Contextualization of psychological treatments for government health systems in low-resource settings: group interpersonal psychotherapy for caregivers of children with nodding syndrome in Uganda

At the initiation of the study, the team reviewed existing government health service guidelines for NS and psychosocial, psychological, and mental health services. We conducted a review of literature on existing evidence-based PTs by non-specialist health workers that had previously been conducted in Uganda [10, 11]. Limiting searches to Uganda, we searched using the keywords: “psychological treatment,” “psychological intervention,” “psychotherapy,” or “psychological therapy.” We searched Ovid MEDLINE, PsycINFO, PubMed (including in-process citations, ahead of print), and Web of Science.

The health system in Uganda is organized into seven levels, and mental health is included in the Minimum Health Care Package provided at all levels of care (Fig. 1) [31]. The study was conducted in Pader district which had more than 50% of reported NS cases in northern Uganda [29] (Fig. 2). The district is subdivided into 12 sub-counties, each served by at least one level II and III health facility. The two study sub-counties each had one health facility designated as a NS treatment center. Atanga sub-county, the intervention site, had a population of 34,103 in 2014 at the time of the program [32].

The principal investigator and principal medical officer (PMO) in charge of NS response at the MOH held face-to-face consultative meetings with 11 stakeholders including district administrators and leaders, health workers, and representatives of NGOs. The goals of the meetings were to develop a partnership with the district leadership, establish administrative consent, and produce a joint agreement on the specific study sites, taking into account existing NGO programs. A 2-day stakeholder contextualization workshop was held with study investigators, representatives from the mental health division of the MOH and PT specialists who had participated in prior IPT-G randomized controlled trials (RCTs) in Uganda. The objective of the meeting was to inform the structure and content of the PT within the public health system, which would then inform the planned pilot of WHO mhGAP in Uganda. The meeting involved three activities: (a) presentations to stakeholders about the proposed study design, the national response to NS, the Uganda public health system, and the experience of World Vision Uganda (an NGO which conducted community-based delivery of IPT-G); (b) group work to make suggestions about possible adaptations to the structure and content of the intervention; and (c) plenary discussions to agree on how the adapted PT structure, content, and process could be implemented.

A qualitative study of caregiver burden for children with NS including focus group discussions (FGD) and in-depth interviews (IDI) was conducted. Results from this study informed the training program, supervision, structure, and content of the intervention [28].

The IPT-G therapists conducted a 7-day training in the local language, Luo, for VHTs from the intervention site. A mixture of didactic lectures and role plays for the first 5 days was used, followed by role plays of the treatment sessions for 2 days. Broad thematic learning areas included definition and presentation of NS, caregiver experiences as described in the formative study, symptoms of mental illness with a focus on depression, working in groups, principles and practice of counseling, IPT-G elements, documentation, and supervision. A 3-day refresher training on the medical management of children with NS was given to all participants in accordance with government guidelines [24]. After the seventh treatment session, a 2-day refresher training was held for VHTs from the intervention site using a variety of teaching methods including lectures, experiential teachings, storytelling, group exercises, and role plays. A rapporteur compiled daily session and activity reports to inform the training evaluations each day. Thirteen VHTs and two health workers from the supervising health facility attended the initial training. Ten VHTs and the two health workers attended the refresher training.

With support from the national coordinator for NS, we sought and received permission from the district administration to access demographic records of caregivers of children with NS. With the registry, VHTs identified caregivers and informed them about the study. Caregivers providing consent were administered the pre-intervention assessment form (Additional file 1). This form recorded the number of depressive symptoms based on the Patient Health Questionnaire (PHQ-9), level of functioning, problem areas, and therapy goals. Therapists together with the VHTs and health workers reviewed the pre-group assessment forms to determine which caregiver was eligible to join a therapy group. A PHQ-9 score ≥ 9 was considered indicative of depression and therefore eligible to participate. The PHQ-9 has demonstrated high specificity and sensitivity in screening for depressive symptoms in Uganda [33].

The intervention was conducted in seven villages of Atanga sub-county, involving 12 weekly treatment sessions with caregivers of children with NS. VHTs facilitated the IPT-G sessions, some of which were attended by the primary health workers who monitored adherence to the IPT-G process using a supervision checklist. Both the VHTs and the health workers received technical and support supervision from the IPT-G therapists and a clinical psychologist at weekly group supervision meetings. Weekly session and field supervision reports were prepared by VHTs, health workers, and a supervising clinical psychologist. IDIs were conducted with a sample of participating caregivers to document experiences of the intervention. Video documentation of sessions was done to guide supervision and increase treatment fidelity. The treatment-as-usual (TAU) control arm comprised care to caregivers of children with NS as recommended by the government guidelines [34].

We conducted post-intervention assessment exercises to explore caregivers’ experiences of the group intervention from the perspective of the VHTs and other members of the intervention team. Two weeks after completion of the group intervention, VHTs conducted a post-group assessment exercise among the 69 caregivers who completed the 12 group sessions. A post-group test instrument (Additional file 1) was used to assess depressive symptoms and functioning. Therapists, VHTs, and health workers reviewed the post-group assessment forms to determine which caregiver had reduction in their symptoms and if they achieved their therapy goal.

After individual caregiver post-group assessments, a debriefing meeting was held for VHTs. In addition, a half-day meeting was held 6 months after the group intervention, between investigators, research assistants, project administrators, and the implementers of the intervention. Study investigators reviewed the meeting minutes to identify lessons learned, challenges, and opportunities for scale-up. For dissemination, a 1-day community meeting was held at the study site, 8 months after completion of the intervention. Presentations from the national coordinator of the NS response, district officials, VHTs, and study team members were supplemented by testimonies from participating caregivers.

We conducted preparatory activities to inform scale-up of the intervention. Following dissemination of results at the study site, the study team developed a VHT training guide (available from first author) including adaptations made during implementation. We estimated costs of implementing components of the intervention package.

The literature review identified 17 publications in Ovid MEDLINE, 31 in PsycINFO, 18 in PubMed, and 14 in Web of Science as of December 31, 2012. After removal of duplicates, there were 34 publications on PTs in Uganda. Nine publications presented results from four RCTs (see PRISMA figure in Additional file 2). The trials include two types of PTs in Uganda: group interpersonal psychotherapy (IPT-G) [10, 11] and Narrative Exposure Therapy (NET) [38, 39]. Both IPT-G and NET had demonstrated efficacy in treatment of mental disorders in adult and adolescent populations. Both trials for treatment of war-affected adolescents were conducted in northern Uganda, a setting similar to that of our study. No prior studies had evaluated a PT in a government health setting.

When determining which of the PTs to select, we found that IPT-G had published literature on adaptation and training [13, 14] which would facilitate development of intervention materials. Though NET trials utilized a cadre similar to VHTs used in our study, NET requires a strictly manualized process for delivery [38]. Because of this, researchers developed another treatment; trauma counseling (TC) to simulate how NET might be used out of a trial setting in a regular care system and that allowed counselors to be flexible in their treatment delivery [38]. However, this was not standardized and would pose challenges in ensuring fidelity to treatment by the lay counselors. In addition, locally adapted materials for low literacy populations in Luo existed for IPT-G that necessitated minimal adaptations for use by VHTs. Training in NET lasts 6 weeks for lay counselors [38] unlike training for IPT-G which lasted 2 weeks at most [13], making it a more feasible option. Of note, group treatments were more culturally acceptable [28] and delivery of IPT-G weekly sessions seemed to be more logistically feasible than the individual NET treatments which required more human and financial resources [28]. Our primary study outcome was depression, and the evidence of treatment effectiveness for depression is stronger for IPT-G in Uganda compared to NET [10, 11, 38, 39]. In comparison to individual treatments, group PTs are thought to have other potential benefits including increased social support, acceptance, hope, and belonging among distressed individuals [40, 41]. Furthermore, IPT is a WHO mhGAP recommended treatment unlike NET and with sustainability in mind, using IPT-G would inform scale-up of psychosocial interventions in line with the MOH strategic plan. Therefore, IPT-G was selected as a potential treatment for caregivers of children with NS.

A number of adaptations to the study design were suggested during the stakeholder adaptation meeting (Table 2).These included shorter periods of training for VHTs compared to that for IPT therapists in previous IPT-G interventions in Uganda in which training was for 2 weeks. Unlike NGO workers who were paid to deliver IPT-G, government VHTs are volunteers and can only afford to be away from their homes and/or livelihoods for a few days. Uganda national guidelines recommend 2 days per training for VHTs. However, following advice from MOH, a 7-day training program for VHTs was acceptable. Another modification was separating health workers from VHTs for training. This would maintain the expected Ugandan healthcare hierarchical structure.

Modifications were made to tools used in previous IPT-G studies. To assess caregiver depression, stakeholders chose to use a Luo version of the PHQ-9 questionnaire rather than the Acholi Psychosocial Assessment Instrument (APAI), which was used in the prior studies [11]. The PHQ-9 has been used in primary care settings within Uganda [42] and in contrast to nine items for the PHQ, the APAI is a greater burden for participants because it has 60 items. In addition, the APAI which was developed for adolescents emotional and behavioral problems, malwor and gin lugero [43], was not applicable to younger children and adult caregivers in our program. The measure of caregiver functioning (Additional file 1) was similar to that used in previous studies [10, 11, 44]. Stakeholders agreed with the study proposal to have VHTs deliver the intervention and health workers as supervisors of VHTs. IPT-G therapists would be overall supervisors of the intervention delivery.

An additional modification was that therapy groups would be co-facilitated by two VHTs of different gender and educational levels rather than only one VHT as we originally proposed. Dual facilitation would assure continuity of the group sessions in the event that one of the VHTs was absent. Government staff, following observations by the IPT-G therapists, supported the decision to shorten the original 16-session IPT-G conducted in Uganda [10] to a 12-session structure (two initial sessions, eight middle sessions, two termination sessions) which would increase the likelihood of delivery when scaled-up in government settings.

Family challenges associated with caring for children affected by NS, particularly those related to caregiver mental health, were identified [28]. Key contextualization issues included acceptance of facilitators of either gender (vs. gender-specific matching of facilitators and groups in earlier trials) or a focus on a specific problem (nodding syndrome vs. focusing on a range of psychological stressors).

The training of VHTs was conducted in 7 days. Training was participatory with VHTs contributing to the modification of materials, such as translation of concepts to less technical, culturally sensitive language to be used in intervention delivery (Table 3). Common local terms representing symptoms of depression, e.g., cwer cwinyi (feeling sad or hopeless), were suggested and agreed upon by VHTs. This exercise also informed further translation of the intervention tools including in-session and pre/post assessment forms. Orientation of health workers to IPT-G supervision was done by one of the IPT-G therapists, in a separate session during the course of the VHT training.

VHTs were able to effectively recruit an adequate number of participants. Eighty-six caregivers participated in the pre-assessment exercise, and 73 fulfilled criteria for inclusion. However, 10 caregivers who were assessed but did not meet inclusion criteria (i.e., their scores on the PHQ-9 were below the cutoff) requested to participate in the intervention, as did three other caregivers with children with NS who were not identified during the recruitment phase. Through consultation with the local leaders and health workers, we decided to include these individuals because when scaled-up into government health services, the intervention could be made available to persons requesting care with distress at subclinical levels. Similarly, once the group sessions began, two caregivers of children with nodding syndrome were allowed to join though they had initially declined participation. This is because under real-life conditions, caregivers may join at different times based on their need and timing of services. Although we originally conceptualized closed-groups, VHTs and district stakeholders stated that open-groups would enhance acceptability and utilization by community members. For the purposes of quantitative evaluation, we used an intention-to-treat approach and only included those participants who met the inclusion criteria.

The intervention was delivered in three stages: initiation or early phase, middle or intermediate phase, and the termination phase (Additional file 1; a sample session is available for viewing online (https://​www.​youtube.​com/​watch?​v=​RQ2eYn-J8w8). Supervision of the intervention by IPT-G therapists and clinical psychologist focused on adherence to IPT-G structure and addressing challenges faced by VHTs during IPT-G delivery. Modified supervision topics included follow-up of caregivers with suicidal behavior. Challenges included poor documentation, difficulties with supporting caregivers experiencing domestic violence, and motivating those caregivers with symptom resolution to stay in treatment. Peer supervision by more literate VHTs was used to support documentation for less educated VHTs. IPT-G therapists observed sample group sessions to document challenges and tailor content of refresher trainings.

Intervention completion was documented through caregiver attendance records at each therapy session. Sixty-nine caregivers attended each of the 12 therapy sessions which translated into more than 90% of study participants. However, only 63 caregivers and children were available during the post assessment exercise (Additional file 1). Reasons for caregiver attrition included relocation to other villages because of new employment opportunities, alcoholism, loss of interest, marital breakup, and one death.

Quantitative results of the trial have been published and demonstrate the effectiveness of IPT-G in treating depression in both caregivers and their children [34]. Significant effects were also observed for psychological distress, stigma, and social support among the caregivers [34].

VHTs found the session reporting format, which required written documentation, demanding because they typically do limited documentation for government health services. VHTs reported that the experience changed their own lives with marked improvement in their personal relationships, reducing alcohol use, and self-grooming. These changes were not reported by VHTs in the control group.

Challenges faced by VHTs during implementation of IPT-G included inadequate financial compensation, managing discussions when IPT-G participants’ partners attended the sessions, and alcohol consumption which affected caregivers’ attendance and participation.

Recommendations for scale-up provided to the MOH included the above findings as well as additional post-intervention suggestions from stakeholders that primary health workers serve as IPT-G facilitators before taking on the role of supervisors. Results from the study were also used to inform a pilot of the WHO mhGAP IPT component by the MOH in northern Uganda. The final result was a training curriculum and supervision guidelines for delivery of IPT-G by VHTs.