UCLA
UCLA Previously Published Works
Title
Family-based promotion of mental health in children affected by HIV: a pilot randomized
controlled trial.
Permalink
https://escholarship.org/uc/item/13c597st
Journal
Journal of child psychology and psychiatry, and allied disciplines, 58(8)
ISSN
0021-9630
Authors
Betancourt, Theresa S
Ng, Lauren C
Kirk, Catherine M
et al.
Publication Date
2017-08-01
DOI
10.1111/jcpp.12729
Peer reviewed
eScholarship.org Powered by the California Digital Library
University of California
Family-based promotion of mental health in children
affected by HIV: a pilot randomized controlled trial
Theresa S. Betancourt,
1
Lauren C. Ng,
2,3
Catherine M. Kirk,
1
Robert T. Brennan,
1
William
R. Beardslee,
4
Sara Stulac,
5,6
Christine Mushashi,
7
Estella Nduwimana,
7
Sylvere Mukunzi,
7
Beatha Nyirandagijimana,
7
Godfrey Kalisa,
7
Cyamatare F. Rwabukwisi,
7
and
Vincent Sezibera
8
1
Department of Global Health and Population, Harvard T. H. Chan School of Public Health, Boston, MA;
2
Department
of Psychiatry, Boston University School of Medicine, Boston, MA;
3
Boston Medical Center, Boston, MA;
4
Department
of Psychiatry, Boston Children’s Hospital, Boston, MA;
5
Brigham and Women’s Hospital, Boston, MA;
6
Partners in
Health, Boston, MA, USA;
7
Partners In Health/Inshuti MuBuzima, Rwinkwavu;
8
College of Medicine and Health
Sciences, University of Rwanda, Butare, Rwanda
Background: Children affected by HIV are at risk for poor mental health. We conducted a pilot randomized controlled
trial (RCT) of the Family Strengthening Intervention (FSI-HIV), a family home-visiting intervention to promote mental
health and improve parent–child relationships in families with caregivers living with HIV, hypothesizing that child
and family outcomes would be superior to usual care social work services. Methods: Eighty two families (N = 170
children, 48.24% female; N = 123 caregivers, 68.29% female) with at least one HIV-positive caregiver (n = 103,
83.74%) and school-aged child (ages 7–17) (HIV+ n = 21, 12.35%) were randomized to receive FSI-HIV or treatment-
as-usual (TAU). Local research assistants blind to treatment conducted assessments of child mental health,
parenting practices, and family functioning at baseline, post-intervention, and 3-month follow-up. Multilevel
modeling assessed effects of FSI-HIV on outcomes across three time points. Trial Registration: NCT01509573, ‘Pilot
Feasibility Trial of the Family Strengthening Intervention in Rwanda (FSI-HIV-R).’ https://clinicaltrials.gov/ct2/
show/NCT01509573?term=Pilot+Feasibility+Trial+of+the+Family+Strengthening+Intervention+in+Rwanda+%28FSI-
HIV-R%29&rank=1. Results: At 3-month follow-up, children in FSI-HIV showed fewer symptoms of depression
compared to TAU by both self-report (b = .246; p = .009) and parent report (b = .174; p = .035) but there were no
significant differences by group on conduct problems, functional impairment, family connectedness, or parenting.
Conclusions: Family-based prevention has promise for reducing depression symptoms in children affected by HIV.
Future trials should examine the effects of FSI-HIV over time in trials powered to examine treatment mediators.
Keywords: HIV; depression; Rwanda; adolescents.
Introduction
The consequences of caregiver HIV on families can
be devastating (Tol et al., 2011) including illness
and death, functional impairments, mental health
problems, stigma, economic distress, and conflict
(Doku, 2009; Nozyce et al., 2006). In HIV-affected
families, poor communication, anxiety, and conflict
increase the risk of mental health problems in
children (Betancourt, Scorza, Kanyanganzi, Fawzi,
Sezibera, Cyamatare, and Stevenson, 2014). The
shifting of adult responsibilities such as economic
burden and caring for young children is associated
with child and caregiver distress (Akresh & De
Walque, 2008).
Rwanda has emerged from the legacy of genocide and
the HIV epidemic with striking improvements in its
health system. Since 2001, the country has halved HIV
prevalence (UNAIDS, 2012b) and overall child mortality
(Ministry of Health & ICF International, 2012). Access
to antiretroviral therapy and prevention of mother-to-
child HIV transmission are widespread (UNAIDS,
2012a). However, affordable and effective interventions
are needed to improve family functioning and prevent
mental health problems in HIV-affected children.
Several mental health interventions have targeted
children orphaned and otherwise made vulnerable by
HIV and HIV-positive children in low- and middle-
income countries (Kumakech, Cantor-Graae, Maling,
& Bajunirwe, 2009; Ssewamala, Han, & Neilands,
2009). Promising family-based interventions such as
the Collaborative HIV Prevention and Adolescent
Mental Health Program in South Africa (Bhana,
McKay, Mellins, Petersen, & Bell, 2010) and the
Together for Empowerment Activities in China (Li
et al., 2011) have targeted HIV risk behaviors and
healthy communication through family group mod-
els. However, few interventions exist to explicitly
prevent mental health problems in school-aged chil-
dren both directly and indirectly affected by HIV. We
adapted an evidence-based, home-visiting preventive
intervention (Beardslee, Wright, Gladstone, & For-
bes, 2007; Betancourt, Ng et al., 2014) and examined
its feasibility and effects on parent–child relation-
ships and child mental health within HIV-affected
households in rural Rwanda (Betancourt et al.,
2012). Rwanda’s robust HIV services present an
opportunity to link mental health services to routine
care for HIV-affected families.
Conflicts of interest statement: No conflicts declared.
© 2017 Association for Child and Adolescent Mental Health.
Published by John Wiley & Sons Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main St, Malden, MA 02148, USA
Journal of Child Psychology and Psychiatry **:* (2017), pp **–** doi:10.1111/jcpp.12729
Methods
In collaboration with Partners In Health/Inshuti Mu Buzima
(PIH/IMB) and the Rwandan Ministry of Health (MOH), we
tested the effectiveness of the Family Strengthening Interven-
tion (FSI-HIV) (Betancourt, Ng et al., 2014) in a pilot random-
ized controlled trial with families with at least one HIV-positive
primary caregiver of school-aged children (7–17 years). We
hypothesized that post-intervention, families assigned to
FSI-HIV would report better caregiver-child relationships and
family protective constructs and as a result, children would
have fewer mental health problems (symptoms of depression
and conduct problems) and less functional impairment com-
pared to control families receiving treatment as usual (TAU) via
routine social work.
Participants and settings
Families were recruited through health centers in Kayonza
District where PIH/IMB provides support to the public health
system. Referrals were made by MOH social workers who were
engaged in ongoing work with HIV-affected households. Inclu-
sion criteria were: (a) being an adult-headed household where
at least one caregiver was HIV-positive; (b) having at least one
child aged 7–17; and (c) willingness to discuss HIV with school-
age children. Eighty-two HIV-affected families were enrolled in
the trial, stratified by equal allocation of single- and dual-
caregiver structures.
IRB approval was obtained from the Harvard T. H. Chan
School of Public Health (Protocol #15440) and Rwanda
National Ethics Committee. Community Advisory Boards
(CABs) were assembled to advise the project and included
groups with clinicians, youth, and one with community lead-
ers, community health workers, and caregivers living with HIV.
With input from these CABs, the FSI-HIV was developed to be
delivered by bachelor-level counselors for reasons of safety and
feasibility testing, with the long-term aim of adapting the
intervention for delivery by community health workers (Betan-
court, Ng et al., 2014). All participants aged 18 and older
provided written consent for themselves; eligible children
provided written assent and a primary caregiver provided
written consent.
Trial registration
NCT01509573, ‘Pilot Feasibility Trial of the Family Strength-
ening Intervention in Rwanda (FSI-HIV-R).’ https://clinicaltria
ls.gov/ct2/show/NCT01509573?term=Pilot+Feasibility+Trial+
of+the+Family+Strengthening+Interv ention+in+Rwanda+%28
FSI-HIV-R%29&rank=1.
Randomization and masking
A randomization sequence in Microsoft Excel was used to
assign families to each condition, with equal balance of dual-
and single-caregiver households, given literature indicating
different support needs and stressors by family configuration
(Human & van Rensburg, 2011). All families were introduced
to the study by the social worker at their nearest health center.
Randomization occurred after baseline assessments. TAU
families were seen at least once by a social worker over the
course of the FSI-HIV intervention study and FSI-HIV families
continued to receive TAU. TAU assistance most often pertained
to food insecurity and assistance with accessing school.
Study measures
Measures were adapted and validated for the Rwandan context
following mixed methods research on local constructs of family
functioning, mental health, and resilience (Betancourt, Mey-
ers-Ohki et al., 2011; Betancourt, Rubin-Smith et al., 2011).
The 20-item Center for Epidemiologic Studies Depression
Scale for Children (CES-DC) (Faulstich, Carey, & Ruggiero,
1986) was adapted and validated to measure depression in
Rwanda (Betancourt et al., 2012). Following item-response
theory (IRT) analyses on a larger sample of Rwandan children
and parents (Betancourt et al., 2012; Betancourt, Scorza
et al., 2014), four items were dropped from the original CES-
DC scale based on poor item functioning in the Rwandan
context, resulting in a 16-item scale with excellent internal
consistency (a = .91). Items were scored 0 (‘Not at all’) to 3 (‘A
lot’) and the scale score was the sum. A locally developed and
validated conduct problems scale (Ng, Kanyanganzi, Mun-
yanah, Mushashi, & Betancourt, 2014) also showed good
internal consistency (a = .90). Responses were on a four-point
scale from 0 (‘never’) to 3 (‘often’) and the scale score was the
mean. Functional impairment was assessed using an adapted
version of the World Health Organization Disability Assess-
ment Schedule for Children (WHODAS-Child) (Scorza et al.,
2013). The 18-item scale (a = .88) assessed six domains:
understanding and communicating, mobility, self-care, getting
along with people, life activities, and participation in society.
Simple scoring was used (World Health Organization, 2014) in
which scores of 0 (‘no difficulty’) to 4 (‘extreme difficulty/
cannot do’) were averaged.
Family connectedness was assessed by a 15-item scale
(a = .93) with indicators developed from qualitative research
(Barerra, Sandler, & Ramsay, 1983; Betancourt, Meyers-Ohki
et al., 2011). Items were scored on a five-point scale from 0
(‘never’) to 4 (‘always’) and were averaged. A 32-item good
parenting scale was reduced to 21 items using IRT analyses
(a = .92), contained 16 locally derived items (Betancourt,
Meyers-Ohki et al., 2011) and five items from the Parental
Acceptance and Rejection Questionnaire (Rohner, Saavedra, &
Granum, 1978) and was scored on four-point scale from 0
(‘never’) to 4 (‘every day’), using the mean as the scale score.
Socioeconomic status was calculated using an asset index
from the Rwanda Demographic and Health Survey (National
Institute of Statistics, 2012), including improved materials for
construction of the home (i.e. cement floor vs. dirt), ownership
of livestock, and ownership of other durable goods in the
household such as a radio or television. Higher scores repre-
sent greater household assets.
Participants in FSI-HIV completed a ten-item questionnaire
assessing their satisfaction with the intervention (e.g. ‘Overall,
how satisfied are you with the FSI-HIV sessions you partici-
pated in?’ and ‘Would you recommend the FSI-HIV to a friend
or neighbor?’).
Intervention
The FSI-HIV is a manualized modular intervention adapted
from the Family Based Preventive Intervention (FBPI). The
theory behind the original FBPI (also called ‘Family Talk’) is
that poor communication and negative parent–child dynamics
due to chronic parental illness can lead to increased risk of
mental health problems in offspring. The FSI-HIV adapted this
family-based approach, originally developed for the prevention
of depression in the offspring of depressed caregivers to the
context of families affected by caregiver HIV. The original FBPI
has a strong evidence base in several diverse cultural settings.
The intervention has received high rankings in the National
Registry of Evidence-based Programs and Practices (NREPP,
2009). It has been adapted for use with different cultural
groups including Latino and Native American populations in
the United States and has been integrated into health services
in Costa Rica and Finland (Beardslee, Solantaus, Morgan,
Gladstone, & Kowalenko, 2013; Beardslee et al., 2011; D’Angelo
et al., 2009; Podorefsky, McDonald-Dowdell, & Beardslee,
2001; Sparrow et al., 2011). In prior trials, participation in the
© 2017 Association for Child and Adolescent Mental Health.
2 Theresa S. Betancourt et al.
intervention was associated with increased family under-
standing of depression, and decreased depressive symptoma-
tology scores in children (Beardslee, Gladstone, Wright, &
Cooper, 2003; Beardslee et al., 2007; D’Angelo et al., 2009;
Podorefsky et al., 2001). Given the FBPI’s focus on improving
the caregiver–child relationship and communication in the
context of chronic caregiver illness, the FBPI was seen as well
aligned with the challenges within families that can contribute
to the higher rates of mental health problems that have been
observed in HIV-affected children (Betancourt et al., 2012;
Betancourt, Scorza et al., 2014). Like the original FBPI, FSI-
HIV is strengths-based and has at its core the development of
a family narrative that draws out the major challenges that
the family has faced together and the inherent strengths that
have helped the family to overcome challenges in the past with
a view toward continued future success as a family through
improved communication and understanding. The narrative
and strengths focus of the FBPI were seen as aligned with
views of resilience among families facing adversity in Rwanda
as illuminated in prior qualitative research on protective
processes in this setting.
Drawing from this strengths-based theory of change, FSI-
HIV is delivered in approximately 90-min weekly home-visiting
sessions which span an initial premeeting, the six core
modules including a culminating family meeting, and a
follow-up to the family meeting in order to debrief together
about what occurred during the family meeting. The core
modules address four themes: (a) psychoeducation about HIV
and its effects on families and how families can succeed
despite HIV; (b) skills development in communication, respon-
sive parenting, and stress management/alternatives to vio-
lence and harsh punishment; (c) development of a family
strengths-based narrative that identifies sources of family
resilience and hope from the perspective of the caregivers, the
children, and a combined family narrative; and (d) problem
solving around eliciting formal and informal support struc-
tures (Betancourt, Ng et al., 2014).
The FSI-HIV model blends meetings solely for the caregivers
and solely for the children together with a meeting with all
family members. In households with children of diverse ages,
the modules for children could be split into separate sessions
for younger children and adolescents as appropriate. The
culmination of the intervention is a family meeting convened by
caregivers with support from counselors. Overall there are at
least two combined family meetings, three sessions just with
caregivers and at least two sessions just with children. The
overview and flow of all sessions is laid out in Figure 1. The
FSI-HIV acceptability and the content, pace, and timing of
modules were pretested in a previous open trial (Betancourt,
Ng et al., 2014).
As needed, supplementary psychoeducation on genocide-
related trauma was developed and provided for families where
the issue arose (see Figure 1 for FSI-HIV modules). Although
most of the caregivers in the trial had lived through the
genocide, caregivers were not pushed to discuss their expe-
riences in the genocide. Instead, the family narrative compo-
nent of FSI-HIV allowed families to choose which important
family events to discuss. Genocide was introduced as a theme
in 15% of intervention families.
Procedures
A Rwanda-based project manager (Masters-degree level) ran-
domized families and provided weekly on-site supervision.
Study leaders, including a child psychiatrist and clinical
psychologist, provided additional weekly supervision to the
intervention team by phone. Three male and three female
bachelor-level Rwandan counselors were given a two-week
training in the FSI-HIV focused on role-play to build skills in
family counseling. Families assigned to TAU received standard
social work services through the Ministry of Health, which
included sessions at the health center and/or home facilitated
by a social worker. TAU social workers received no additional
training and were asked to provide usual social work services.
Families in both treatment conditions were offered referrals
for additional mental health services, health insurance
enrollment, health services, and information on the health
system as needed. Counselors maintained clinical notes on
challenges FSI-HIV families faced, including intimate partner
violence, alcohol abuse, and effects of the Rwandan genocide,
in addition to process data on intervention progress and
session attendance.
Trained research staff conducted blinded baseline, post, and
three-month follow-up assessments via in-person interviews
using hand-held Android devices. Assessments were conducted
within a month of both beginning and completing the interven-
tion and then three months after the end of the intervention.
Children reported on their own mental health, functional
impairment, and the parenting they received. The caregiver
who knew each child best reported on the mental health and
functional impairment of the child and parenting that the child
received from caregivers. Caregivers also reported on their own
parenting and sense of family connectedness. Caregivers pro-
vided written informed consent for themselves and their chil-
dren. Children provided independent written assent. All
consent and assent procedures were witnessed and docu-
mented by a study staff member. Six families refused to
participate in the study (6.8% of all eligible families sampled).
Statistical analysis
Assuming a standard alpha level of 0.05, 80 families with two
eligible respondents per family on average, and assumptions of
moderate intraclass (within-family) correlation (approximately
0.5), this pilot RCT had power of 0.80 to detect a standardized
‘medium’ effect size of approximately 0.50 (Cohen, 2013).
We used mixed models (also known as multilevel models) to
assess the three-wave longitudinal effects of FSI-HIV on
caregiver-child relationships and child mental health and
functioning, accounting for family clustering and clustering
of the three time points within persons. Descriptive analyses
were conducted in Stata 14 (StataCorp LP) and multilevel
models were estimated using HLM 7 (Scientific Software
International, Inc.). Where normality assumptions for residu-
als could not be sustained due to skewed scale scores, we
employed a generalized linear model with a Poisson link
function. All models predicting child outcomes were adjusted
for child sex, child age, family SES, family type (single- or dual-
caregiver family), and child HIV status. For caregiver-reported
outcomes, caregiver sex, age, family type, and caregiver HIV
status were included. Time dummy variables (post-treatment
and three-month follow-up) were included to account for
change, and treatment (FSI-HIV dummy) by time dummy
interaction terms were included to test the effect of the FSI-HIV
treatment on outcomes. The mode of analysis was intent-to-
treat with 20 multiply-imputed data sets created in Stata 14
using chained equations to account for missing values and
participants who did not complete all assessments (9.2% by
three-month follow-up) (Tang, Song, Belin, & Un
€
utzer, 2005).
Results
Baseline characteristics
Forty-one families were randomized to FSI-HIV. One
FSI-HIV family withdrew prior to the intervention,
and one TAU family relocated outside of the study
area (Figure 2).
Descriptive statistics are displayed in Table 1.
Caregivers who reported on the children were most
© 2017 Association for Child and Adolescent Mental Health.
Promotion of mental health in children affected by HIV
3
frequently biological mothers (75.0%) and fathers
(19.5%). There were 42 single-caregiver families with
43 caregivers (in one family the mother was enrolled,
however, mid-study, the grandmother moved in to
help care for the children), and 40 dual-caregiver
families enrolled in the research. All single-
caregiver families were headed by females. One
dual-caregiver family was headed by a mother and
a grandmother, and all others were headed by a man
and a woman. While all single caregivers were HIV-
positive, sixteen (40.0%) of the dual-caregiver fami-
lies were serodiscordant. Of 170 children enrolled in
the trial, 21 (12.35%) were HIV-positive.
Attendance
The overall attendance rate was 93.4% including the
family that withdrew and a family that could not
continue due to children leaving the home. Fully 150
participants attended at least one module, and only
3% of those who participated missed more than one
module. The mean number of modules attended was
6.5 of 7 modules for caregivers (SD = 1.3) and 3.8 of
4 modules for children (SD = 1.1).
The number of sessions ranged between 8 and 21,
with families taking 10.0 sessions on average (SD =
3.9) to complete the premeeting, six core modules,
Pre-Meeting - Caregivers, Children, and Interventionist
Introduction of the intervention and main goals.
Identify who in the family will participate in the intervention and what the family hopes to get out of it.
Module 1 - Caregivers and Interventionist
Develop the family narrative, or family history, with a focus on family strengths and challenges
Introduce strategies for positive parenting
Module 2 - Caregivers and Interventionist
HIV/AIDS psychoeducation for caregivers
Continuation of the family narrative with a focus on how illness has affected the family
Module 3 - Children and Interventionist
Develop the family narrative from the children's perspective
Children identify the strengths and challenges of the family
HIV/AIDS psychoeducation for children and discussion of illness in the family
Module 4 - Caregivers and Interventionist
Discuss with caregivers sources of resilience in the family and the children's concerns
Build parenting and communication skills
Prepare caregivers for the family meeting, including role plays and setting the agenda of topics
Module 5 - Children and Interventionist
Discuss with children sources of resilience in the family
Build coping and communication skills
Prepare children for the Family Meeting with caregiver(s), utilizing new communication skills through
role plays
Module 6 - Family Meeting with Caregivers, Children, and Interventionist
Conduct a family meeting to establish shared goals and expand the family narrative
Discuss challenges and strengths of family communication
Family Meeting Review - Caregivers, Children, and Interventionist
Review the family meeting, discuss family goals, and inquire about family functioning
Review key psychoeducation on HIV/AIDS and resilience
Discuss previous and new family concerns
Figure 1 FSI modules
© 2017 Association for Child and Adolescent Mental Health.
4 Theresa S. Betancourt et al.
