practice and supervision groups were helpful and motivating	15
	Training was good – organized, comprehensive, experiential, fun, etc.	9
	Training should be longer	4
Suggestions for Improvement
	Create a physical, permanent center for therapy	9
	Adjust the training in TF-CBT (lengthen, update with materials)	7
	Create awareness of sexual abuse and this treatment in the community	6
	Reduce length of TF-CBT treatment for clients	6
	Improve patient referral system	5
	Improve patient attendance	4
	Provide more funding	4
	Scale up program	4
	Need for programs like this in Zambia	4

TF-CBT is a good program

All counselors reported positive perspectives about the TF-CBT model, each focusing on different areas such as liking the structure of the model, as well as the flexibility. For example:

I like its simplicity, its structure, its applicability, user friendly, structure; I like structure so for me this works just fine. It's not something you are just doing without knowing what the benefits are. When you use the model you can almost immediately see how it works. You get results; you get feedback so I like that.

They also reported that TF-CBT skills were useful not only for the clients, but also for themselves. For example, counselors talked about using the cognitive coping component often themselves to deal with daily stressors. As one counselor put it, “It is useful in that we are helping the children… it's also useful as person I'm learning a lot…” Counselors also discussed how the model seemed to empower caregivers and children, as well as help build relationships. For example, counselors appreciated TF-CBT, as the inclusion of caregivers and family empowered the latter to help their children. Additionally, counselors saw this connection as improving relationships between family members. One counselor said, “When you pair with the patient and caregiver, it becomes like you are creating that relationship between them so they have a good social support system…”

Challenhges of the TF-CBT program

The majority of counselors reported difficulty in attendance and commitment from the families they were working with.

I think it's difficult when you're given a case and the case just falls through, and I think it's even more difficult when you know that the client is really into a tough time […] you really want to help them, but the caregivers won't come anymore, then it's hard.

They also said that there were too many sessions of TF-CBT for families. This was thought by counselors to be largely due to families' unfamiliarity with this type of therapy, and the fact that families were used to HIV/AIDS counseling (i.e., Voluntary Counseling and Testing or VCT) which was described as usually one brief (5-15 minute) session. One commented, “For them they think is like counseling 2- 4 sessions then comes to end – like VCT. They don't know TF-CBT has some stages.” Another common challenge reported by multiple counselors was talking about sex and sexual body parts with the children and caregivers.

Perceived changes in children and families

The majority of counselors stated overall positive change/growth giving specific examples such as better attitudes, more trust in others, sleeping better, goes to school, keeps good hygiene, is less angry, and more relaxed. Most counselors also specifically commented on how the child was much more “open” after TF-CBT. For example, children were described as feeling free to talk about trauma and things that were taboo (e.g., sex), and talk with/to the parent(s) about problems they are having. Counselors stated,

Child may come to you…at the beginning child is shy and tense, you may hear from caretaker, that the child is isolating himself and avoiding the company of people. As you move on with some sessions, see the child try to look you in the face, able to relate, those are just positive impact of TF-CBT on children.

First time [you] met the family, [the] father comes in, mother behind, and child lagging behind. As we continued therapy they would walk as a unit. Towards the end of the program […] they could easily tell their daughter that they were sorry which is not easy for most Zambians to tell their children that they are sorry.

There were other behavioral changes listed such as becoming more friendly, having improved relationships, and being able to reach out when they need help. Counselors also reported improved child/caregiver relationships with more communication and interactions, as well as better parenting skills (e.g., from beating to using praise and consequences for bad behavior). Counselors noticed more understanding from the parent of the child's traumatic experience(s), less fear from the child towards the parent, more trust within the family, and more “fun time” spent as a family. Some counselors also spoke of seeing the clients (both caregivers and children) learn how to make their own decisions, and start looking at situations in a more positive way. For instance, “There is growth not only in the child but in the parent, parents learn how to communicate positively.”

Cultural adaptations

Counselors discussed that TF-CBT principles made sense to them and their culture, and modifications were made primarily in activities used. For example one counselor explained,

What I mean is, almost all the components that are on the TF-CBT are acceptable in our culture context. You don't find them conflicting with our culture, depending on how you implement them of course. They are applicable in our culture context.

Most modifications that were made centered around language, examples, activities, and analogies. For example, with relaxation they role played with the child as a cooked and uncooked okra (versus noodles often used in the U.S.). Many reported needing to make modifications due to the use of tribal languages, such as finding ways to understand different feelings when there was only one word used to describe all of them. One counselor explained, “Transferring what I meant, make it more child friendly, transferring it in a way they can understand local language, certain aspects that are in the English language that just are not there in the African languages.” Counselors also stated that talking about sex is taboo in Zambian culture, and that they had to engage an adult to discuss how, in this program, it was okay to talk about those things. Interviews also discussed the need to make adaptations around the parenting skills component of TF-CBT. For example, counselors reported that giving praise to a child is not common in most families within the Zambian culture and would be new to caregivers. Counselors used caregiver's personal experience (e.g., have you ever been praised by your husband for cooking a good meal?) to explain the rationale. In addition, skills like “time out” were difficult to implement due to space constraints.

For example when look at parenting skills, sitting there telling a parent to start to praise their child when they grew up not hearing the parents saying I love you or that is great what you are doing. So not easy to put across to someone who has never heard these words before.

TF-CBT training and supervision

A theme that came out of the counselor interviews was specific to the training and supervision model. Almost all the counselors discussed the importance of the group practice and supervision groups for learning, motivation and support.

Our supervisory groups – yea they're great and very helpful for running through your plan of what you're going to do with your client and next time you meet with them you get different ideas and suggestions from others in group and supervisor.

The Apprenticeship model used is very different from the usual “one-off” trainings. Many particularly commented on the usefulness of role-plays to learn: “On the role plays you know the teaching we are used to is lecture type, but what the trainer did was to enhance the role playing.” Respondents discussed the overall training as having many positive characteristics including that it was comprehensive, well organized, and flexible. However, some counselors felt that the initial live trainings (there were 2 5-day trainings) should be longer, that they were “short and intense.” One explained, “The first challenge was it was too compressed, like 5 days you have to go for training from 8 to 16.30 in afternoon and you would have a lot of things to do. You find there are certain things maybe later in the afternoon, you would not concentrate.”

Suggestions for improvement

The counselors recommended a number of suggestions for improvement. On a community level, many stated there is a need to create a designated space for therapy in Zambia that would be safe, and separate from all the other clinics (e.g., HIV services that are heavily stigmatized). They all suggested that the community could benefit from more awareness around sexual abuse and a treatment like TF-CBT. One counselor said,

You can even advertise TV, radio, flyers, so people know if you have a problem of this nature, this is where you can go. I think it would help, because of stigma that's attached to University Teaching Hospital for whatever reason, I think a lot of cases go unreported.

Counselors also suggested an improved client referral system for mental health, more funding for mental health and sexual abuse, and scaling up TF-CBT to other areas in Zambia and even other countries. More related to TF-CBT as a model, they recommended shortening the session length, and developing more ways to improve patient attendance.

Caregiver and Child Responses

Qualitative interviews were given to 18 (86%) of the children and 16 (76%) of the caregivers who completed TF-CBT (Total completed; N=21). Four of the families were not able to meet with interviewers due to moving from the area, or not being available because of school and/or work. Two caregivers were not able to meet due to other livelihood commitments (e.g., selling at the market), although their children did respond. Of the 18 children, 10 were interviewed twice, 7 children were only interviewed once, and one child was interviewed 3 times. Of 16 caregivers, 8 were interviewed twice, 7 were interviewed once, and 1 participated in three interviews. Themes included, a) The program helped, b) Challenges/Dislikes, and c) Recommendations to make the program better. The most frequently mentioned ‘included terms’ (four or more responses) are presented in Table 2.

Table 2. Zambian Client Perspectives (N=18 Child; N=16 Caregivers).

COVER TERM	Included term summary	Child	Caregiver	Total: Number reporting
The program helped
	Relationship building/enhanced relationships family and social	15	13	28
	Program helped/ “Program was good”	8	10	18
	Increased/improved communication	8	10	18
	Improved functioning	10	7	17
	Able to modulate affect; positive coping	10	2	12
	Overall improvement in behavior/ positive change	3	8	11
	Learning about thoughts/feelings/behaviors	6	5	11
	Enhanced parenting skills	1	10	11
	Avoid negative peer influence; decrease alcohol/sex behavior	7	3	10
	Learned safety skills	10	0	10
	Decrease in avoidance; decrease reaction to trauma	7	3	10
	Learned a lot	5	5	10
	Decrease in symptoms/makes people better	8	1	9
	Helped child change inaccurate/negative thoughts	5	4	9
	Program helps through relaxation	6	2	8
	Child learned to be free/open	3	4	7
	Helped through novelty of program; different from usual	3	4	7
	No longer blame child	1	5	6
	Strong therapeutic relationship	2	4	6
	Helped through psychoeducation of child sexual abuse	2	4	6
	Increased enjoyment in life/program changed life	4	1	5
	Self-efficacy	4	0	4
“Dislikes”
	Nothing negative to say, liked entire program	6	3	9
Recommendations
	Continue program	7	4	11
	Program should expand	2	4	6
	Program should be linked with law enforcement and justice system	0	5	5

The program helped

All children and caregivers talked about how the program helped in response to one or more of the different questions asked. Some responses were more broad such as, “the program helped,” saying there was overall improvement in behavior, or stating [we] “learned a lot.” Many responses described more specific changes they felt were due to TF-CBT. There was fairly consistent overlap between caregiver and child responses. The most frequently discussed impact of TF-CBT was enhanced social and/or family relationships (28 out of 31). One child said, “We started knowing each other in the family. We worked together in a good way.” Other frequently mentioned specific changes the program helped with include: increased/improved communication, improved functioning, ability to modulate affect, improved cognitive processing, and avoiding negative peer influence, alcohol and sexual behaviors. One caregiver commented, “She has changed like I said she now does not drink alcohol, does household chores, she plays with friends and does not sleep around with men.” Enhanced parenting skills was a common response, but almost always listed by caregivers. Interestingly, learning safety skills and ‘decrease of avoidance and reaction to trauma’ was frequently mentioned but primarily by children. Another response made by caregivers and children was that the “child learned to be free and open.” For example, one child stated, “It has helped me to be free and not to be scared of anything and to tell when something is bothering me.” Notably, this is a phrase in the local language that also came out in an initial qualitative study (Murray et al., 2006) – specifically that a symptom of trauma is that a child is “not free or open.” There were three included terms under the cover term of “the program helped” (improved cognitive processing, helped child change inaccurate/negative thoughts, and no longer blame child) that deal in some way with cognitive reprocessing - one of the core elements of most cognitive behavioral therapy treatments. For example, one child stated: When I was raped I used to cry when I think about it. I would blame myself that it is because of me that's why I was raped. But due to the program and the counselor I should not be blaming myself about what happened to me because it was not my fault.

Challenges/Dislikes of TF-CBT

When asked this question, most caregivers and children stated they had nothing negative to say and liked the program. Criticisms of the program that were reported (primarily one or two responses only) related to life circumstances and/or poverty. For example, caregivers reported that lack of transportation or the time required for the intervention were negative aspects of the program. Other comments had to do with disliking how small the program was, expressing the need to expand services and/or make services more available. For example, one caregiver said, “The problems were monetary and also problems of sharing time to do activities- going to UTH and coming here.”

Recommendations to make the program better

Overall, the most frequently mentioned recommendation was to continue the program. Caregivers added that the program should be expanded (e.g., to adults, and other areas of Zambia), and recommended more capacity building with law enforcement and justice system to improve the program.

I have learnt that this program is only in few areas, it would be appreciated if it would be made available in most areas of the country and made known to children so that even on their own they can access this service. (caregiver)

You should continue helping people. Some people out there are being abused, some of them it's their parents who are abusing them. So you should continue helping people. I teach my friends about abuse. (child)

Discussion

To our knowledge, this is the first study examining counselor and client (both caregiver and child) perspectives on TF-CBT in a LMIC. Based on interviews, we found overall positive perspectives to and acceptability of TF-CBT as a treatment. Counselors felt that the treatment's structure and flexibility were good, and appreciated the inclusion of families. They described cross-cultural adaptations as being primarily focused on games, activities and language. Caregivers and children primarily discussed positive outcomes of TF-CBT. All clients suggested continuing and expanding the program. The overall positive tone of the interviews suggests that TF-CBT would be an acceptable mental health intervention from the perspectives of local counselors and clients.

Collectively, the results suggest that respondents perceive that TF-CBT was responsible for a variety of positive changes. All respondents identified a decrease in the child's symptoms, and an increase in the child's ability to function both at home and school with many giving specific examples. All groups discussed the positive impact on social and family relationships and enhanced communication. Although TF-CBT is focused on the treatment of children, both counselors and parents identified that this treatment helped in their own lives as well. Children and youth themselves reported a decrease in symptoms and increase in their ability to cope and form healthy relationships as a result of TF-CBT. This is supported by the quantitative decrease seen in symptomatology, albeit from a small sample. These findings also speak to the clinical significance of TF-CBT, which is important to examine alongside efficacy studies and particularly critical for implementation, uptake, and sustainability (Schoenwald et al., 2008). Counselors' responses made an important contribution to understanding task shifting (i.e., the training of lay workers to implement EBT) and the use of EBT from a provider perspective. They voiced that the modifications made to EBT are centered on language and the way in which core elements are implemented (opposed to core constructs), such as using different games and analogies and simplifying language. This is supported by a small literature on examining the cultural adaptation of EBT in LMIC such as Interpersonal Psychotherapy (IPT) (Patel et al, 2011; Verdeli et al. 2003; Verdeli et al, 2008) and Cognitive Processing Therapy (CPT) (Kaysen et al., 2013). Still, ongoing research on adaptation of EBT in LMIC is warranted, specifically with children and families. Counselors also positively discussed the Apprenticeship Training and Supervision model which differs from the “train and hope” model (i.e., training with no follow-up; Kelly et al., 2000). This study suggests that this model may be particularly valued by providers when using a task-shifting approach. Equally important are the recommendations that the training seemed “short and intense” – so much so that our team has now altered the model to an initial 10 day training, rather than 5 days (Murray et al., 2011). Some of the positive responses from these lay counselors are different from findings in the United States (U.S.). For example, Zambian counselors reported liking the structure of the program, where a frequent complaint of EBT from therapists in the U.S. is the lack of flexibility (Kadzin, 2008; Kendall, 2011; Kendall & Beidas, 2007). Zambian counselors also made positive statements about role plays, feedback and were hungry for more training time. These responses may have differed if asked of experienced mental health providers suggesting that certain implementation strategies may be more palatable to newcomers to the field of mental health. Finally, the case completion rate, one of the largest challenges cited by our counselors, indicates a strong need for additional engagement in the program. This is supported by clinical research elsewhere that engagement of families, particularly those with multiple stressors such as poverty, mental health, and trauma, is highly critical (Interian, Lewis-Fernandex & Lixon, 2013; McKay et al, 2004). Following the study, our team has worked with the counselors to add a session on engagement including an additional explanation upfront about TF-CBT and how it differs from the services people may be more accustomed to such as 1-2 sessions of VCT.

Clients raised some critical feedback around how to effectively implement EBT in this LMIC. One challenge was logistics, such as transport and the money required for this. Lusaka is quite a large city (2,191,225 people) and most rely heavily on “mini-buses” which frequently break down, are delayed, or full. Another challenge raised was the stigma attached to the location where children were referred from, and possibly seen, which our clients felt were related to HIV. A suggestion was a free-standing, safe, and child friendly location dedicated to services like TF-CBT. Although a small sample, the clients spoke to the clinical significance of TF-CBT by listing all the positive changes they experienced from participating.

This study generated some new hypotheses and ideas about the impact of TF-CBT. Some caregivers and children indicated an increase in ability to make better choices and to identify and stay away from negative peer influences due to TF-CBT, and also a decrease in substance use and risky sexual behaviors. Many of these statements were specifically focused on HIV. For example: “I found the program to be helpful to the behavior of the youths. It changes bad behavior such as sexual behavior of boys and girls, drinking alcohol at a tender age….” This finding suggests a hypothesis that a mental health focused intervention could have an impact on risky sexual behavior and other negative behavior patterns. This is supported by some literature from the U.S. documenting that TF-CBT does show improvements in behavior problems, although this is defined more generally rather than specific to HIV (Cohen et al., 2010; Cohen et. al., 2004). Small but promising research from other LMIC suggests that cognitive-behavioral interventions in general can make a difference on risky sexual behavior (Sikkema et al., 2010). Future research should examine whether a model like TF-CBT can help both with mental health symptoms and also be effective in decreasing risky sexual behavior.

Notably, there were many statements indicating an increase of knowledge and use of safety skills – despite the contextual challenges. One challenge in this setting is that many times perpetrators are not removed from the community or family systems due to underdeveloped health and legal systems, lack of shelters, and poverty. For example, in Zambia, there were cases where the perpetrator remained in the home because they were the bread-winner and necessary for survival. The idea that despite situations like these, caregivers and children felt their safety was enhanced is a significant finding.

Across all the respondents, the majority of challenges focused on organizational or broader systems issues such as lack of space to provide services, limited understanding from the local population of mental health treatment (versus services such as VCT counseling), and other public health systems problems (e.g, unknown or limited referral systems, funding…etc.). This feedback is supported by literature reinforcing the need for support at various organizational levels (e.g., service organization, greater health systems) for effective implementation of EBT (Cummings, Estabrooks, Midodzi, Wallin & Hayduk, 2007; Greenhalgh, Robert, Bate, Macfarlane & Kyriakidou, 2005; Schoenwald et al., 2008). More research is warranted on the multiple aspects of implementing EBTs in LMIC, including organizational outcomes.

Limitations

Several important study limitations merit attention. The caregivers and children/youth who participated in this study represent those that completed TF-CBT in full. This could have significant effects on their positive reactions indicating that the full course may be needed to see such beneficial changes, and/or that this is a sub-population that had a different perspective from the start about a program like TF-CBT. Indeed, there were statements from the counselors about families starting and not finishing the TF-CBT program. Non-completers could not be found for the qualitative study due to the time lapse and the limited resources for the project. Our research team has now begun including a case retention specialist in studies in LMIC to aid in this common challenge of loss to follow-up. There may also be potential biases due to a TF-CBT counselor (not the one that provided treatment to the family) performing the interviews. Families could have still seen them as a TF-CBT counselor and then hesitated to report negative feedback. It may also be possible that families saw the counselors as holding some authority, and thus reluctant to criticize the program. We decided to use TF-CBT counselors due to their known skills to talk to children and families appropriately, however, future studies may utilize outside interviewers when interviewing families to reduce bias. Another important consideration is that this study was based on a small feasibility study with too low a sample size to make any firm conclusions. The sample of children was also all female, limiting generalizability across genders. It is important to remember that qualitative studies are good for getting local perspectives and suggesting next steps of research. This study shows the local perceptions of TF-CBT but is not proof of impact. Although many caregivers suggested expanding TF-CBT to rural areas, the overall positive perceptions in Lusaka (urban) should be taken with caution without considering possible contextual differences between rural and urban Zambia. This study was also conducted on one group of local lay workers trained in TF-CBT, and may not fully represent all providers that may partake in training and implementation. Finally, if resource permitted, we would have preferred to have analysis done completely by local individuals.

Despite these limitations, the findings of this study are particularly relevant because they are the first that we are aware of to document the perspectives of lay workers and clients in a LMIC on an EBT (TF-CBT) to treat mental health symptoms of trauma for children and families. There continues to be an energetic debate in the field of global mental health of whether the use of evidence-based mental health treatments in LMIC is ethical, valuable, transportable, effective, and culturally sensitive. There is indeed growing evidence that EBTs can be more effective for decreasing symptoms as well as transportable and adaptable (Bolton et al., 2003; Bolten et al, 2007; Patel et al., 2011; Rahman et al., 2008). Research from the USA is examining whether EBTs are in fact effective with cultures and ethnicities outside those for which they were developed (see Huey & Polo 2008 for a review). However, there are also documented concerns about the implementation of EBTs cross-culturally that needs to be considered (Wessells, 2009). Therefore, it is critical that the field of global mental health take into account perspectives of the very individuals learning and implementing such models, as well as those receiving them.

In summary, this study was part of a longer chain of studies rooted in ethnographic work that suggested trauma as a locally important problem, and an interest in testing a model like TF-CBT. Results from the perspectives of local lay workers trained in TF-CBT, and children and caregivers who received TF-CBT indicate overall positive perspectives on TF-CBT, particularly noting the positive changes in symptoms, family relations and functioning. This study highlighted various changes in implementation that may be warranted such as bringing on a retention specialist, and a longer training period. Results also put forth ideas for future research such as conducting a trial of TF-CBT to examine effectiveness on mental health and HIV risk behaviors in this context, and increased focus on implementation factors such as broader acceptability, penetration, uptake, sustainability, organizational variables, and costs (Proctor et al., 2009). Studies like this will move the field towards understanding effective and sustainable treatment options in LMIC for children to help reduce the enormous treatment gap (Kohn, Saxena, Levav & Saraceno, 2004; Wang et al., 2007).