Implementing combined WHO mhGAP and adapted group interpersonal psychotherapy to address depression and mental health needs of pregnant adolescents in Kenyan primary health care settings (INSPIRE): a study protocol for pilot feasibility trial of the integrated intervention in LMIC settings

Background: Addressing adolescent pregnancies associated health burden demands new ways of organizing maternal and child mental health services to meet multiple needs of this group. There is a need to strengthen integration of sustainable evidence-based mental health interventions in primary health care settings for pregnant adolescents. The proposed study is guided by implementation science frameworks with key objective of implementing a pilot trial testing a full IPT-G version along with IPT-G mini version under the mhGAP/IPT-G service framework and to study feasibility of the integrated mhGAP/IPT-G adolescent peripartum depression care delivery model and estimate if a low cost and compressed version of IPT-G intervention would result in similar size of effect on mental health and family functioning as the Full IPT-G. There are two sub- studies embedded which are: 1) To identify multi-level system implementation barriers and strategies guided by the Consolidated Framework for Implementation Research (CFIR) to enhance perinatal mhGAP-depression care and evidence-based intervention integration (i.e., group interpersonal psychotherapy; IPT-G) for pregnant adolescents in primary care contexts; 2) To use findings from aim 1 and observational data from Maternal and Child Health (MCH) clinics that run within primary health care facilities to develop a mental health implementation workflow plan that has buy-in from key stakeholders, as well as to develop a modified protocol and implementation training manual for building health facility staff’s capacity in implementing the integrated mhGAP/IPT-G depression care. Methods: For the primary objective of studying feasibility of the integrated mhGAP/IPT-G depression care in MCH service context for adolescent perinatal depression, we will recruit 90 pregnant adolescents to a three-arm pilot intervention (unmasked) trial study (IPT-G Full, IPT-G Mini, and wait-list control in the context of mhGAP care). Pregnant adolescents ages 13-18, in their 1st-2nd trimester with a depression score of 13 and above on EPDS would be recruited. Proctor’s implementation evaluation model will be used. Feasibility and acceptability of the intervention implementation and size of effects on mental health and family functioning will be estimated using mixed method data collection from caregivers of adolescents, adolescents, and health care providers. In the two sub-studies, stakeholders representing diverse perspectives will be recruited and focus group discussions data will be gathered. For aim 2, to build capacity for mhGAP-approach of adolescent depression care and research, the implementation-capacity training manual will be applied to train 20 providers, 12 IPT-G implementers/health workers and 16 Kenyan researchers. Acceptability and appropriateness of the training approach will be assessed. Additional feedback related to co-located service delivery model, task-shifting and task-sharing approach of IPT-G delivery will be gathered for further manual improvement. Discussion: This intervention and service design are in line with policy priority of Government of Kenya, Kenya Vision 2030, World Health Organization, and UN Sustainable Development Goals that focus on improving capacity of mental health service systems to reduce maternal, child, adolescent health and mental health disparities in LMICs. Successfully carrying out this study in Kenya will provide an evidence-based intervention service development and implementation model for adolescents in other Sub-Saharan African (SSA) countries. The study is funded by FIC/NIH under K43 grant.

Figure 1 provides an overview of the study procedure and design for the pilot randomization trial. Below we provide detailed methodological approach for the pilot feasibility trial and the two sub-studies. Considering high prevalence of adolescent pregnancy, high risk for depression in this population, and low resource in MCH clinics, the WHO recommended Full version of IPT-G may not be practical in meeting population service demands and needs. Therefore, we will test a Mini version of IPT-G in comparing to the Full version of IPT-G in the context of mhGAP mental health service delivery model. For this Aim, our primary goal is to evaluate the feasibility of implementing the integrated mhGAP-IG/IPT-G service model in the MCH contexts as well as feasibility of applying RCT design to study differential impacts of the IPT-G Full and Mini version in MCHs. Our secondary goal is to estimate the sizes of effect for the mhGAP/IPT-G Full and Mini versions. Figure ​Figure11 provides an overview of the pilot trial design. The two MCH clinics that we build capacity in Aim 2 will be the pilot Implementation sites. During the 1st implementation cycle (about 2–3 months), we will work with only 1 MCH clinic (Cohort 1), and then work with another MCH clinic in the 2nd implementation (Cohort 2). For each cohort, 45 pregnant adolescents with moderate to severe depression will be recruited. As the interventions will be using self-reported measures and the adolescent participants would be directly involved the intervention arms would be unmasked. Pregnant adolescents (ages 13–18), screened with moderate to severe depression during the 1st to 2nd trimester (defined as Edinburgh Postnatal Depression Screen (EPDS) score ≥ 13 or meet the APA’s Diagnostic and Statistical Manual for Mental Disorders IV-TR version (DSM-IV-TR) criteria for major depression disorder and perinatal depression), who agree to give informed consent (by adolescents and their adult caregivers), are willing to come for group sessions at MCH clinic, can participate in evaluation assessment and give consent to publish the findings of the work, will be eligible. Participants will be excluded from the study if they suffered from or showed evidence of severe personality disorder, acute psychosis, suicidality, (however will be given referral to appropriate services upon identification of any significant risk) or where there will be significant substance abuse, or evidence of severe comorbidities. Participants will be randomly assigned to IPT-G Full, IPT-G Mini, and wait-list control using computer random generator by research staff (n = 15 for each arm, 45 per cohort). Schedule of Enrolment, Interventions, and Assessments.**Recommended content can be displayed using various schematic formats. See SPIRIT 2013 Explanation and Elaboration for examples from protocols To evaluate feasibility of the mhGAP-IG/IPT-G implementation model in the MCHs, a mixed methods data collection approach will be applied and data will be collected from multiple sources (i.e., implementers, MCH leaders, adolescent mothers and their family members) to minimize data collection bias. As described above, the Conceptual model of Implementation Outcomes [56] is applied to guide the construct selection for the implementation feasibility and impact estimation. Feasibility of the mhGAP/IPT-G mental health service delivery model will be assessed using three indicators: fidelity (i.e., % of sessions is delivered with > 80% of contents), level of engagement (e.g., % reach targeted population, % participating in intervention sessions), quality of implementation (e.g., acceptability, appropriateness, usefulness, satisfaction of the mhGAP/IPT-G service model), and perceived mental health service (e.g., mental health service quality, MCH climate). Feasibility of applying a trial design for mhGAP/IPT-G impact evaluation in MCHs will be assessed using two indicators: % of adolescent-caregiver pairs that are exposed to intervention contents within the assigned condition; and % of data collection performed as planned. Impacts/effect size estimation for the IPT-G Full and Mini versions (the secondary outcomes) will be assessed for adolescents’ mental health and family functioning outcomes. Both quantitative data and qualitative data will be collected. Qualitative data will be collected through focus groups (2 groups for each intervention arm) at the end of implementation cycle. Tables ​Tables11 and ​and44 lists the feasibility outcomes measures and sources of data collection (under implementation feasibility, mental health service outcomes, and trial design feasibility and effect size estimation). Implementation feasibility outcome data, collected from the MCH staff/intervention implementers and adolescents who participated in the two intervention arms (IPT-G Full and Mini), will be gathered after each group session throughout the two implementation periods. For effect-size estimation, all study participants will complete a baseline (Time 1) interview (on mental health and family functioning outcomes) before randomization and intervention implementation, followed by a 1- or 2-month implementation period (2 months for the IPT-G Full and 1 month for the IPT-G Mini). Only when a participant is physically unwell, or has to terminate due to a pregnancy related emergency or any other severe mental health issue will their participant grouping be unblinded and an appropriate referral made. They will receive a post-intervention interview assessment at the end of implementation cycle (Time 2, about 3–4 months after the Time 1). To facilitate coordination of program implementation activities, participating MCH will be provided with two tablets and a small incentive which could offset implementation costs. Research implementation steps and data gathering activities Community advisory (17members), Technical advisory board (5 members) a. FGDs, KIIs b. Short survey for conjoint experiment a. ethnographic- observational study, b. training workshops with the researchers, c. training workshops in the health facilities a. Stakeholder engagement through advisory board meetings and group discussions, b. Expert feedback Several planned publications including: a. mhGAP/IPT-G adaptation for adolescent mothers with depression; b. Acceptability and Appropriateness of the Brief and Long versions of IPT-G intervention implemented by CHWs (results from a mixed method study), c. D&I context measures validation paper, d. Implementation Effectiveness Evaluation (using RCT implementation study data to study effectiveness for brief and long version of IPT-G) A randomly selected sample of 90 adolescent-parent dyads will be recruited. It has been suggested that a minimum of 30 participants per group/arm is sufficient for individual IPT and IPT-G pilot studies based on similar intervention studies in LMICs [17, 18, 23]. Data collected through 2 cohorts will be analyzed. Quantitative analysis will be focused on examining the following questions: The following stop-go criteria will be used to determine whether to proceed to the main trial: the recruitment proceeds well so that the first IPT-G cohort can proceed by the end of the first month of the specified recruitment period. The IPT therapists that is the trained CHVs and nurses demonstrate adequate levels of treatment fidelity to the WHO IPT-G checklist adherence available in the manual and 70% of participants attend at least 5 of the available sessions in the course of their treatment and 80% are successfully followed up at primary endpoint. Quantitative analysis for implementation outcomes will be based on data collected throughout the implementation period, and analyzed separately for each time point as well as jointly to create overall summary scores. Analysis for the MCH service outcomes will be based on pre-post intervention data using paired-t tests and repeated measures. We will examine patterns separately and jointly for two MCH clinics. Analysis for estimating size of impact on mental health (depression) and family functioning will be based on multi-level modeling, adjusting for family/group nesting effects. Post-intervention effects/impacts will be estimated as a function of baseline levels of the corresponding outcome variables and intervention status. Effect size (regardless the statistical testing significant level) will be computed. Quantitative data for implementation outcomes, MCH service outcomes, and adolescent mental health effectiveness outcomes will be examined separately. Qualitative data will be transcribed and analyzed using similar techniques described in Aim 1. We will review all focus group discussion data and use grounded theory approach to search for emerging themes using open and axial coding. Analysis will focus on themes related to MCH service quality (on providing both MCH and mental health care), feedback on implementation process and intervention contents, and perspectives on theory of change. To understand ‘fit’ of the mhGAP/IPT-G and study barriers/strategies (in 5 CFIR domains) to promote utilization and impacts of mhGAP/IPT-G, we will carry out a mixed methods study, including qualitative and quantitative data collection. Quantitative data collection will focus on D&I readiness contexts and new D&I tools pilot testing (see Table 4 and Fig. 3 for data gathering activities for the entire study and SPIRIT chart). Group Interpersonal Psychotherapy full versus mini (or brief) versions We will conduct eight focus groups (n = 40) with representative diverse stakeholders (i.e., including two groups for pregnant adolescents and new adolescent mothers, two groups for adolescent caregivers and partners, two groups for MCH staff, 1 group for community advocates and policy leaders, and 1 group for child and adolescent mental health professionals). Each focus group discussion (FGD) will have about 5 participants. Each group will participate in three FGD meetings. One meeting will focus on barriers/strategies related to mhGAP-IG, and the other meeting will focus on barriers/strategies related to IPT-G. The third FGD would be focusing on mental health treatment preferences of pregnant adolescents. Each focus group will last 1½ to 2 h. Participants will participate in an FGD and complete a survey, and these will be carried out in English or Kiswahili (as English is the official language in Kenya). Focus group guides/questions and quantitative survey will be adapted from existing D&I studies [48], or existing D&I measures (see measures under sub-study 1). Stakeholders’ views on barriers and strategies in five CFIR domains will be discussed separately for mhGAP and IPT-G. Informed by our previous needs’ assessment work, strategies related to livelihood problem, partner/family engagement, knowledge gap, and stigma reduction will also be explored. Key information that we will gather from different groups of stakeholders is summarized below. Results from this study will be used to inform the adaptation of mhGAP and IPT-G contents, adolescents/ family members’ engagement activities, implementation process, and training of the implementation team. An adapted manual for mhGAP-IG and IPT-G (Full and Mini versions) will be developed. To better identify mhGAP/IPT-G workflow barriers and strategies in MCH contexts, MCH site visits and workflow observation will be arranged for the Community and Technical Advisory Board members (n = 22) (see Table 3). The observation will focus on MCH structure and perinatal service workflow. For the four adolescent representatives, we will seek two pregnant adolescents and two new adolescent mothers. The eligibility criterion would be as follows: willing to consent to participate, one pregnant adolescent and one new mother between ages 13–16, another pregnant adolescent, and new mother between ages 16–18 willing to join the meetings held at the community center or at the health facility. For all other stakeholders of the community advisory, the eligibility would be as follows: to sign informed consent to participate in the study and for the deliberations to be recorded, willing to engage with the issues around mental health and adolescent pregnancy, and willing to commute to the health facility or community center for meetings three times a year. Additionally, there would be a technical advisory board. The technical advisory board comprises of members whose organizations provided support letters during the grant and formally committed to providing technical guidance. Their eligibility includes the following: adults above 18 of years and technical persons representing mental health and health policy implementing partners in Kenya, willing to sign consent form for participation in the study, and willingness for discussion to be recorded. This will allow the Board and research members to gain better understanding about the clinical set up, implementation barriers and strategies, which will facilitate better and more sustainable workflow plan/strategy development to support mhGAP/IPT-G implementation. To ensure confidentiality, informed consent for all participants will be obtained prior to the data collection. Participating provider personnel and family members will be compensated for their time. Participating mental health professionals and officers of NGO/ Ministry of Education/ Ministry of Health will also be compensated. Results from sub-studies 1 and 2 will be used to inform the adaptation of mhGAP/IPT-G implementation protocol in MCH context and training manual development for mhGAP/IPT-G implementation. Additional feedback will be sought from Advisory Board before using these. To build implementation research capacity, the lead investigator along with her co-investigators and mentors will provide a short/intensive 3-day course on child mental health implementation methodology for research team members, relevant stakeholders, and graduate level students (n = 20 in year 2). To build service capacity for adolescent perinatal depression, we will train MCH implementation team (n = 16 from two MCH clinics) for task-shifting/sharing collaborative approach of mhGAP-IG and IPT-G implementation. Members of the implementation team (to be determined in sub-study 1, which may include family members/adolescents if peer leader/empowerment implementation strategy will be used) will receive a 4-day training (including one-day practice training) to prepare them for the implementation. Additional 8-group coaching/support from the clinical team (provided after each IPT-G group session) during the first intervention implementation period will also be provided. The coaching support after the first implementation will be based on the identified needs. Quality and impacts of training will be assessed through attendance tracking, after training satisfaction evaluation, pre- to post-training knowledge/competence assessment with trainees. The measures will be developed during the study period, and adapted from the research teams’ previous D&I studies [57]. We anticipate capacity building will significantly improve trainees’ (nurses and community health workers) knowledge and skills in implementation research and delivering of depression intervention. To document feasibility, qualitative training notes and feedback data will be gathered. Themes related to appropriateness and usefulness of the training, partnership and lessons learned for mhGAP/IPT-G implementation capacity building will be analyzed and reported for improvement in future capacity building approach. Quantitative data related to knowledge and competence will be analyzed using pair-t tests and repeated measures to explore impact across different types of MCH personnel. The protection of the rights of research participants is ensured by several aspects of the study protocol: All information collected from study participants during the course of the project will be kept confidentially. This includes information collected and stored in written and electronic form. As indicated previously, data storage also ensures integrity of research data and rights of study participants. Paper copies of identifying information, assessment measures, and other study materials will be maintained in locked research files in locked offices. Electronic data are secured by server maintenance that includes password protection, limited access to data by staff, different levels of access depending on the person’s specific level on the staff, and server securities, all of which ensure a high degree of protection from unauthorized users. Information will be coded by participant identification number. Linking of identifying information to research data will be kept to a minimum and even then, under key and lock and only accessible to key study staff. Identity of study participants will not be revealed in presentations or publications of study findings. Participants will be given the opportunity to consent to (or decline) the use of any specific data collected from them during the intervention phase for use in presentations or publications. The project staff will continuously evaluate the experience of study participants, with particular regard to participant reactions and risk exposure. Trends in data and findings will be examined periodically (yearly) in order to identify any changes in risk/benefit ratios that might necessitate a modification of the assessment protocol. In the unlikely event that monitoring of data reveals unanticipated or otherwise negative findings; the PI would report these to the field in the form of a presentation or publication. Furthermore, it would be critical to assess, to the degree possible, what characteristics of the study protocol contributed to the findings. Given that the study is nested in a career development grant given to the first author and that it is a pilot trial, the mentors and technical advisory board would act as a data monitoring board along with Fogarty/NIMH annual research reporting. General Purpose of the board: This is not a formal clinical trial DSMB board however a mechanism to address any adverse events that might occur during the study period. Adverse events occurring in any study participants will be reported to the Kenyatta National hospital and University of Nairobi institutional review board within 48 h of occurrence and the external review boards (advisory committee members of the K43 proposal and key mentors at University of Nairobi, University of Washington St Louis and New York University) within 72 h. On a monthly basis the study team will review summary reports of the adverse events. A local Data Safety and Monitoring Board (DSMB) consisting of DW, BA, and OG has been established and will review interim comparisons of the trial arms at 3 and 6 monthly intervals. The general purpose of this board is to maximize the safety and privacy of all study participants, and ensure the integrity, validity, and confidentiality of the data collection and analysis procedures. The protocol described herein details the integration of the goals of the DSMB into all aspects of study design, implementation, and review. The study team will report any serious and unexpected adverse events, or any problems that involve risk to the participants or others, to the Nairobi city council operational research review board and the Kenyan IRB. The study team will report any serious adverse events, along with the Kenyan IRB review response/outcome, to the NIH.

N/A

• Dima Health