Background: Asset-based economic empowerment interventions, which take an integrated approach to building human, social, and economic capital, have shown promise in their ability to reduce HIV risk for young people, including adolescent girls, in sub-Saharan Africa. Similarly, community and family strengthening interventions have proven beneficial in addressing mental health and behavioral challenges of adolescents transitioning to adulthood. Yet, few programs aimed at addressing sexual risk have applied combination interventions to address economic stability and mental health within the traditional framework of health education and HIV counseling/testing. This paper describes a study protocol for a 5-year, NIMH-funded, cluster randomized-controlled trial to evaluate a combination intervention aimed at reducing HIV risk among adolescent girls in Uganda. The intervention, titled Suubi4Her, combines savings-led economic empowerment through youth development accounts (YDA) with an innovative family strengthening component delivered via Multiple Family Groups (MFG). Methods: Suubi4Her will be evaluated via a three-arm cluster randomized-controlled trial design (YDA only, YDA + MFG, Usual Care) in 42 secondary schools in the Central region of Uganda, targeting a total of 1260 girls (ages 15–17 at enrollment). Assessments will occur at baseline, 12, 24, and 36 months. This study addresses two primary outcomes: 1) change in HIV risk behavior and 2) change in mental health functioning. Secondary exploratory outcomes include HIV and STI incidence, pregnancy, educational attainment, financial savings behavior, gender attitudes, and self-esteem. For potential scale-up, cost effectiveness analysis will be employed to compare the relative costs and outcomes associated with each study arm. (Continued on next page) * Correspondence: email@example.com; firstname.lastname@example.org Brown School, Washington University in St. Louis, 1 Brookings Drive, St. Louis, MO 63130, USA Columbia University School of Social Work, 1255 Amsterdam Ave., New York, NY 10027, USA Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Ssewamala et al. BMC Public Health (2018) 18:693 Page 2 of 11 (Continued from previous page) Conclusions: Suubi4Her will be one of the first prospective studies to examine the impact and cost of a combination intervention integrating economic and social components to reduce known HIV risk factors and improve mental health functioning among adolescent girls, while concurrently exploring mental health as a mediator in HIV risk reduction. The findings will illuminate the pathways that connect economic needs, mental health, family support, and HIV risk. If successful, the results will promote holistic HIV prevention strategies to reduce risk among adolescent girls in Uganda and potentially the broader sub-Saharan Africa region. Trial registration: Clinical Trials NCT03307226 (Registered: 10/11/17). Keywords: HIV, Adolescent girls, Assets, Economic empowerment, Family strengthening, Combination interventions Background connections and more open communication between a Sub-Saharan Africa (SSA) remains the world’smost af- child and his/her primary caregiver can predict mental fected region in the HIV epidemic, home to 71% of health outcomes, delays in onset of sex, and overall people living with HIV worldwide  and girls account- child adjustment [26–31]. Additionally, better parenting ing for 7 out of 10 new infections among adolescents skills have been associated with adolescents having less (ages 15–19 years) . This gender disparity has in- susceptibility to peer pressure [31, 32]. Thus, particu- creased recognition that adolescent girls need more atten- larly in low-resource settings, supporting families with tion if the global community is to achieve an AIDS-free economic opportunities and strengthening family support- generation. Studies and theory suggest causal pathways ive processes may minimize risk taking behaviors, discour- between family economic resources, education, mental age school separation, and address mental health stressors health, and HIV risk [3, 4]. Being out-of-school is one of among adolescents. the key characteristics found to increase young women’s Given the complex and multi-dimensional drivers of vulnerability to HIV as it is associated with numerous risk increased HIV risk among adolescent girls in SSA and the factors, including age-disparate and transactional sex, failure of most single interventions to significantly decrease early marriage, inconsistent condom use, and limited these rates, investments in combination interventions are power in relationships – most significantly the ability critical to provide an interdisciplinary, multi-level response to negotiate safe sex [5–11]. Alongside these risks exist needed to reduce new HIV and STI infections in a way that mental health challenges associated with economically single interventions alone have not yet been able. To motivated sex (both age-disparate and transactional), address the identified challenges, this trial, entitled “Suubi4- which have been shown to have a bi-directional rela- Her” (also known as Hope for Girls),aimstoexamine the tionship with depression, low self- esteem, and anxiety impact of and cost associated with an innovative combin- for young women [12, 13]. Moreover, higher depression ation intervention on HIV risk behaviors among adolescent among young females has been associated with girls. The intervention is composed of two programs: 1) co-factors of HIV risk . Given the heightened risk Asset-based financial inclusion (specifically youth develop- for HIV infection in adolescent girls, there is an urgent ment accounts – (YDAs), and 2) An evidence-based, family need to address the complex and multilayered economic strengthening approach to enhancing youth behavioral and psychosocial issues facing this population in SSA. health delivered via multiple family groups (MFG). In many SSA countries, including Uganda, access to Designed within a prevention framework, the Suubi4Her education remains strongly associated with household intervention seeks to support vulnerable adolescent girls economic stability . Lack of financial resources is the before they drop out of school when their exposure to most commonly cited reason why adolescent girls fail to HIV-risk taking behaviors increases. The integration of attend school [16–19]. Moreover, cultural norms can be efforts towards strengthened family relationships and influential and families may feel pressure to prioritize male improved communication among household members may education when resources are few. Several traditions in be crucial in helping families navigate adolescent transi- SSA are passed down generationally and encourage strati- tions to adulthood. Suubi4Her pairs two innovative and fication of gender roles, such as adolescent marriage and evidence-based interventions together, an asset-based early childbearing, both of which can prompt separation financial inclusion/economic strengthening model and from school for adolescent girls [20–22]. a family strengthening approach to enhancing youth At the same time, family economic stability influences behavioral health delivered via MFGs, recognizing the the quality of family relationships with poverty adversely possibility that mental health may be a critical compo- impacting parent-child communication and involvement nent intersecting between poverty and HIV risk for [23–25]. Studies have documented that strong positive young females. Ssewamala et al. BMC Public Health (2018) 18:693 Page 3 of 11 Methods R34MH081763), schools and the local district adminis- Study aims tration will be relied upon to identify participants and The study aims are to: 1) Examine whether the Suubi4- help with recruitment. Existing school enrollment proce- Her intervention is effective in protecting adolescent dures will be leveraged to invite caregivers with an eli- girls against known HIV risk factors; 2) Elucidate the ef- gible child to contact the school for further details. In fects of the Suubi4Her intervention on mental health addition, community development officers and imple- functioning and examine the effects of these variables as menting partners will distribute flyers during their fre- potential mechanisms of change, mediating the relation- quent community visits to inform caregivers ship between each intervention and HIV risk reduction; whose children meet the inclusion criteria but may not and 3) Evaluate the cost-effectiveness of each interven- yet have reported back to school. Adolescents and care- tion condition. givers who indicate interest will be invited to come to the school in-person for a one-on-one information meet- Setting ing with the research team, during which they will be Over the course of five years (August 2017 – July given details of the project. 2022), Suubi4Her will be implemented in the Central The project will take place in a highly stable region of Region of Uganda, an area heavily affected by HIV with Uganda where geographical moves are rare, facilitating aprevalencerateof10.6%,three percentage points ability to track and retain the sample. We will ask partic- higher than the national average of 7.4% . In 2013, ipants to give the telephone number, names, addresses, the number of new HIV infections for adolescents (ages and contact information for three people who will al- 15–19) in Uganda was twice as high for girls (est. ways know how to reach them. Participants will be told 10,000) as it was for boys (est. 4500) . With data that if we contact the people listed, we will not discuss from 2014, UNAIDS reported that 570 adolescent girls any details about them or their study participation. We and young women (ages 15–24) were acquiring HIV will use these records to help track their location only if per week in Uganda . we have lost contact. We will also be in contact with all participants regularly during school roll calls to deter- Study population, recruitment, and retention mine enrollment and attendance. Moreover, we will have A total of 1260 secondary school-going girls (ages 15–17 monthly contact with children in the two treatment con- at enrollment) in their first year of secondary school will ditions through the distribution of their monthly savings be enrolled and followed for four years. Given the high banks statements. This frequent contact will enable the incidence of HIV in the study area and to avoid stigma research team to continually engage all participants, and that surrounds being HIV positive in the region, no girl minimizing loss to follow-up. Based on prior studies in will be excluded by virtue of her HIV status. Given the the same region, we expect attrition by end of follow-up high prevalence of HIV in our proposed study sample, to be no more than 17%. to further ensure that HIV serostatus is balanced across interventions, HIV status will be used as a blocking Intervention description factor during the cluster randomization process using The study has been designed as a three-arm, cluster the strategy outlined in Bellamy et al. . randomized-controlled trial (RCT), consisting of two Adolescents will be included within the study if they treatment arms and one control arm (Fig. 1). meet the following criteria: 1) female; 2) enrolled in first year of secondary school in Rakai, Kyotera, Masaka, Treatment arm 1 - YDA Lwengo or Kalungu districts; 3) age 15–17 years; 4) living YDAs are guided by asset-theory [3, 4] which posits that within a family (broadly defined and not an institution or financial and tangible assets not only impact economic orphanage, as those in institutions have different familial stability for individuals and households but also have needs). Exclusion criteria: Girls are ineligible if: 1) they important developmental and psychological benefits in- have a severe cognitive or severe psychiatric impairment cluding future-oriented thinking, feelings of security, and that would prevent comprehension of study procedures as self-efficacy. The argument advanced by asset-theory is assessed by trained staff during the Informed Consent consistent with several behavioral and psychosocial theor- process; or 2) they are unwilling or unable to commit to ies that have guided studies on sexual risk-taking and completing the study. We will not exclude girls because of mental health, including Bandura’s Social Cognitive their HIV, STI and/or pregnancy status. Analysis will be Theory  and the Theory of Reasoned Action . adjusted to account for these baseline factors. Adolescents in both treatment arms will be enrolled in Using the same recruitment procedures tested in three a 1:1 matched savings program at a national financial earlier studies (Suubi+Adherence R01HD074949, Brid- institution. Each account will be in the name of the ado- ges to the Future R01HD070727 and Suubi Maka lescent, with her primary caregiver as a co-signer, until Ssewamala et al. BMC Public Health (2018) 18:693 Page 4 of 11 Fig. 1 Suubi4Her Study Design she turns 18 years of age, at which time a co-signer will and children to communicate in a safe setting with no longer be required. This is consistent with the Ugan- other families who have shared experiences thus allow- dan banking law which prohibits children below age 18 ing each family to benefit from the contributions of one from independently entering into a binding contract and another . Advice and insight from other families is operating a bank account. The matching funds will be often seen as less threatening than feedback given by a kept in a separate account from the participants’ own therapist . In addition, MFG focuses on reducing savings. When a girl is ready to pay for school fees, the stigma by normalizing shared experiences. The MFG check for the matching funds will be written in the intervention acknowledges poverty as a stressor that name of the school she attends or directly wired to the may undermine parenting while also recognizing the school’s bank account. This process is intended to elim- contextual challenges that contribute to poor mental inate the risk of family pressure on the girl to withdraw health functioning for adolescent girls, including high money set aside for education and skills training. rates of poverty, violence, and family loss due to HIV Participating students will be allowed to use up to 30% and other health threats [41–43]. The MFG approach of their total matched savings to invest in a family-based will allow girls and their families to share their experi- income-generating activity (IGA). The remaining 70% of ences with others in similar situations, thus building the savings will be restricted to fund the education and hope by providing social support, normalization of skills training of participating adolescent girls. Consistent similar experiences and struggles, and the sharing of ef- with earlier studies, a participant’s access to the matching fective solutions . Suubi4Her will utilize the MFG funds is conditional upon completion of 12 financial man- approach to specifically target family communication, agement workshops over 12 months. The workshops, to support, and gender equality both within and across be implemented by collaborating community agency, families from the same community setting. Reach the Youth-Uganda, in collaboration with the finan- cial institutions holding the YDAs, will consist of 12 gen- Control arm – Usual care eral workshops that cover basic principles of financial In Uganda, an Adolescent Sexual and Reproductive management including income generation, use of financial Health curriculum is required of all secondary school stu- institutions, saving, and asset-building. dents . As such, these curricula will be considered usual care and all enrolled participants, both in control Treatment arm 2 – Combination intervention (YDA + MFG) and treatment conditions, will be exposed. The Adoles- The combined treatment arm will consist of a YDA cent Sexual Reproductive Health content is dispersed (detailed above) and a family-based dialogue and training across a range of academic subjects in secondary schools. delivered via MFGs that aim to strengthen family relation- In each class, students receive information about sexual ships and address mental health challenges that com- activity, HIV prevention, and gender studies relevant to monly occur in adolescence. that subject. Teachers and students all receive a sex The MFG has adopted the strengths of multiple and health education handbook. The content related to therapeutic methods and theories to create an ex- HIV and sexual risk-taking behaviors includes delaying tremely flexible approach which has been applied to a sex; using condoms and contraception; preventing variety of target populations struggling with a diverse forced sex; and preventing substance abuse. This range of issues [38–40]. MFG is based on building sup- curriculum also includes education on gender equality port for families by providing opportunities for parents and importance of delayed marriage. Prior to study Ssewamala et al. BMC Public Health (2018) 18:693 Page 5 of 11 implementation, the research team will hold induction across the three experimental groups. Each of the 42 meetings for all teachers involved in the study to ensure schools will be randomly assigned to one of the three uniform delivery of the Ministry of Education approved study conditions such that all selected adolescent girls sex education curriculum. from a particular school will receive the same interven- tion (to reduce contamination). In sum, of the 42 sec- Randomization, sample size, and power analysis ondary schools, 14 will be randomly assigned to receive Stratified random sampling will be used to allocate a YDA (n = 420 students), 14 to combination interven- schools to four strata based on two variables: 1) student tion: MFG + YDA (n = 420 students), and the final 14 to population size (medium size vs. large), and 2) geo- the usual care intervention comprising of standard graphical location (rural vs. urban), to ensure balance health and sex education provided in secondary schools on those variables. The restricted randomization tech- (n = 420 students) (Fig. 2). nique of Hayes and Moulton  will be implemented Power analyses were generated using the two-group within thefourstratatoassureoverall school balance repeated measures modules in NCSS PASS 14 to Fig. 2 CONSORT Flow Diagram – Suubi4Her Ssewamala et al. BMC Public Health (2018) 18:693 Page 6 of 11 compute minimum detectable effect sizes for the pri- Research Assistants will contact girls who test positive mary analysis. The study will begin with 1260 partici- 30 days after telling them of their HIV status, to confirm pants from 42 schools evenly assigned to the three that they have connected with follow up care and have ac- study groups. Conservatively assuming 20% attrition, cess to comprehensive care services. Girls testing positive data from 1008 participants will be available for analysis for HIV, STIs (Gonorrhea, Trichomonas, or Chlamydia) at all time points. Given anticipated within-school cor- or pregnancy will be referred for care and support. All relations, theeffectivesamplesize(ESS) waslowered to study procedures were approved by the Washington be ESS = N/DEFF, where DEFF is the design effect or University in St. Louis Review Board (IRB) – home institu- variance inflation attributable to using correlated data. tion of the PI – and by in-country local IRBs in Uganda: Conservatively assuming DEFF = 4.00, ESS = 1008/4 = 252, Uganda Virus Research Institute (UVRI), and Uganda division of 252 by 3 study groups yielded 84 per study National Council of Science and Technology (UNCST). arm. Assuming α = .017, power = .80, and ESS = 168, we computed for binary STI and risk outcome variables used Quantitative assessment and analytic plan to evaluate the minimum detectable (inverted) odds ratio Assessments will occur at baseline, 12-, 24-, and (OR), proportion difference (pdiff), and standardized pro- 36-months (Fig. 2) and will take place at the partici- portion difference (h) for the proposed time-averaged pant’s home or at school (in a private room) with each 2-group paired comparisons for outcome base rates ran- lasting about 60 min. Interviews will be conducted in ging from 10 to 50%. For continuous mental health out- English or Luganda depending on participants’ English comes, we computed the standardized mean difference d. proficiency. All interviewers will be fluent in both lan- Since within-subject correlations among the various out- guages. A list of standardized instruments that will be in- come measurements following intervention are unknown, cluded in the main statistical analyses are outlined in the within-subject correlation ρ was varied between .20 Table 1. All measures used have been or will be pre-tested and .70. Effect size estimates for our primary analyses and made culturally appropriate to the Ugandan context. range from .33 to .45, which are between benchmark For questions measuring sensitive behaviors, computer thresholds of .20 and .50 for small and medium standard- assisted self-interviews will be implemented. Non-sensitive ized effect sizes , suggesting that our proposed pri- questions will be interviewer-administered using Qualtrics. mary analyses have sufficient power to detect small to The two primary outcome measures for the study are: medium effects across a variety of analysis conditions. 1) change in HIV risk behavior (occurrence of risky sexual behaviors as measured by Timeline Follow Back (TLFB) Ethics and consent for Specific Aim 1; and 2) change in mental health func- Written informed assent and consent will be obtained tioning (continuous variables of depression, self esteem, from the adolescent participant and their caregiver, re- and hopelessness as measured by Beck’sHopelessness spectively. The processes for adult caregivers and adoles- Scale and adapted versions of the Child Depression Inven- cents will be separate to avoid any coercion. Consent tory and Tennessee Self-Concept Scales, all of which have and assent forms will be translated and back-translated been pre-tested in earlier Suubi trials, demonstrating between English and Luganda. While adolescent partici- strong internal consistency [20, 23–25, 39]) (Table 1)for pants will be English speaking, some may be more com- Specific Aim 2. Secondary exploratory outcomes include fortable using Luganda. Therefore, the assent process HIV and STI incidence (including Gonorrhea, Trichomo- will be conducted in English or Luganda depending on nas, and Chlamydia), pregnancy incidence, educational at- participants’ English proficiency. Interviewers will make tainment, financial savings behavior, gender attitudes, and it explicitly known to the participants that they may self-esteem. All study tools were translated into Luganda refuse to answer a question or decline to undergo a and back-translated into English. procedure, at any time. The study team will receive training on Good Clinical Practice (GCP) so that sensi- Primary analyses tive topics and issues are handled appropriately. Add- An intent to treat (ITT) approach will be employed. itionally, for questions measuring sensitive behaviors Prior to initial analyses, the study team will determine like sexual risk-taking and mental health, we will use collectively whether to estimate marginal population computer-assisted self-interviews where the participant average effects versus conditional subject-specific effects. takes the survey herself on a mini laptop/ipad, provid- Interest in population average vs. subject-specific effects ing additional privacy and confidentiality. and the relative importance of explicitly estimating In line with the Uganda HIV disclosure policy, study school-level and person-level variance components will participants will receive their HIV results independent be considered during these discussions. of their caregiver. Upon receipt of the child’s consent to For marginal effects estimation, generalized estimating do so, a parent or guardian will then be informed. equations (GEE) will be used to perform the proposed Ssewamala et al. BMC Public Health (2018) 18:693 Page 7 of 11 Table 1 Suubi4Her Instruments Variable Measurement Reliability Time point Demographics Age; orphan status (single vs. double); socioeconomic status; Socio-demographic questionnaire n/a B, 12, 24, 36 family composition/structure; caregiver educational level Moderators Rural/urban/semi-urban; exposure to outside HIV/STI-related B, 12, 24, 36 programs; economic/household income; asset accumulation Primary Outcomes HIV/Sexual risk taking behavior Time-line Follow Back (TFLB)  n/a B, 12, 24, 36 Mental Health Functioning Beck Hopelessness Scale  0.79 B, 12, 24, 36 Adapted Tennessee Self-Concept Scale  0.81 Adapted Child Depression Inventory  0.65 Secondary Exploratory Outcomes & Potential Mechanisms of Change STI Biomedical data: Gonorrhea, Chlamydia, Trichomonas, n/a B, 12, 24, 36 a a a a HIV, Viral load , CD4 count , Pregnancy Incidence Biomedical data: HIV, Viral load , CD4 count , n/a B, 12, 24, 36 Pregnancy test Self-efficacy Adapted Tennessee Self-Concept Scale (TSC-2)  0.81 B, 12, 24, 36 Education plans/aspirations Adapted Monitoring the Future Survey  n/a B, 12, 24, 36 Motivation to participate Questions tested in previous studies [20, 23–25] n/a B, 12, 24, 36 Family Support Social Support Behaviors Scale (SS-B)  0.77 B, 12, 24, 36 Family Cohesions Scale , Krauss Interview  0.69 Parent Child Relationship Inventory (PCRI)  0.91 Family Stability Socio-demographic questionnaire n/a B, 12, 24, 36 Self-esteem Rosenberg Self-Esteem Scale  0.77–0.88 B, 12, 24, 36 Attitudes towards gender roles; Decision-making; Gender Norm Attitudes Scale  0.67–0.70 B, 12, 24, 36 Communication HIV/STI knowledge HIV/STI knowledge 0.80 B, 12, 24, 36 Condom Negotiation Self-Efficacy Condom negotiation self-efficacy scale  0.80 B, 12, 24, 36 Sexual Communication Skills Sexual Communication Scale  0.80 B, 12, 24, 36 Social Support MSPSS  0.84 B, 12, 24, 36 Savings Deposits Bank statements n/a B, 12, 24, 36 Financial Literacy Financial Literacy knowledge  0.80 B, 12, 24, 36 Access to services RBA Services  .66–.83 B, 12, 24, 36 Cost of staff time, supplies, overhead for YDA and for MFG Project records; Admin. review n/a ongoing For participants testing positive for HIV primary analyses. The alternating logistic regression three-level models containing random intercepts for (ALR) approach implemented in SAS PROC GENMOD schools and random intercepts and slopes for girls will will be employed to address the 3-level clustering of be considered with cases with partial outcome data auto- observations within participants and participants within matically included under the missing-at-random (MAR) schools. Multiple imputation (MI) will be used prior to assumption. Alternative covariance structures may be fitting GEEs to include cases with partial data under the considered to facilitate convergence and improve model missing-at-random (MAR) assumption. The QIC statis- fit. Covariance structures will be compared using tic will be used to select an optimal working correlation information-theoretic criteria such as the BIC to select structure for GEEs. the optimal covariance structure. For conditional effects estimation, multilevel or similar HIV risk behavior is the primary outcome to address (e.g., latent growth curve) models will be used. Initial Specific Aim 1. Mental Health functioning is the primary Ssewamala et al. BMC Public Health (2018) 18:693 Page 8 of 11 outcome to address Specific Aim 2. We hypothesize that estimates of cost-effectiveness. For outcomes that are following baseline: 1) YDA will have lower odds of HIV measured similarly in other intervention studies, such as risk behavior and higher mean mental health functioning improvements in education or health, we will be able to vs. control participants; 2) YDA + MFG will have lower compare the cost-effectiveness of YDA-alone and Com- odds of HIV risk behavior and higher mean mental health bination intervention (YDA + MFG) to other interven- functioning vs. control participants, and 3) YDA + MFG tions in developing country settings. will have lower odds of HIV risk behavior and higher mean mental high functioning vs. YDA participants. For Discussion each primary outcome, these hypotheses will be tested by There is an increasing interest in and use of social safety three pairwise planned time-averaged comparisons of nets to achieve health outcomes for children. At the post-baseline measurements. Alpha (α) will be set at .05/3 same time, there is need to understand how social inter- = .017 for these three planned comparisons to maintain a ventions can complement these economic programs for nominal α = .05. Any additional post-hoc comparisons enhanced impact. The present study protocol describes (e.g., paired comparisons of groups at each time point) will a three-arm cluster randomized-controlled trial with the maintain nominal alpha of .05 through the use of multiple primary outcomes of analysis as HIV risk and mental comparison adjustment methods (e.g., simulation-based health functioning among school-going adolescent girls. stepdown methods). The study will also provide critical information on cost, comparing budget expenditures and adolescent out- Secondary exploratory analyses comes across study arms, indicating which arm provides We will also describe the proportion of new HIV cases better value for money. The findings will yield important among girls who are HIV-negative at study entry and who insights on the effectiveness of asset-promotion pro- seroconvert during the study. Although our study is not grams and MFG dialogues to reduce HIV risk and formally powered to test for differences in the proportion improve behavioral health while also exploring the ways of HIV incident cases across the three study groups, we in which mental health may serve as a mediator to the will explore the effect of YDA and YDA + MFG on the success of HIV risk reduction programs for adolescent proportion of HIV incident cases observed across the girls in Uganda. post-baseline study period. The same multilevel modeling We do not anticipate any major threats to study im- and GEE approaches described above will be used to com- plementation though we recognize potential concerns pare the odds and mean levels of other secondary out- and have adapted accordingly. We have a stringent re- comes across study groups for binary and continuous tention plan for attendance at the MFG and Financial outcomes, respectively. To explore hypothesized mecha- Management sessions. We expect to achieve enrollment nisms of change, secondary exploratory analyses will also and retention goals (90% at 36-month post intervention, investigate whether mental health constructs at 12 and a statistic based on our current and previous studies 24 months mediate the relationship between intervention among school going children in the same study area – group assignment and HIV risk outcomes at 24 and see Study Population, Recruitment, and Retention). How- 36 months and whether HIV-serostatus moderates these ever, we are aware that older adolescents are more likely associations. Exploratory analyses will also examine the to leave school and migrate for work. As such, we have potential mediating effects of other variables such as fam- put in place detailed tracking procedures including each ily support, self-efficacy, and condom negotiation skills on participant providing names, physical addresses, and primary outcome measures. phone numbers for a minimum of three relatives or friends who would always know the whereabouts of the Cost analysis participant. Moreover, should our recruitment, enroll- Following standard practice of measuring cost-effectiveness ment, or retention rates deviate from anticipated goals, of interventions, we will measure costs on a per person the study team, comprising both the research and field basis. The intervention costs will include all program implementation teams, will determine collectively how costs, including the YDA savings match as well as all costs best to adjust the field outreach activities. incurred for running the program. Research costs will not Another limitation of the proposed study is that the be included. Costs from multiple years will be adjusted YDA intervention may only appeal to families that feel for inflation, depreciation, and discounting. The out- they have expendable resources to contribute to savings comes analyses described above will be used to estimate accounts. Others may see YDAs, which defer consump- how much Combined Intervention (YDA + MFG) vs tion, to be impractical and detrimental to meeting their YDA-alone increased particular outcomes, such as school- basic needs. However, with the proposed 12 sessions of ing. The per-person costs of YDA + MFG, and YDA alone financial management tailored specifically to the needs will then be divided by the relevant effect sizes to produce of adolescent girls and their families in Uganda, we Ssewamala et al. BMC Public Health (2018) 18:693 Page 9 of 11 expect that even those with the fewest resources will be the results will promote holistic HIV prevention strategies equipped to make well-informed saving and investment to reduce risk among adolescent girls in Uganda and po- decisions—however modest the amount. We will carefully tentially the broader SSA region. track the saving behavior, including deposit frequency and withdrawals of each family enrolled in the study. Trial status Despite these limitations, Suubi4Her, as a combin- At the time of manuscript submission, the trial had ation intervention, is innovative and poised to make a received IRB approvals and was making preparations for contribution to the literature on HIV prevention for recruitment. The study is ongoing. adolescent girls in SSA for a number of reasons. First, it applies a theoretically guided economic empower- Acknowledgements We are grateful to Abel Mwebembezi at Reach the Youth –Uganda; Rev. Fr. ment intervention that uses incentivized/matched sav- Joseph Kato Bakulu at Masaka Catholic Diocese; Godfrey Kigozi at Rakai ings accounts. These accounts have been widely used Health Sciences Program in Uganda, Flavia Namuwonge, Phionah Namatovu with a younger primary-school going population and and Sarah Namutebi at the International Center for Child Health and Development (ICHAD) for their respective contributions to the study design demonstrated positive effects but have never been used and implementation. In addition, we are grateful to the financial institutions with older adolescents nor explicitly combined with an that agreed to work with the adolescent girls in opening savings accounts, evidence-based approach to enhancing youth behavioral and the extension workers who have committed time to train the adolescent girls in conducting income-generating activities. Our thanks also health. Increasing the economic resources of adolescent go to the Ugandan Government Ministry of Education and the 42 secondary girls, whileemphasizing theimportanceofgirls’ educa- schools that have agreed to participate in the Suubi4Her study. We also wish tion through a family strengthening approach, may help to thank Rabab Ahmed for her administrative support in the drafting of this protocol. lower HIV risks, improve mental health functioning, and increase opportunities for adolescent girls. More- Funding over, the MFG method is culturally consistent with the The study outlined in this protocol is supported by the National Institute of sub-Saharan (and Ugandan) collective approach of sup- Mental Health (NIMH) under Award Number 1R01MH113486-01(PI: Fred M. Ssewamala, PhD). The content is solely the responsibility of the authors and porting and raising children. Second, given that sexual does not necessarily represent the official views of NIMH or the National behavior and mental health are often stigmatized and Institutes of Health. prone to misreport, exclusive reliance on self-report can lead to misclassification, masked intervention ef- Authors’ contributions FS is the Lead/Principal Investigator and is responsible for overall study fects, and ambiguity when study findings are inter- conceptualization, design, and implementation. MMM, IG, LS, CM, OSB, and preted . Suubi4Her addresses this concern by GN serve as co-investigators. MM, CD, JN, and AK will provide overall study measuring sexual risk-taking behaviors objectively using management locally. TN is the statistical consultant. LGB and FS drafted the manuscript. All authors have reviewed and commented on drafts and have the following biomedical data: new HIV infection, STIs approved of the final manuscript prior to submission. (including Gonorrhea, Trichomonas and Chlamydia) and pregnancy tests. The study will also address this Ethics approval and consent to participate issue by using computer assisted self-interview to ob- All study procedures were approved by the Washington University in St. Louis Review Board (#201703102) and by in-country local IRBs in Uganda: tain more reliable responses on sensitive topics, such as Uganda Virus Research Institute (UVRI – GC/127/17/07/619)), and Uganda sexual risk taking behaviors and mental health. Lastly, National Council of Science and Technology (UNCST – SS4406). Written Suubi4Her builds upon a decade of partnerships (in- informed assent and consent will be obtained from adolescent participants and their caregivers, respectively, as part of study registration. cluding but not limited to schools, financial institu- tions, government departments, clinics, and community Competing interests organizations) to ensure interventions are rooted in a The authors declare that they have no competing interests. practical and localized understanding of the needs of older adolescent girls. These partnerships help ground Publisher’sNote the study in the community, building the capacity and Springer Nature remains neutral with regard to jurisdictional claims in helping to ensure eventual scale up, if findings warrant. published maps and institutional affiliations. Author details Conclusion Brown School, Washington University in St. Louis, 1 Brookings Drive, St. Suubi4Her is one of the first prospective studies to exam- 2 Louis, MO 63130, USA. Columbia University School of Social Work, 1255 ine the impact and cost of a combination intervention in- Amsterdam Ave., New York, NY 10027, USA. School of Medicine, University of California San Francisco, 550 16th Street, San Francisco, CA 94158, USA. tegrating economic and social components to reduce Rakai Health Sciences Program, Old Bukoba Road, 279 Kalisizo, Uganda. known HIV risk factors and improve mental health func- 5 Department of Population and Family Health, Columbia University Mailman tioning among adolescent girls, while exploring mental School of Public Health, 60 Haven Ave B-4 Suite 432, New York, NY 10032, USA. HIV Center for Clinical & Behavioral Studies, New York State Psychiatric health as a mediator in HIV risk reduction. The findings Institute and Columbia University, 1051 Riverside Dr, New York, NY 10032, will illuminate the pathways that connect economic need, 7 USA. International Center for Child Health and Development Field Office, mental health, family support, and HIV risk. If successful, Plot 23 Circular Rd, Masaka, Uganda. Ssewamala et al. BMC Public Health (2018) 18:693 Page 10 of 11 Received: 4 December 2017 Accepted: 24 May 2018 18. Ombati V, Ombati M. Gender inequality in education in sub-Saharan Africa. Journal of Women's Entrepreneurship and Education. 2012;3–4:114–36. 19. Ssewamala FM, Wang JSH, Karimli L, Nabunya P. Strengthening universal primary education in Uganda: the potential role of an asset-based development policy. Int J Educ Dev. 2011;31(5):472–7. https://doi.org/10. References 1016/j.ijedudev.2010.11.001. 1. UNAIDS. The gap report [internet]. UNAIDS; 2014. Available from: http:// 20. Ssewamala FM, Ismayilova L, McKay M, Sperber E, Bannon W Jr, Alicea S. www.unaids.org/en/resources/documents/2014/20140716_UNAIDS_gap_ Gender and the effects of an economic empowerment program on report. Accessed 16 Nov. 2017. attitudes toward sexual risk-taking among AIDS-orphaned adolescent youth 2. UNICEF. Turning the tide against AIDS will require more concentrated focus in Uganda. J Adolesc Health. 2010;46(4):372–8. https://doi.org/10.1016/j. on adolescents and young people [internet]. In: Adolescents and young jadohealth.2009.08.010. PMID:20307827 people. UNICEF; 2017. https://data.unicef.org/topic/hivaids/adolescents- 21. UNESCO. New global educatioen goals must prioritize girls. Education for all young-people. global monitoring report [Internet]. UNESCO; 2014. Available from: http://www. 3. Sherraden M. Assets and the poor: a new American welfare policy. Armonk, unesco.org/new/en/media-services/single-%20view/news/new_global_ NY: ME Sharpe; 1991. education_goals_must_prioritize_girls/#.Vnl3dvkrJph. Accessed 16 Nov. 2017. 4. Sherraden M. Stakeholding: notes on a theory of welfare based on assets. 22. Jukes M, Simmons S, Bundy D. Education and vulnerability: the role of Soc Serv Rev. 1990;64(4):580–601. schools in protecting young women and girls from HIV in southern Africa. 5. Cho H, Hallfors DD, Mbai II, Itindi J, Milimo BW, Halpern CT, Iritani BJ. AIDS. 2008;22(4):S41–56. https://doi.org/10.1097/01.aids.0000341776.71253. Keeping adolescent orphans in school to prevent human 04. PMID: 19033754 immunodeficiency virus infection: evidence from a randomized controlled 23. Nabunya P, Ssewamala FM, Ilic V. Family economic strengthening and trial in Kenya. J Adolesc Health. 2011;48(5):523–6. https://doi.org/10.1016/j. parenting stress among caregivers of AIDS-orphaned children: results from jadohealth.2010.08.007. PMID:21501814 a cluster randomized clinical trial in Uganda. Child Youth Serv Rev. 2014;44: 6. Glynn JR, Caraël M, Auvert B, Kahindo M, Chege J, Musonda R, Kaona F, 417–21. https://doi.org/10.1016/j.childyouth.2014.07.018. PMID: 25136142 Buvé A, et al. Why do young women have a much higher prevalence of HIV 24. Ismayilova L, Ssewamala FM, Karimli L. Family support as a mediator of than young men? A study in Kisumu, Kenya and Ndola, Zambia. AIDS. 2001; change in sexual risk-taking attitudes among orphaned adolescents in rural 15:S51–60. PMID: 11686466 Uganda. The Journal of Adolescent Health. 2012;50(3):228–35. https://doi. 7. Gregson S, Garnett GP. Contrasting gender differentials in HIV-prevalence org/10.1016/j.jadohealth.2011.06.008. PMID: 22325127; PMCID: PMC3279703 and associated mortality increase in eastern and southern Africa: Artifact of 25. Ssewamala FM, Karimli L, Chang-Keun H, Ismayilova L. Social capital, savings, data or natural course of epidemics. AIDS. 2000;14(Supplement 3):S85–99. and educational performance of orphaned adolescents in sub-Saharan PMID:11086852 Africa. Child Youth Serv Rev. 2010;32(12):1704–10. https://doi.org/10.1016/j. 8. Hallfors D, Cho H, Rusakaniko S, Iritani B, Mapfumo J, Halpern C. Supporting childyouth.2010.07.013. adolescent orphan girls to stay in school as HIV risk prevention: evidence 26. McNeely C, Shew ML, Beuhring T, Sieving R, Miller BC, Blum RW. Mothers’ from a randomized controlled trial in Zimbabwe. Am J Public Health. 2011; influence on the timing of first sex among 14-and 15-year-olds. J Adolesc 101(6):1082–8. https://doi.org/10.2105/AJPH.2010.300042. PMID: 21493943 Health. 2002;31(3):256–65. https://doi.org/10.1016/S1054-139X(02)00350-6. 9. Hargreaves J & Boler T. Girl Power. The impact of girls' education on HIV PMID:12225738 and sexual behavior [Internet]. ActionAid International; 2006. Available from: 27. Resnick MD, Bearman PS, Blum RW, Bauman KE, Harris KM, Jones J, Tabor J, https://www.actionaid.org.uk/sites/default/files/doc_lib/girl_power_2006.pdf. Beuhring T, et al. Protecting adolescents from harm: findings from the Accessed 17 Nov. 2017. national longitudinal study on adolescent health. JAMA. 1997;278(10):823– 10. Nobelius AM, Kalina B, Pool R, Whitworth J, Chesters J, Power R. Sexual 32. https://doi.org/10.1001/jama.1997.03550100049038. PMID:9293990 partner types and related sexual health risk among out-of-school 28. Amerikaner M, Monks G, Wolfe P, Thomas S. Family interaction and adolescents in rural south- West Uganda. AIDS Care. 2011;23(2):252–9. individual psychological health. J Couns Dev. 1994;72(6):614–20. https://doi. https://doi.org/10.1080/09540121. PMID: 21259139 org/10.1002/j.1556-6676.1994.tb01691.x. 11. Pettifor AE, Levandowski BA, Macphail C, Padian NS, Cohen MS, Rees HV. 29. Klein M, Gordon S. Sex education in Walker CE, Roberts MC (eds). Handbook of Keep them in school: the importance of education as a protective factor clinical child psychology. 3rd ed. New York, NY: John Wiley & Sons, Inc.; 2001. against HIV infection among young South African women. International 30. Kotchick BA, Dorsey S, Miller KS, Forehand R. Adolescent sexual risk-taking Journal of Epidemiology. 2008;37(6):1266–73. https://doi.org/10.1080/ behavior in single-parent ethnic minority families. J Fam Psychol. 1999;13(1): 09540121.2010.507736. PMID: 18614609 PMCID: PMC2734068 93–102. https://doi.org/10.1037/0893-3184.108.40.206. 12. Gunn JK, Roth AM, Center KE & Wiehe SE. Gunn JK, Roth AM, Center KE, 31. Miller KS, Forehand R, Kotchick BA. Adolescent sexual behavior in two Wiehe SE. The unanticipated benefits of behavioral assessments and ethnic minority samples: the role of family variables. J Marriage Fam. 1999; interviews on anxiety, self-esteem and depression among women engaging 61(1):85–98. https://doi.org/10.2307/353885. in transactional sex. Community Ment Health J 2016;52(8):1064–1069. DOI: 32. Brooks-Gunn J, Furstenberg FF Jr. Adolescent sexual behavior. Am Psychol. https://doi.org/10.1007/s10597-015-9844-x; PMID: 25712538. 1989;44(2):249–57. https://doi.org/10.1037/0003-066X.44.2.249. 13. Meier A, Erickson GA, McLaughlin H. Older sexual partners and adolescent 33. The Republic of Uganda. The HIV and AIDS Uganda Country Progress females’ mental health. Perspect Sex Reprod Health. 2016;48(1):25–33. Report [Internet]. The Republic of Uganda; 2014. Available from: http:// https://doi.org/10.1363/48e8316. PMID:26918326 www.unaids.org/sites/default/files/country/documents/UGA_narrative_ 14. Barhafumwa B, Dietrich J, Closson K, Samji H, Cescon A, Nkala B, Davis J, report_2015.pdf. Accessed 16 Nov. 2017. Hogg RS, et al. High prevalence of depression symptomology among adolescents in Soweto, South Africa associated with being female and 34. UNAIDS, UNICEF. UNICEF Annu Rep, Uganda [Internet]. UNAIDS, UNICEF; cofactors relating to HIV transmission. Vulnerable Children and Youth 2014. Available from: https://reliefweb.int/sites/reliefweb.int/files/resources/ Studies. 2016;11(3):263–73. https://doi.org/10.1080/17450128.2016.1198854. Uganda_2.pdf. Accessed 16 Nov. 2017. 15. Agbor J. Poverty, inequality and Africa's education crisis [Internet]. 35. Bellamy SL. A dynamic block-randomization algorithm for group- Brookings; 2012. Available from: https://www.brookings.edu/opinions/ randomized clinical trials when the composition of blocking factors is not poverty-inequality-and-africas-education-crisis/. Accessed 16 Nov. 2017. known in advance. Contemporary Clinical Trials. 2005;26(4):469–79. https:// doi.org/10.1016/j.cct.2005.02.005. PMID:16054579 UNICEF & The World Bank. (2009). 36. Bandura A. Health promotion by social cognitive means. Health Educ Behav. 16. The World Bank. Abolishing school fees in Africa: Lessons from Ethiopia, 2004 Apr;31(2):143–64. Kenya, Malawi and Mozambique [Internet]. World Bank; 2009. Available from: http://documents.worldbank.org/curated/en/780521468250868445/ 37. Fisher WA, Fisher JD, Rye BJ. Understanding and promoting AIDS-preventive Abolishing-school-fees-in-Africa-lessons-from-Ethiopia-Ghana-Kenya-Malawi- behavior: insights from the theory of reasoned action. Health Psychol. 1995; and-Mozambique. Accessed 16 Nov. 2017. 14(3):255. 17. Colclough C, Rose P, Tembon M. Gender inequalities in primary schooling: 38. Dennison ST. A multiple family group therapy program for at risk the roles of poverty and adverse cultural practice. Int J Educ Dev. 2000;20(1): adolescents and their families. 1st ed. Springfield, Illinois: Charles C Thomas 50–27. https://doi.org/10.1016/S0738-0e593(99)00046-2. Publisher; 2005. Ssewamala et al. BMC Public Health (2018) 18:693 Page 11 of 11 39. Mellins CA, Nestadt D, Bhana A, Petersen I, Abrams EJ, Alicea S, Holst H, 62. Microfinance Opportunities. Global financial education program [internet]. Myeza N, John S, Small L, McKay M. Adapting evidence-based interventions Microfinance Opportunities; 2002. Available from: https://www. to meet the needs of adolescents growing up with HIV in South Africa: the microfinanceopportunities.org/4-work-with-us/mfo-in-the-field/project-list/ VUKA case example. Global Social Welfare. 2014;1(3):97–110. fecc/. Accessed 16 Nov. 2017. 40. McKay MM, Gonzales JJ, Stone S, Ryland D, Kohner K. Multiple family 63. National Institute on Drug Abuse. Risk behavior assessment, 3rd ed. therapy groups: a responsive intervention model for inner city families. [Internet]. National Institute on Drug Abuse; 1993. Retrieved from http://bit. Social Work with Groups. 1995;18(4):41–56. https://doi.org/10.1300/ ly/RBA_inst. J009v18n04_04. 41. Keiley MK. The development and implementation of an affect regulation and attachment intervention for incarcerated adolescents and their parents. Fam J. 2002;10(2):177–89. https://doi.org/10.1177/1066480702102007. 42. Kumpfer KL, Alvarado R, Smith P, Bellamy N. Cultural sensitivity and adaptation in family-based prevention interventions. Prev Sci. 2002;3(3):241– 6. https://doi.org/10.1023/A:1019902902119. PMID:12387558 43. Wahler RG, Dumas JE. Attentional problems in dysfunctional mother-child interactions: an interbehavioral model. Psychol Bull. 1989;105(1):116–30. https://doi.org/10.1037/0033-2909.105.1.116; PMID:2648437 44. Jewell TC, Downing D, McFarlane WR. Partnering with families: multiple family group psychoeducation for schizophrenia. J Clin Psychol. 2009;65(8): 868–78. https://doi.org/10.1002/jclp.20610; PMID: 19530233. 45. UNESCO, UNFPA. Sexuality education: A ten-country review of school curricula in East and Southern Africa [Internet]. UNESCO, UNFPA; 2012. Available from: http://unesdoc.unesco.org/images/0022/002211/221121e. pdf. Accessed 16 Nov. 2017. 46. Hayes R, Moulton LH. Cluster randomized trials.1st ed. Boca Raton. Florida: CRC Press; 2009. 47. Crosby RA, Rothenberg R. In STI interventions, size matters. Sex Transm Infect. 2004;80(2):82–5. 48. Gallo MF, Steiner MJ, Hobbs MM, Warner L, Jamieson DJ, Macaluso M. Biological markers of sexual activity: tools for improving measurement in HIV/sexually transmitted infection prevention research. Sex Transm Dis. 2013;40(6):447–52. https://doi.org/10.1097/OLQ.0b013e31828b2f77; PMID: 49. Sobell LC, Sobell MB. Timeline Followback (TLFB) User's manual. Toronto: Addiction Research Foundation; 1994. 50. Beck AT, Weissman A, Lester D, Trexler L. The measurement of pessimism: the hopelessness scale. J Consult Clin Psychol. 1974;42(6):861–5. 4436473 51. Fitts WH, Roid GH. Tennessee self concept scale. Nashville, TN: Counselor Recordings and Tests; 1964. 52. Kovacs M. The child depression inventory (CDI). North Tonawanda, NY: Multi-Health Systems, Inc; 1992. 53. Levy SR, Lampman C, Handle A. Young adolescent attitudes towards sex and substance abuse: implications for AIDS prevention. AIDS Educ Prev. 1993;5(4):340–51. 19207597 54. Vaux A, Riedel S, Stewart D. Modes of social support: the social support behaviors (SS-B) scale. Am J Community Psychol. 1987;15(2):210–37. https:// doi.org/10.1007/BF00919279. 55. Nabunya P, Ssewamala FM. The effects of parental loss on the psychosocial wellbeing of AIDS-orphaned children living in AIDS-impacted communities: does gender matter?. Child Youth Serv Rev 2014;43 Suppl C:131–7. DOI: https://doi.org/10.1016/j.childyouth.2014.05.011; PMID:25067869. 56. Krauss BJ. Calm down, Mom, let’s talk about sex, drugs, and HIV: 10–13 year old girls prescription for HIV Preventieon Conversations in their high HIV seroprevalence neighborhood. Washington DC: Paper presented at the HIV Infection in Women Conference; 1995. 57. Gerard AB. Parent-child relationship inventory (PCRI). 1st ed. Los Angeles: western. Psychol Serv. 1994; 58. Rosenberg M. Rosenberg self-esteem scale (RSE) [Internet]. Acceptance and commitment therapy. Measures package, 61; 1965. Available from: http:// fetzer.org/sites/default/files/images/stories/pdf/selfmeasures/Self_Measures_ for_Self-Esteem_ROSENBERG_SELF-ESTEEM.pdf Accessed 16 Nov. 2017. 59. Nanda G. Compendium of gender scales [internet]. Nanda G.; 2011. Available from: https://www.c-changeprogram.org/content/gender-scales- compendium/pdfs/C-Change_Gender_Scales_Compendium.pdf. Accessed 16 Nov. 2017. 60. Wingood GM, DiClemente RJ. The influence of psychosocial factors, alcohol, drug use on African-American women’s high-risk sexual behavior. Am J Prev Med. 1998;15(1):54–9. 9651639 61. Zimet GD, Dahlem NW, Zimet SG, Farley GK. The multidimensional scale of perceived social support. J Pers Assess. 1988;52(1):30–41. https://doi.org/10. 1207/s15327752jpa5201_2.
BMC Public Health – Springer Journals
Published: Jun 5, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
All the latest content is available, no embargo periods.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud