The proposed study will be an individually randomized controlled trial with children (age 1 to 13 years at the time of screening for inclusion in the study) living in residential care, reintegrated back into family-based care, in Uganda. It is designed to evaluate the impact of adding a household-based parenting program to a standardized reintegration package that includes individualized case management support and a reunification cash grant, aimed at improving the reintegration of children living in residential care back into family-based care. The study population will include children living in residential care facilities (RCFs) in Mpigi, Mukono, Masaka and Greater Masaka districts in Uganda.

Study participants will be randomized to one of two arms of the study: the comparison arm and the intervention arm. The target sample size is 640 children with 320 in each arm of the study. Children assigned to the comparison arm will receive a standard reintegration package that includes individualized case management support and a reunification cash grant. Those in the intervention arm will receive the enhanced reintegration package that includes individualized case management support, reunification cash grant and a parenting intervention. Data will be collected at baseline (while the child is still living in the RCF), 6 months post-placement and 12 months post-placement.

Data will be collected in the local language by a project trained local data collection partner on the following six domains of reintegration: Child health and development, Psychosocial health and wellbeing of the child and primary caregiver, Protection and safety of the child, Caregiver-child relationship, Child's and caregiver's sense of social and community belonging, Education access, quality, and achievement (where age-appropriate).

The sources of data are a) interviews with primary caregiver, b) interviews with RCF caregiver, c) interviews with older children (8-13 years of age), d) standardized assessments of child cognitive functioning for all the study children, and e) focus groups and interviews with participants, parenting facilitators, and case managers.

Background:

There is an expanding global evidence base that illustrates the negative effects of placement in residential care on the physical, cognitive and socioemotional development of children (Smyke, 2007). Based on this evidence, legal and policy frameworks and programming have focused on reducing reliance on residential care and strengthening family-based environments for children (United Nations General Assembly, 2009). Recent care reform initiatives in Uganda promote the reunification and reintegration of children living in RCFs back into family-based care. Reintegration encompasses more than just the relocation of separated children into family-based care; it includes multiple dimensions of children's and families' wellbeing over time.

Uganda's population is young and vulnerable, with as many as 80 to 90 percent of children living in RCFs have at least one living parent (Rotabi, 2016). In Uganda, a study by Walakira, Ddumba-Nyanzi, & Bukenya (2015) found that more than two-thirds of the children living in 28 RCFs had at least one living parent and many more had a contactable relative. Children enter residential care for a variety of reasons and while poverty is often a driving factor, other factors such as inadequate access to quality education, health care, or social services, disability, child behavior problems, parental alcohol or drug abuse and death of one or more caregiver, may also be contributing or precipitating factors.

There have been several studies that have shown beneficial effects of parenting interventions on child cognitive outcomes in Uganda (e.g., Boivin et al., 2013). A recent systematic review of 12 randomized controlled trials (RCT) examining parenting inventions in low and middle income countries found evidence that parenting interventions positively affected a range of outcomes, including parent-child interactions, parental knowledge and attitudes, and reduced self-reported harsh or abusive parenting. The authors concluded that parenting interventions "hold some promise for improving parenting practices and reducing risk factors for child maltreatment in low and middle-income countries" (Knerr et al., 2013).

Study Rationale:

While there is a growing interest and emphasis in finding ways to transition children from residential care to family-based care, there is very little rigorous evidence that identifies specific interventions that help to facilitate this process, especially within sub-Saharan Africa. The proposed study aims to contribute to the evidence base by determining whether provision of a household-based parenting program contributes to reintegration success in Uganda. The six 'domains' of reintegration success are based on the following outcomes:

1. Child Health and Development Child

   a. Child anthropometry including height, weight, BMI; access to food and health care; child cognitive development; protective factors and resilience (only for older children)

2. Psychosocial Health and Wellbeing:

   1. Child internalizing and externalizing behaviors; self-esteem (only for older children); child depressive symptoms (only for older children);
   2. Caregiver's self-esteem; symptoms of mental disorders; quality of marriage; parental stress; loneliness

3. Child Protection and Safety

   1. Child attitudes towards physical punishment (only for older children); experiences of physical, emotional, sexual abuse, and neglect (only for older children)
   2. Caregiver attitudes towards physical punishment; experiences of physical, emotional, sexual abuse, and neglect

4. Caregiver-Child Relationship

   a. Caregiver's involvement, monitoring and supervision of the child; caregiver's parenting behaviors; disturbances in attachment in young children; parent-child activity scale

5. Social and Community Belonging

   1. Child's social support (only for older children); sense of community belonging (only for older children)
   2. Caregiver's social support; sense of community belonging

6. Education Access, Quality, and Achievement

   1. Child's school enrollment, attendance, progression, and grades (if feasible);
   2. Child's satisfaction and happiness with school (only for older children

Study Partners:

The study is part of the Coordinating Comprehensive Care for Children (4Children), which is a five-year United States Agency for International Development (USAID)-funded consortium of organizations led by Catholic Relief Services (CRS) with partners IntraHealth, Maestral, Pact, Plan International, and Westat. 4Children is designed to improve health and wellbeing outcomes for orphans and vulnerable children (OVC). 4Children's Keeping Children in Healthy and Protective Families (KCHPF) project is supported by the Displaced Children and Orphans Fund (DCOF). It focuses on strengthening family care among households where there is a high risk of child separation, or where children can be reintegrated into family-based care after having been placed in residential care. Westat is leading the data collection effort, in close collaboration with our partner, the Department of Social Work and Social Administration at Makerere University, which will serve as the local research partner. Westat also lead protocol development and will be responsible for sampling, data management, data analysis, and preparation of the final report and publication. The study protocol was approved by Mildmay, on May 17th, 2017.

Site Selection:

All RCFs known to district authorities in the districts of Greater Masaka, Mpigi, and Mukono were evaluated for inclusion in the study. These three districts were selected for the study given their proximity to Kampala and the fact that these districts do not have currently active, externally-funded family reintegration programs. Additionally, the Ministry of Gender, Labor and Social Development (MGLSD) and the authorities in these districts (including the Probation and Social Welfare Officers [PSWOs]) have indicated their support of the KCHPF project.

Enrollment Procedures:

A team of case managers affiliated with the KCHPF project will review the child records at each of the participating RCFs and complete bio-data forms on all of the children. This form will include information on the child's background, identity of the child's birth family or next of kin, as well as the family's location, circumstances and ability to be reunified with the child. Using this information, the case managers will conduct a rapid assessment that will allow KCHPF team to determine which children are potentially eligible for inclusion, pending further follow up with and assessment of the family. The information, aggregated over the RCFs, will be sent to Westat to help plan sample selection and randomization. Case managers will conduct family tracing efforts on potentially eligible children to assess the family's capacity and suitability for reunification. Those who meet the eligibility criteria and agree to being reunified by the KCHPF project will be included in the sampling frame.

All enrolled child-caregiver pairs will be randomized to the comparison or intervention arms of the study. The components of the reintegration packages provided in each arms are described below.

'Esanyu Mu Maka' will be delivered on a household level. The primary caregiver will be required to attend all sessions. In addition, all adults who participate in caring for the reintegrated child and ensuring her/his wellbeing will be invited to participate in the program. This will include neighbors if the neighbors also share in the caregiving of the child. Each visit will be about an hour long and will be participatory with discussions regarding home practice activities during the previous week, working through illustrated stories together, role-playing new parenting skills, and assignment of home practice for the following session. Caregiver and family participation in the parenting sessions will be documented in the parenting M&E system and may be used in the analysis. The 'Esanyu Mu Maka' curriculum will cover the list of topic shown below. However, given the broad age range of children being reintegrated (1 to 13 years), the curriculum will be customized to address the specific needs of caregivers of children 1 to 3 years, 4 to 7 years, and 8 to 13 years.

The qualitative data component will provide additional contextual data on the satisfaction and feasibility of the intervention components for children and caregivers as well as the program implementers (parenting facilitators and case managers), and the effect of the individual intervention components on the reintegration process from the perspectives of the target audience and implementers. These data will also help provide insight into local values or concepts, barriers and facilitators to implementing what was learned in the parenting program, and an understanding of the elements that participants value in terms of reintegration. Data will be collected from focus groups and child-friendly individual activity-based approaches with reunified children from both comparison and intervention arms.

Data Management: Quantitative data will be collected by trained interviewers on Google Nexus tablets using Open Data Kit (ODK) software. Tablets will be password protected with encrypted hard drives. At the end of each data collection day, the data on each tablet will be transferred to Westat via a secure file transfer protocol (FTP).

Data Analysis A primary objective of the study is to determine whether there is an association between the parenting intervention and the six domains of reintegration among reunified children and their caregivers. The outcomes across the six outcome domains of reintegration will be summarized using mean, median, and standard deviation for continuous measures, and frequency and percentage for categorical measures, both within and across the two groups. The investigators will then conduct Analysis of Variance (ANOVA) to examine effects of participation in the parenting program on the outcome variables. In addition, based on preliminary analysis, the investigators may select covariates, and conduct Analysis of Covariance (ANCOVA). For outcome variables that are categorical, Chi-square tests will be conducted. To examine the changes over time within and across the groups, the investigators plan to conduct Generalized Linear Mixed Models (GLMM) with repeated measures at multiple time points.

For the main analysis to assess whether participation in a household-based parenting program results in higher rates of reintegration, the investigators will consider several options for the main analysis. One possibility will be to develop a scale using factor analysis to analyze the various domains that constitute reintegration. Exploratory Factor Analysis (EFA) will help us identify items that measure reintegration success conceptually, and have strong contributions to specific constructs. The investigators will use information based on EFA as well as the known conceptual framework for these domains to decide the number of factors and items in each factor. Confirmatory Factor Analysis (CFA) will then be performed to validate the factor structure developed based on findings from the EFA. Another possibility will be to conduct multivariate GLMM to form composites of the outcome variables that can be analyzed together. That is, variables within each dimension and across the dimension can be conceptualized, categorized and then entered as dependent variables in the multivariate GLMM analysis to examine the effect of the intervention on outcomes over time. The investigators will also use GLMM to examine whether the improvement in outcomes over time in the intervention group are significantly greater than the improvement in outcomes in the comparison group.

Inclusion Criteria:

• The case assessment indicates that the child can be reunified with the family/kin.
• Child is being reintegrated with family or kin in one of the study districts.

Exclusion Criteria:

• Child requires an alternative care arrangement including foster care, adoption, or independent/group home living.
• Child has severe developmental disabilities.
• Children whose caregivers plan to send them boarding school

Randomized assignment will be done by the individuals who are evaluating the study outcomes, but will be blind to participant assignment