Randomized Controlled Trial of Ways to Improve OVC HIV Prevention and Well-being
Randomized Controlled Trial of Ways to Improve OVC HIV Prevention and Well-being
The purpose of this study is to compare the effectiveness of two types of counseling, Psychosocial Counseling (PC) and Trauma-Focused Cognitive Behavioral Therapy (TF-CBT), in addressing outcomes of orphans and vulnerable children (OVC) including mental and behavioral health, well-being, social support, and HIV risk behaviors. The study will be conducted in Lusaka, Zambia.
This is a randomized controlled trial to evaluate the effectiveness of TF-CBT compared to PC in improving a range of OVC outcomes including HIV risk behaviors but also mediating factors which constitute important outcomes - well-being, mental health, functioning, and caregiver health and support. The study design also includes measurement of the impact of factors that may moderate effects - age, gender, religion, ethnicity, HIV knowledge and attitudes, education, and physical health. This is an effectiveness trial in that the interventions will be provided by Ministry of Health (MoH) and organizations already providing services to OVC.
Assessments: Study instruments will be administered using an Audio Computer Assisted Self-Interviewing (ACASI) system developed by Tufts University. The Tufts ACASI system allows the responder control to navigate and answer sensitive questions privately and discreetly. It was developed using Macromedia Authorware (v7.0) and can be programmed in any language by having a native speaker record the screen text. The Tufts ACASI system offers both live interview and data entry modes and can deliver a variety of question types, including yes/no, single-answer multiple choice (select one answer), and multiple-answer multiple choice (select all that apply) questions which allow numeric keypad entry, as well as scale questions which allow the user to click on a point along a scale, and text entry for interviewer-administered, open-ended questions
Procedure: Study participants will be recruited via several methods. The first method of recruitment is from urban communities through outreach by existing Home Based Care Workers (HBCWs) and other peer educators under the Archdiocese of Lusaka, local churches, and other local organizations that have outreach activities in the community. The HBCWs and peer educators will introduce the study to families in the community who they believe may be eligible for the study. Investigators will also ask these peer educators and HBCWs to use a recruitment script to introduce the study to both the adolescent and his/her caretaker. If the adolescent and/or caretaker are not interested in hearing more about the study, their information will not be passed on to study staff and we will have no contact with that family. The HBCW will make follow-up appointments with families who are interested in hearing more about the study and arrange a time for them to go to a local parish. At these follow-up appointments one of the study assessors will discuss the study in detail with the potential participants (adolescent and caretaker). This is preferable to clinic-based recruitment because, 1) many OVC do not come to clinics for reasons such as stigma, fear, mental health problems or transport problems 2) HBCWs will assist with retention, and 3) study staff will have no contact with any families who have not already indicated to their HBCW that they would like to hear more about the study
As a second recruitment strategy, investigators will also hold general community meetings in each of the study sites in which all members of the community will be invited. Investigators will explain the general purpose of the study at these meetings and answer any questions. Investigators will invite families to contact us directly if they are interested in hearing more about the study from one of our study assessors or to talk with their HBCW who can put the family in touch with our research staff. The community will also be given specific dates when the study team will be at a central parish where they may come to be assessed and/or ask additional questions.
Third, we may use a school-based recruitment. For this strategy investigators will first hold meetings with school officials (e.g., Principals, head teachers) in schools within our catchment area in Lusaka. Investigators will also gain approval from the Ministry of Education of Zambia if needed. Investigators will also hold general meetings at the school open to all staff, students, and parents to let them know about the study. If the school officials agree to have recruitment in the school, investigators will ask that they introduce guidance counselors. Investigators will train the guidance counselors in the school, who have knowledge of which students in the school might be eligible for the study, on the purpose of the study and how to introduce the study to families that they believe would be eligible. Meet with head teacher, explain study. Also talks to other staff members. Guidance counselor would meet with the parents and child to see if they are interested. Additionally, we will provide our study contact information at the general school meeting so that interested families may contact us directly. Investigators will have study staff on-site on selected days to meet privately with interested families for explaining the study in full, obtaining informed consent, and proceeding with study activities.
Fourth, investigators will use a snowball sampling strategy. A large number of adolescents and caregivers enrolled in the study have informed us that they know other families in their communities they think would benefit from mental health and psychosocial services. Investigators plan to have counselors ask current study participants if they know of specific families who they think would be interested in hearing more about the study. If participants know other families who they think would be interested, investigators will ask the participants to bring those families to meet with the counselor. The counselor will then give a brief introduction to the study. The counselor will then refer families who are still interested to the research team and a day/time will be setup for the family to privately meet with an assessor for consenting and screening as per the original study protocol.
The assessors will be at the parish locations, or if needed would go with an HBCW (if an HBCW was involved in the recruitment) to the families' homes that indicated interest in the study. The assessor will conduct the adult consent and caretaker permission process with the caretaker first. Consent and permission will be obtained at this time for all parts of the study-- screening, interview using ACASI, and intervention. In all cases, the caretaker will have to provide consent to participate in the interview and intervention for him/herself and will also have to provide permission for the adolescent to participate in the screening, interview, and intervention. For these cases, in the response section the assessor will indicate if the caretaker consents to participate in each of the sections (screening/interview and intervention) separately and gives permission for their adolescent to participate in each of the 3 sections (screener, interview and intervention) separately.
If the caregiver and child respond yes to the consent and the caretaker provides permission for the adolescent, the assessor will administer the screener. This screener will be scored immediately and if the child is determined eligible for the study, the assessor will ask if they are able to complete the full assessment battery at this time (approximately 2 hours) or if they would like to schedule a separate time to complete this. If a separate time/date is preferred, the assessor will schedule that with the family. The HBCW will not be present during the consent or assessment process.
If the adolescent is found to be eligible after the screening interview (through the results of either the adolescent screening, caretaker screening, or both), then the assessor will allocate a study ID number to the adolescent. When the baseline is designated to be completed, the assessor will set up two notebook computers for the baseline interview: one for the adolescent and one for the caretaker. The adolescent and caretaker will complete the interview simultaneously but separately (in separate rooms), with the assessor not present but close by in case of problems or questions. On completion of the baseline interview the assessor will let the family know that someone from their nearest MoH clinic will contact them.
At the end of each day all assessors will report the allocated study IDs to the on-site Study Director, who will send them to a Johns Hopkins-based research staff who will retain the confidential master list linking study IDs to interventions based on random assignment. The Johns Hopkins researcher will respond with the allocations for those study IDs. These data will be used to inform the clinics which intervention each new participant will receive prior to their first visit. Both interventions will be available at all study sites. Neither assessor, caretaker, nor adolescent will know which intervention the adolescent will receive.
Whether or not an adolescent agrees to join the study s/he will have access to other services under the Archdiocese of Lusaka. These services include Psychosocial Counseling, Home Based Care services, medical and education support, voluntary counseling and testing (VCT), Community Sensitization Campaigns, Behavioral Change, livelihood support, tuberculosis care, awareness programs on communicable diseases. The adolescent will also have access to the HIV prevention strategies available at the MoH clinics including, VCT, counseling, safe sex interventions and primary care services. For those in the study, we will monitor utilization of these services and include that information as a variable in the analysis, to determine their effects as potential moderators of treatment effects.
The adolescent will either go to a MoH clinic or a different, pre-identified community center for their first appointment with their counselor to receive either TF-CBT or PC. Later sessions will either be at the clinic or at another location that the counselor and family agree on, and locations can vary from week to week if necessary and feasible. The caretaker may also participate in these sessions (if they have consented to participate in the study along with the adolescent). Because the Archdiocese of Lusaka sites are deeply embedded into the community, other neighborhood structures-such as churches-may be used as meeting points.
Follow-up Assessments with both adolescent and caretaker will be done at 0 and 6 months after completion of the interventions. The 6 month time point is considered the primary time point of interest. If one treatment shows substantially greater efficacy at 6 months it will be offered to those that received the other intervention and a 12 month assessment will be done only with the group originally allocated to the more effective intervention.
We expect to have 3-4 study contacts with research participants and their caregivers:
The outcomes for Specific Aim 1 are indicators of HIV risk behaviors. This includes questions about social circles engaging in risky sexual behaviors such as multiple sexual partners, having sex without a condom, sexually transmitted infections (STIs) and substance use (HIV risk assessment screening tool). The WAF instrument will also assess broader HIV risk indicators such as knowledge, attitudes, self-efficacy and sexual practices. The outcomes for Specific Aim 2 are the scores on various mental health and well-being tests, including Child PTSD Symptom Scale (CPSS), Child Behavior Checklist (CBCL), self-perception, Achenbach Youth Self Report, ASSIST (substance use), and functionality.
All primary analyses will be based on an intent-to-treat approach. For all continuous outcomes, linear mixed models will model each outcome over time by incorporating a random intercept term and possibly a random slope term to account for within subject correlation on repeated measures. The test scores at completion of intervention and at 6 and 12 months after completion will be modeled separately using treatment arm, baseline symptom score, time, and an interaction between each treatment arm and time as fixed covariates in each model. This analysis will allow estimation of the mean difference in each symptomatology score between the two treatment arms at each time, adjusting for baseline score if needed. Mean differences and 95% confidence intervals will be calculated and presented. For dichotomous variables, a generalized linear mixed effects models will be used. A risk ratio and 95% confidence interval comparing each binary outcome between the two arms at each time will be reported. Potential modifiers and mediators will be handled as described below.
Moderator and mediator analyses: The study will evaluate the impact of hypothesized moderators (child age, gender, religion, ethnicity, HIV knowledge and attitudes, family formation, education, physical health) on treatment response. We have also hypothesized that changes in the course of therapy in caregiver levels of distress and support and changes in the child's mental health and self-perception will mediate the impact of treatment on children's outcomes.
All of the significant moderating and mediating variables then will be simultaneously entered into multiple regression analyses with respect to the individual outcome measures to ascertain which variables contribute unique variance to estimating the outcome variables. Significant moderators and mediators will then be tested for specific treatment effects using multiple linear-regression analyses. Multiple regression techniques will examine the pattern of relationships between treatment, hypothesized mediating influences and outcomes. Using the steps suggested by Baron and Kenny, we will examine whether a) each treatment arm significantly predicts mediator; b) mediator significantly predicts symptom outcome; and c) whether the significant relationship between each treatment condition and symptom outcome decreases significantly when the change in mediator is entered into the regression model.
In addition to primary analyses on the clinical effectiveness of the interventions, we will also analyze cost effectiveness and implementation challenges.
Service use, employment-related and caregiver data will be collected for a retrospective period of 3 months at baseline, and at 1 month, 6 months and 12 months post-treatment using a version of the Client Service Receipt Inventory adapted for OVC. Service use and carer inputs will be turned into cost measures by applying local and national unit cost values.
We will also monitor and evaluate system implementation factors, including acceptability, adoption, appropriateness, feasibility, fidelity, penetration and sustainability from the perspective of multiple stakeholders including, providers, organizations and policy leaders. Semi-structured interviews will be conducted with these stakeholders to examine perspectives on the barriers and facilitators of implementation.
Inclusion Criteria:
Exclusion Criteria: