Research Article

Research on Social Work Practice

2022, Vol. 32(6) 635–650
Interventions to Improve Outcomes of © The Author(s) 2022
Article reuse guidelines:
Grandchildren Raised by Grandparents: A [Link]/journals-permissions
DOI: 10.1177/10497315221079352
[Link]/home/rsw
Systematic Review

Yanfeng Xu1 , Sarah Pace1, Lauren P. McCarthy2, Theresa M. Harrison3, and Yao Wang2

Abstract
Purpose: This systematic review aims to examine the effectiveness of interventions that seek to improve outcomes of
grandchildren raised by grandparents. Method: A systematic review following the Preferred Reporting Items for Systematic
Reviews and Meta-Analyses guidelines was undertaken. We searched in Education Resources Information Center (ERIC),
Family Studies Abstracts, PubMed, PsycINFO, Social Work Abstracts, Sociological Abstracts, Web of Science, and Google
Scholar. Results indicated that grandchildren who participated in these interventions had improvement in their socio-emotional,
mental, and behavioral outcomes but had mixed educational outcomes, regardless of research designs and evaluation methods.
Two out of eight studies were of high quality, while the rest needed improved rigor in designs and analyses. Conclusions: The
review summarized evidence for interventions to improve grandchildren’s outcomes and pointed out directions for future
research and practice.

Keywords
interventions, grandparents raising grandchildren, custodial grandparents, socio-emotional well-being, mental health, behavioral
health, education

Grandparents raising grandchildren, or custodial grandparents, of grandparents living with grandchildren continue to work, close
are defined as those who raise their grandchildren, ages 18 or to 19% of them live below the federal poverty line, and about
younger, without biological parents living in the home (Hayslip, 25% experience one or more disabilities (Generations United,
Shore, Henderson, & Lambert, 1998). These children are referred 2021; U.S. Census Bureau, 2018).
to as custodial grandchildren (Hayslip et al., 1998). On a global Grandparents raising grandchildren face a myriad of chal-
scale, grandparents raising grandchildren has become increas- lenges, such as stress from their unexpected role, legal challenges
ingly common. Similarly, the number of grandchildren raised by related to custody, a lack of resources, experiencing social iso-
their grandparents has increased to about 2.7 million in recent lation, dealing with parenting challenges and children’s prob-
years in the United States alone (Generations United, 2021; U.S. lems, and navigating their relationship with biological parents
Census Bureau, 2018). Although reasons for grandparents to take (Choi, Sprang, & Eslinger, 2016; Generations United, 2021;
the role of parents vary across countries, similar trigger events for Goodman et al., 2004). Similarly, the separation from biological
grandparents raising grandchildren have been observed across parents and transition into grandparental care make grand-
western countries (Dolbin-MacNab, 2020). For instance, the children especially vulnerable. Due to childhood trauma and
most common reasons leading to grandparents raising grand- difficult life transitions, grandchildren face additional stress and
children in the United States included parental child abuse and challenges, such as physical health concerns, socio-emotional,
neglect, parental substance abuse, mental or other illness, in- mental, and behavioral problems, and poor educational outcomes
carceration, death of the biological parents, deployment, and
financial challenges (Campbell, Carthron, Miles, & Brown,
2012; Goodman, Potts, Pasztor, & Scorzo, 2004; Smith, 1
University of South Carolina College of Social Work, Columbia, SC, USA
Hayslip, Hancock, Strieder, & Montoro-Rodriguez, 2018). Re- 2
University of Maryland School of Social Work, Baltimore, MD, USA
3
garding characteristics of these grandparents in the United States, University of South Carolina College of Education, Columbia, SC, USA
about half are over 60 years of age; about 66% are White; over
Corresponding Author:
19% are African American; and about 20% are Hispanic or Yanfeng Xu, University of South Carolina College of Social Work, 1512
Latino (Generations United, 2021; U.S. Census Bureau, 2018). Pendleton St., Hamilton College, Columbia, SC 29208, USA.
The U.S. national data further suggests that although around 55% Email: yanfeng@[Link]
636 Research on Social Work Practice 32(6)

(Campbell et al., 2012; Generations United, 2021; Goodman grandparents in the United States had relatively lower academic
et al., 2004; Smith et al., 2018). performance, lower school engagement levels, and a higher
likelihood of needing special education than their counterparts,
Socio-Emotional, Mental, and Behavioral who were raised by biological parents, and some of them
and Educational Outcomes of Grandchildren needed special education (Cross & Day, 2008; Conway & Li,
2012; Pilkauskas & Dunifon, 2016; Smithgall et al., 2009).
Raised by Grandparents
More specifically, 80% of grandchildren had poor educational
Custodial grandchildren often experience significant adverse performance (Smithgall et al., 2009). Furthermore, some
childhood experiences (ACEs). As a result of these ACEs, a grandchildren were much older than their classmates in the
large number of grandchildren raised by grandparents expe- same grade level. Others requested reduced course loads but
rience socio-emotional, mental, and behavioral problems needed extensive assistance and after-school tutoring programs
(Dolbin-MacNab, 2020). It is estimated that 21% to 33% of to manage their coursework (Cross & Day, 2008).
these grandchildren have internalizing and externalizing be- Risk factors associated with grandchildren’s low educational
havior problems based on African American custodial performance included grandparents’ low education, physical and
grandchildren ranging from 2 to 16 years old (Kelley, Whitley, mental health problems, insufficient understanding of the school
& Campos, 2011). Internalizing behavior problems refer to curriculum, different educational expectations, older ways of
problems such as anxiety, depression, social withdrawal, and learning, lack of communication with teachers, a poor rela-
somatic complaints, while externalizing behavior problems tionship between grandparents and grandchild grandchild, in-
include aggressive and delinquent behaviors (Achenbach, sufficient supports and financial resources, and low quality of
1991). A study based on 1990 census data in Alameda parenting (Ingersoll-Dayton, Tangchonlatip, & Punpuing, 2020;
County, California found that custodial grandchildren expe- Lee & Blitz, 2020; Nadorff, Scott, & Griffin, 2021; Peterson,
rienced behavioral disorders, such as 31.4% with ADHD, 15% Scott, Ombayo, Davis, & Sullivan, 2019). Furthermore,
with a severe learning disability, and 11% with an emotional grandchildren’s socio-emotional, mental, and behavioral prob-
disorder (Blackburn, 2000). lems affect their educational outcomes (Kelch-Oliver, 2011;
Prior research has suggested that custodial grandchildren’s Peterson et al., 2019). More specifically, grandchildren’s learning
socio-emotional, mental, and behavioral problems are im- difficulties, attention-deficit/hyperactivity disorder, aggressive
pacted by multiple risk factors. At the grandchild level, a behaviors, and depressive and anxious symptoms may result in
number of factors were associated with children’s socio- their difficulty concentrating on schoolwork and even lead to
emotional, mental, and behavioral problems. These factors school suspensions (Kelch-Oliver, 2011; Peterson et al., 2019).
included the grandchild’s gender (e.g., being a male), race
(e.g., being White), age (e.g., being older), prior traumatic
experiences, low self-esteem, low social support, and low life
Gaps in the Literature and Research Purposes
satisfaction (Bernedo, Fuentes, & Fernández, 2008; Kelch- The socio-emotional, mental, behavioral, and educational
Oliver, 2011; Park & Lee, 2016; Yorgason et al., 2014). At the challenges facing grandchildren have adverse effects on short-
grandparent level, grandparents’ mental health and parenting and long-term child development, which calls for evidence-
practices were the two most significant predictors of grand- based interventions to address these issues. However, the vast
children’s behavioral problems (Goodman, 2012; Smith et al., majority of previous interventions have focused on improving
2018; Smithgall, Mason, Michels, LiCalsi, & Goerge, 2009). grandparents’ well-being (refer to published reviews of in-
Yet, the closeness between the grandparent and grandchildren terventions for grandparents raising grandchildren; Chan
was a protective factor for children’s behavioral problems. In et al., 2019; McCarthy, 2021; McLaughlin, Ryder, &
essence, grandchildren’s behavioral problems were found more Taylor, 2017; Sumo, Wilbur, Julion, Buchholz, & Schoeny,
often when there were poor familial relationships (Poehlmann 2018). The format of interventions for grandparents varies,
et al., 2008). In addition, grandparents’ poor health, low ed- including case management programs, support groups, psy-
ucation, low household income, and marital distress contributed choeducational interventions, and cognitive-behavioral in-
to their grandchildren’s socio-emotional, mental, and behav- terventions (McCarthy, 2021; McLaughlin et al., 2017).
ioral problems (Smith et al., 2008; Smith & Hancock, 2010), as Interventions can be community-, home-, and group-based,
well as grandfamilies’ low social support, insufficient coping with in-person and online modalities (Chan et al., 2019;
resources, and an unsupportive family environment (Kelley McCarthy, 2021; McLaughlin et al., 2017). Meta-analyses
et al., 2011; Smith, Cichy, & Montoro-Rodriguez, 2015). have indicated that interventions for grandparents have small
Regarding custodial grandchildren’s educational out- to moderate effect sizes for increasing grandparents’ social
comes, multiple indicators have been used in previous studies, support and parenting skills (Chan et al., 2019), and small
such as their academic achievement scores, grade levels effect sizes for improving grandparents’ mental, physical, and
corresponding to their ages, school engagement, and attitudes social relations (Sumo et al., 2018). A few interventions were
toward school (Cross & Day, 2008; Smithgall et al., 2009). theoretically driven, relying on the health promotion model,
Prior studies indicated that grandchildren raised by the person, environment, stress, and coping theory, and the
Xu et al. 637

resiliency model of family stress, adjustment, and adaptation Inclusion and Exclusion Criteria
(Sumo et al., 2018). In summary, these interventions aim to
address grandparents’ health concerns, decrease grandparents’ We used the following inclusion criteria to select articles: (1)
parenting stress and psychological distress, and increase articles that described and evaluated the effectiveness of inter-
grandparents’ parenting skills (Dolbin-MacNab, 2020). Al- ventions designed to address custodial grandchildren’s socio-
though improving grandparents’ well-being has positive and emotional, mental, behavioral and/or educational outcomes re-
indirect effects on the well-being of grandchildren (Smith gardless of research design and data analysis methods, and (2)
et al., 2018), more tailored interventions are needed to address articles with empirical evidence that were published in peer-
grandchildren’s specific needs. reviewed journals in English since 2000. Of note, due to a lim-
To our knowledge, no study has systematically reviewed the ited number of interventions that aimed to improve grand-
effectiveness of existing interventions on custodial grandchildren’s children’s socio-emotional, mental, behavioral, and educational
outcomes. Thus, this systematic review aims to critically appraise outcomes, we included studies regardless of their evaluation
and summarize interventions targeting grandchildren, particularly methods. The exclusion criteria were (1) articles did not target
interventions that seek to improve grandchildren’s socio- custodial grandparents and/or their custodial grandchildren; (2)
emotional, mental, and behavioral and educational outcomes. interventions did not address custodial grandchildren’s socio-
More specifically, this systematic review seeks to answer the emotional, mental, behavioral and/or educational outcomes; (3)
following research questions: (1) what interventions targeting interventions’ targeted custodial grandparents’ outcomes without
custodial grandchildren’s socio-emotional, mental, and behavioral measuring custodial grandchildren’s outcomes; (4) non-
and educational outcomes have been implemented and evalu- intervention studies; (5) non-empirical studies; (6) conceptual
ated?; and (2) what are these interventions’ effects on grand- articles; and (7) non-English articles.
children’s outcomes? As childhood development has significant
impacts on individuals’ short- and long-term outcomes, im- Article Selection and Data Extraction Process
plementing early prevention and interventions is important in
mitigating the effects of these adverse outcomes. This review will Our search yielded scholarly articles (n = 17,863) from seven
provide some directions for developing, implementing, and databases, and we further identified two additional relevant ar-
evaluating interventions that address grandchildren’s needs and ticles from Google Scholar. COVIDENCE software was used to
improve their outcomes, which eventually benefits grandchildren manage and screen all articles (Veritas Health Innovation, n.d.).
and grandfamilies in the long term. Team members conducted data extraction independently, and
the first two authors double-checked the accuracy of data ex-
traction. Extracted study characteristics included authors, study
Method country, the published year, study designs, interventions, sample
sizes, data analysis methods, measures, and findings on the ef-
Search Procedures fectiveness of interventions on socio-emotional, mental, behav-
This review followed the Preferred Reporting Items for Sys- ioral, and educational outcomes. Findings on the effect of
tematic Reviews and Meta-Analyses (PRISMA) guidelines interventions on these outcomes from quantitative, qualitative, and
(Page et al., 2021). We searched in seven databases, including mixed-methods studies were extracted for a narrative synthesis.
ERIC, Family Studies Abstracts, PubMed, PsycINFO, Social
Work Abstracts, Sociological Abstracts, and Web of Science, in
July 2021. In addition to these seven databases, we searched in
Quality Assessment
Google Scholar to identify additional articles. Comprehensive Three assessment tools, depending on the study design of each
literature searches were conducted based on the following three study, were used to comprehensively assess the quality of the
concepts: (1) custodial grandparents, (2) interventions, and (3) eight included studies. For two randomized control trial studies
grandchildren’s socio-emotional, mental, behavioral, and ed- (N’zi et al., 2016; Smith et al., 2008), the Quality Assessment of
ucational outcomes. Search terms were developed by the team Controlled Intervention Studies (14 items) was employed to
with input from a university librarian. The exact search terms examine the quality of these studies (National Heart, Lung, and
used were as follows: (grandparent* OR grandfather* OR Blood Institute, n.d.). The Quality Assessment Tool for Before-
grandmother* OR grandma OR grandpa OR grandchild*) After (Pre-Post) Studies with no control group (12 items;
AND (intervention* OR training* OR group* OR treatment* National Heart, Lung, and Blood Institute, n.d.) was employed
OR program* OR therapy* OR support* OR counseling*) to assess the quality of four pre-experimental design studies
AND (emotion* OR mental health OR mental illness OR with pre- and post-tests (Campbell et al., 2012; Leder,
behavior* problem* OR behavior* disorder* OR internalizing Grinstead, Jensen, Bond, & 2003; Littlewood, Strozier, &
OR externalizing OR depress* OR anxiety OR aggress* OR Whittington, 2014; Smith & Dannison, 2003). For the other
Educat* OR school* OR class* OR college* OR teach* OR two qualitative pre-experimental studies with one-shot case
learn* OR train* OR diploma* OR certificate* OR tutor* OR designs (Dannison & Smith, 2003; Dare, Marquis, Wenden,
achiev* OR perform* OR academic OR graduat* OR grade*). Gopi, & Coall, 2020), we employed a qualitative study
638 Research on Social Work Practice 32(6)

assessment tool (7 items), which was derived from the Mixed article’s title and abstract, and then screened each of the re-
Methods Appraisal Tool (MMAT; Hong et al., 2018). All the maining full-text articles using the established inclusion and
items across three tools were scored trichotomously as “yes” exclusion criteria. Disputes were resolved by the first author.
denoted met criterion; “no” denoted unmet criterion; “?” Articles were excluded if they (1) did not discuss custodial
denoted either “cannot determine,” or “not reported”; and grandparent-headed families (n = 8), (2) did not discuss
“N/A” denoted “not applicable.” Specific items for these grandchildren’s socio-emotional, mental, behavioral or edu-
three tools are listed at the end of the Table 1. cational outcomes (n = 8), (3) only included interventions
outcomes for grandparents (n = 16), or (4) were non-empirical
Results studies and/or conceptual papers (n = 33). Finally, eight ar-
ticles remained in our current review (see details in Figure 1).
Study Selection Result
COVIDENCE automatically removed 3,751 duplicates before
any screening took place. At the first stage, each of 14,114
Analysis of Study Design
articles’ titles and abstracts were screened, which resulted in In the current study, we reviewed eight studies, including two
14,041 articles being excluded. At the second stage, the re- randomized control trials and six pre-experimental designs (four
maining 73 full-text articles were screened, resulting in the studies with pre-post design without a control group; two studies
exclusion of 65 articles. Team members first screened each with one-shot case design). A summary of these studies is

Table 1. Results of Quality Assessment of Included Studies.

Randomized Control Trial 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 1.10 1.11 1.12 1.13 1.14

N’zi et al. (2016) Yes Yes Yes ? ? Yes Yes Yes Yes Yes Yes No Yes No
Smith et al. (2018) Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
Pre-experimental design: Pre-post design without a 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.10 2.11 2.12
control group
Campbell et al. (2012) Yes Yes Yes Yes Yes Yes Yes ? No No No N/A
Leder et al. (2003) Yes Yes Yes Yes Yes No Yes ? No Yes No N/A
Littlewood et al. (2014) Yes Yes Yes Yes Yes Yes Yes ? No Yes No N/A
Smith and Dannison (2003) No Yes Yes Yes Yes No Yes ? No No No N/A
Pre-experimental design: One-shot case design 3.1 3.2 3.3 3.4 3.5 3.6 3.7
Dannison and Smith (2003) Yes Yes Yes No Yes Yes Yes
Dare et al. (2020) Yes Yes Yes Yes Yes Yes Yes
Note. Yes” denotes met criterion; “no” denotes unmet criterion; “?” denotes either “cannot determine,” or “not reported”; “N/A” denotes “not applicable.”
Randomized control trial quality assessment questions (National Heart, Lung, and Blood Institute, n.d.): 1.1 Was the study described as randomized, a
randomized trial, a randomized clinical trial, or an RCT? 1.2 Was the method of randomization adequate (i.e., use of randomly generated assignment)? 1.3 Was
the treatment allocation concealed (so that assignments could not be predicted)? 1.4 Were study participants and providers blinded to treatment group
assignment? 1.5 Were the people assessing the outcomes blinded to the participants’ group assignments? 1.6 Were the groups similar at baseline on important
characteristics that could affect outcomes (e.g., demographics, risk factors, and co-morbid conditions)? 1.7 Was the overall drop-out rate from the study at
endpoint 20% or lower of the number allocated to treatment? 1.8 Was the differential drop-out rate (between treatment groups) at endpoint 15 percentage
points or lower? 1.9 Was there high adherence to the intervention protocols for each treatment group? 1.10 Were other interventions avoided or similar in the
groups (e.g., similar background treatments)? 1.11 Were outcomes assessed using valid and reliable measures, implemented consistently across all study
participants? 1.12 Did the authors report that the sample size was sufficiently large to be able to detect a difference in the main outcome between groups with at
least 80% power? 1.13 Were outcomes reported or subgroups analyzed prespecified (i.e., identified before analyses were conducted)? 1.14 Were all randomized
participants analyzed in the group to which they were originally assigned, (i.e., did they use an intention-to-treat analysis)?
Pre-experimental design: Pre-post design without a control group quality assessment questions (National Heart, Lung, and Blood Institute, n.d.):
2.1 Was the study question or objective clearly stated? 2.2 Were eligibility/selection criteria for the study population prespecified and clearly described? 2.3
Were the participants in the study representative of those who would be eligible for the test/service/intervention in the general or clinical population of interest?
2.4 Were all eligible participants that met the prespecified entry criteria enrolled? 2.5 Was the sample size sufficiently large to provide confidence in the findings?
2.6 Was the test/service/intervention clearly described and delivered consistently across the study population? 2.7 Were the outcome measures prespecified,
clearly defined, valid, reliable, and assessed consistently across all study participants? 2.8 Were the people assessing the outcomes blinded to the participants’
exposures/interventions? 2.9 Was the loss to follow-up after baseline 20% or less? Were those lost to follow-up accounted for in the analysis? 2.10 Did the
statistical methods examine changes in outcome measures from before to after the intervention? Were statistical tests done that provided p-values for the pre-
to-post changes? 2.11 Were outcome measures of interest taken multiple times before the intervention and multiple times after the intervention (i.e., did they
use an interrupted time-series design)? 2.12 If the intervention was conducted at a group level (e.g., a whole hospital and a community) did the statistical analysis
take into account the use of individual-level data to determine effects at the group level?
Pre-experimental design: One-shot case design quality assessment questions (Hong et al., 2018): 3.1 Were there clear research questions? 3.2 Did
the collected data allow researchers to address the research questions? 3.3 Was the qualitative approach appropriate to answer the research question? 3.4 Were
the qualitative data collection methods adequate to address the research question? 3.5 Were the findings adequately derived from the data? 3.6 Was the
interpretation of results sufficiently substantiated by data? 3.7 Was there coherence between qualitative data sources, collection, analysis and interpretation?
Xu et al. 639

Figure 1. PRISMA chart.

presented in Table 2. Two randomized controlled trials (N’Zi, Scales-Children’s Forms (PES; Ihilevich, Gleser, Gritter,
Stevens, & Eyberg, 2016; Smith et al., 2018) tracked grand- Kroman, & Watson, 1982), a modified version of the Pictorial
children’s outcomes at three time points (pre-, post-, and follow- Scale of Perceived Competence and Acceptance for Young
up) with more advanced quantitative methods (e.g., ANCOVA, Children (Harter & Pike, 1984), and the Teacher Report of Child
repeated measure t-tests, and second-order latent change models) Behavior (Schaefer & Edgerton, 1978). A few studies examined
and examined the intervention’s short- and long-term effects on the effects of interventions on grandchildren’s educational out-
grandchildren. Among pre-experimental design studies, two comes. These studies captured children’s involvement in sports
studies (Dannison & Smith, 2003; Dare et al., 2020) employed and other activities and academic performance in school. In
qualitative methods; three studies (Campbell et al., 2012; Leder addition, Littlewood et al. (2014) used Home Observation for
et al., 2003; Smith & Dannison, 2003) primarily used quanti- Measurement of the Environment (HOME) inventory (Caldwell
tative methods, such as univariate analysis, t-tests, and corre- & Bradley, 2001) to measure child learning, language, and ac-
lations; and one study (Littlewood et al., 2014) employed mixed- ademic performance during infancy and early childhood. Finally,
methods, which combined univariate and bivariate analyses and Smith and Dannison’s (2003) study measured the frequency of
qualitative case studies, to evaluate outcomes of an intervention. grandparents reading to their grandchildren using a four-point
The sample size for most studies was relatively small, ranging scale with responses “a whole lot,” “pretty much,” “sometimes,”
from 14 to 50, but three studies (Leder et al., 2003; Littlewood and “never.”
et al., 2014; Smith et al., 2018) had sample sizes larger than 100.
Regarding grandchildren’s outcomes of these included
Study Quality Assessment Results
studies, five studies used standardized measures to capture
grandchildren’s socio-emotional, mental, and behavioral out- The results of the quality assessment of included studies
comes. These measures included the Child Behavior Checklist are presented in Table 1. Two out of eight studies met all
(CBCL; Achenbach & Rescorla, 2001), the Strengths & Diffi- evaluation criteria (Dare et al., 2020; Smith et al., 2018). For
culties Questionnaire (Goodman, 2001), the Progress Evaluation two randomized control trial studies, Smith et al.’s (2018)
 640

Table 2. A Summary of Included Intervention Studies that Aim to Improve Custodial Grandchildren’s Socio-emotional, Mental, and Behavioral and Educational Outcomes.
Educational Measures
Socio-Emotional/Mental/Behavioral Measures and Intervention Effects
Author, Year, and Intervention Effects on Grandchildren’s on Grandchildren’s
Country Intervention Sample Size Sample Characteristics Data Analysis Socio-Emotional/Mental/Behavioral Outcomes Educational Outcomes

Randomized control trial

N’zi et al., The child directed interaction training Custodial grandmother, Grandmothers: Mage = 56.5; gender: ANCOVA; repeated- Measures: Child behavior Checklist Measures: N/A
(2016), program targeted custodial grandmothers including great- female = 100%; race not reported. measure t-tests (Achenbach & Rescorla, 2001), including Findings: N/A
United and great-grandmothers with their 2- to 7- grandmothers (n = 14; 7 Grandchildren: Mage = 5.2; gender: internalizing and externalizing problems.
States year-old children, including eight sessions. in an intervention group female = 50%, male = 50%; race: White Findings: Participants in the intervention
The curriculum consisted of positive and 7 in a control group) = 64%, African American = 22%, group reported that children had fewer
following skills, praise, modeling, children Hispanic = 7%, and biracial = 7% externalizing behavior problems in the
and stress, and support for caregivers and intervention group than in the control
children group between T1 and T2 (p = 0.03).
Differences between the intervention and
waitlist control groups were not detected
in grandchildren’s internalizing problems
(p = 0.97) between T1 and T2.
No significant differences between T2 and 3-
month follow-up assessments were found
for grandchildren’s externalizing behavior
problems (p = 0.69). However,
grandchildren’s internalizing behavior
problems were significantly improved
between T2 and follow-up assessments (p
< 0.001)
Smith et al., It compared three interventions: Behavioral Custodial grandmothers (n Grandmothers: Mage = 58.46; gender: Second-order latent Measures: Strengths & difficulties Measures: N/A
(2018), parent training (BPT), cognitive-behavioral = 343), including female = 100%; race: White = 44.3%, change (score) questionnaire (Goodman, 2001), including Findings: N/A
United therapy (CBT) and information only BPT (n = 115), CBT (n = African American = 43.4%, Hispanic = models internalizing and externalizing problems
States control (IOC) for custodial grandmothers 128), and IOC (n =100) 11.1%, and other = 1.2%; Findings: No significant differences in
who provided care to children between grandchildren: Mage = 7.83; gender: grandchildren’s internalizing and
ages 4 and 12. It included 10 two-hour female = 49%, male = 51%; race not externalizing problems between BPT and
group sessions reported CBT from T1 to T2 or T3 were found.
T1 to T2 comparison between CBT and IOC
found a significant difference in decrease of
internalizing (p = .03) and externalizing (p
= .01) symptoms.
T1 to T3 comparison between BPT and IOC
found a significant difference in decrease of
internalizing (p = .05) and externalizing (p
= .03) symptoms.
T1 to T3 comparison between CBT and IOC
found a significant difference in decrease of
internalizing (p = .01) and externalizing (p
= .002) symptoms
Pre-experimental design: Pre-post design without a control group

(continued)
Research on Social Work Practice 32(6)
Table 2. (continued)
Educational Measures
Socio-Emotional/Mental/Behavioral Measures and Intervention Effects
Xu et al.

Author, Year, and Intervention Effects on Grandchildren’s on Grandchildren’s

Country Intervention Sample Size Sample Characteristics Data Analysis Socio-Emotional/Mental/Behavioral Outcomes Educational Outcomes

Campbell et al., The case management program targeted Grandparents (n = 50) Grandparents: Mage = 56.7; gender: Descriptive analysis, Measures: The child behavior Checklist Measures: The activity
(2012), custodial grandfamilies, which included Grandchildren (n = 33) female = 96%, male = 4%; race: African paired t-tests (note: (CBCL; Achenbach & Rescorla, 2001), subscale assesses the
United assessing the strengths and needs of the American = 83%, other = 17%; the authors did not including internalizing and externalizing child’s involvement in
States family, developing the case management grandchildren: Mage = 9.4; gender: female conduct t-tests for problems activities (e.g., sports
plan, implementing and evaluating the plan = 54%, male = 46%; race not reported included target Findings: or recreational activities,
and outcomes. The program was provided outcomes) Fewer children’s clinically significant jobs, and chores).
by nurses and social workers, and had internalizing problems were reported at The school subscale
weekly small groups, visits, and/or phone the end of the case management program assesses the child’s
calls than at the beginning. performance in academic
Fewer children’s clinically significant subjects, involvement in
externalizing problems were reported at remedial
the end of the case management program services, failures, and other
than at the beginning school problems.
Findings: There was a slight
decrease in the activities
score after participation
in the case management
program (pre: 15.2%;
post:12.1%).
There was a decrease in the
school score after
participation in the case
management program
(pre: 6.1%; post: 3.0%)
Leder et al. The psychotherapeutic treatment targeted Grandparents (n = N/A) Grandparents: Mage = 52; gender and Descriptive analysis, Measures: The progress evaluation scales- Measures:
(2003), custodial grandchildren Grandchildren (n = 207) race not reported; correlation, t-test Children’s forms (PES; Ihilevich et al., The progress evaluation
United grandchildren: Mage = 8.8; gender: female 1982) measures therapeutic outcomes scales- Children’s forms
States = 37.2%, male = 62.8%; race: White = with subscales in several areas, such as (PES; Ihilevich et al.,
38.2%, African American = 41.5%, family interaction, getting along with 1982) has a subscale that
Hispanic = 9.7%, biracial = 7.2%, and others, feelings and mood, use of free time, measures
Native American = 2.4% problems, school, and attitude toward self. grandchildren’s school
Findings: 43.5% of grandchildren met the performance.
therapeutic goals. There were positive and Findings: There was no
statistically significant changes in six statistically significant
subscales of PES (i.e., family interaction, change in the academic
getting along with others, feelings and performance at school
mood, use of free time, problems, and subscale
attitude toward self; p < 0.05) between
pre- and post-treatment scores

(continued)
641
Table 2. (continued)
642

Educational Measures
Socio-Emotional/Mental/Behavioral Measures and Intervention Effects
Author, Year, and Intervention Effects on Grandchildren’s on Grandchildren’s
Country Intervention Sample Size Sample Characteristics Data Analysis Socio-Emotional/Mental/Behavioral Outcomes Educational Outcomes

Littlewood Kin as teachers (KAT) was a modification of Grandparents, including Grandparents: Mean age not reported; age Paired sample t-tests; Measures: N/A Measures: The home
et al. (2014), the parents as teachers program designed great grandparents (n = ranges: 20–29 = 6.9%, 30–39 = 9.8%, qualitative case study Findings: The first case study indicated that observation for
United to meet the special needs of relative 123) 40–49 = 20.8%, 50–59 = 31.9%, 60–69 analysis the grandchild with developmental delays Measurement of the
States caregivers, including custodial = 25.0%, 70+ = 5.6%; gender: female = (e.g., IQ = 67) and ADHD. The program environment (HOME)
grandparents, who raised children from 77.7%, male = 4.4%, gender not offered services to teach him new skills inventory (Caldwell &
birth to kindergarten entry. KAT specified = 17.9%; race: White = and divert some of his negative behaviors. Bradley, 2001).
consisted of eight sessions of support 34.9%, African-American = 53%, After 2 years of developmental services, Findings: There was a
groups which were held every other week Hispanic = 9.6%, and other = 2.4%; the grandchild was able to pass a significant increase in the
for caregivers, monthly home visits, grandchildren: Mage = 3.35; gender and kindergarten readiness exam and enter Infant Toddler HOME
referrals for children to appropriate race not reported kindergarten, spending half of his day in a learning subscale (p <
therapists and other services, case traditional classroom. His grandmother .001) in comparing pre-
management, and a curriculum for also learned how to provide better and post-scores.
children organization and structure in the home to There was a statistically
help with his ADHD. significant improvement
The second case illustrated the grandchild in the following early
had behavioral problems in school and was childhood HOME
often suspended, and he was referred for subscales: Learning (p <
counseling due to his frequent outbursts. .001), language (p =.001),
The kin educator provided developmental and academic (p < .001)
screenings for the youngest grandchild and in comparing pre- and
determined he had developmental and post-scores
social-emotional delays. The KAT
program made referrals and he was given a
speech therapist as well as therapeutic
services. The grandparent also received
education on the importance of age and
developmentally appropriate
surroundings. As a result, the family
became stable
Smith and The 8-14 week program targeted custodial Grandparents (year 1: n = Grandparent ages, gender, and race not Descriptive analysis Measures: Modified version of Pictorial Measures: The frequency
Dannison grandparents and their grandchildren ≤ 6 29; year 2: n = 15) reported; scale of Perceived competence and that grandparents read to
(2003), years old, and had weekly support groups Grandchildren (year 1: n = grandchildren: age range = 3-6 years; Acceptance for young children (Harter & grandchildren was
United for grandparents and playgroups for 33; year 2: n = 16) mean age, gender, and race not Pike, 1984); Teacher report of child measured on the
States grandchildren. Activities were designed to reported behavior (Schaefer & Edgerton, 1978). following scale: A whole
enhance self-esteem, improve social skills, Findings: Grandchildren reported increases lot, pretty much,
and celebrate family diversity. The in their ability to interact with peers and in sometimes, and never.
program also had educational their self-esteem. Findings: There was an
opportunities for school personnel Classroom teachers reported that increase in reading
grandchildren’s behaviors improved, frequency (pre: pretty
among those who participated in the much or a whole lot:
grandchildren’s groups. 60%; sometimes: 33%;
There were increases in interaction with never: 7%; post: pretty
grandparents, including talking more (47% much or a whole lot:
pre vs. 88% post), more outings together 87%; sometimes: 13%;
(53% pre vs. 87% post), more smiling (47% never: 0%)
pre vs. 75% post), and preparing food they
liked (33% pre vs. 87% post)
Pre-experimental design: One-shot case design

(continued)
Research on Social Work Practice 32(6)
 Xu et al.

Table 2. (continued)
Educational Measures
Socio-Emotional/Mental/Behavioral Measures and Intervention Effects
Author, Year, and Intervention Effects on Grandchildren’s on Grandchildren’s
Country Intervention Sample Size Sample Characteristics Data Analysis Socio-Emotional/Mental/Behavioral Outcomes Educational Outcomes

Dannison and The intervention was an 8-week educational Grandparents: (n = 33) Grandparent age, gender, and race not Qualitative analysis Measures: N/A Measures: N/A
Smith and support group project for custodial Grandchildren: (n = 41) reported; Findings: Grandchildren presented with a Findings: N/A
(2003), grandparents, custodial grandchildren, and grandchildren: preschool age; gender, and lack of familiarity with social play and
United school personnel. It included separate race not reported behavioral outbursts at the beginning of
States support groups for grandchildren and the program. Grandchildren’s social skills
grandparents, and an integrated group for and self-confidence improved markedly
grandparents and grandchildren. School over the 8-week period. They developed
personnel participated in in-service friendships, displayed appropriate social
meetings focused on providing behaviors, and became supportive of each
information about grandparents raising other. Regular classroom teachers also
grandchildren commented about the changed behaviors
among these grandchildren
Dare et al., The intervention included a 4-day overnight Grandparents (n = 34) Grandparent mean age not reported, but Qualitative analysis Measures: N/A Measures: N/A
(2020), leadership residential camp for custodial Grandchildren (n = 49) the highest percentage between ages 60 Findings: There were significant Findings: N/A
Australia grandchildren. The camp had activities and 69 years was 47.3%; gender: female psychosocial benefits to custodial
designed to build personal and leadership = 86%, male = 14%; race not reported; grandchildren, including an increased
skills through cooperative problem solving grandchildren: Mage = 14.4; gender: sense of belonging, normalization of
and challenges for grandchildren female = 47.8%, male = 52.2%; experiences of being raised by
ethnicity: Australian = 70.8%, grandparents, increased awareness of
Australian-Aboriginal = 22.9%, grandchildren’s situation, increased
European = 2.1%, and New Zealand = empathy for grandparents, an increased
4.2% ability to discuss their emotions and
experiences with grandparents, improved
social skills, more independence, new
relationships, and newly developed
resilience, self-confidence/efficacy, and
leadership skills
643
644 Research on Social Work Practice 32(6)

study met all evaluation criteria, while N’zi et al. (2016) did predominantly female (Campbell et al., 2012; Dare et al., 2020;
not report whether study participants, providers, and re- Littlewood et al., 2014; Smith et al., 2018). Only three studies
searchers were blinded to treatment group assignments. In reported grandparents’ race/ethnicity with most being White or
addition, N’zi et al.’s (2016) pilot study had a relatively small Black (Campbell et al., 2012; Littlewood et al., 2014; Smith et al.,
sample size (n = 14; seven were in the intervention group, and 2018). Regarding grandchildren’s characteristics, grand-
seven were in the control group), which might reduce its children’s gender distribution was mostly even (Campbell et al.,
statistical power. 2012; Dare et al., 2020; Leder et al., 2003; N’zi et al., 2016;
Among four studies utilizing pre-experimental pre-post Smith et al., 2018), and they were between preschool age to high
design without a control group (Campbell et al., 2012; school age (Campbell et al., 2012; Dare et al., 2020; Leder et al.,
Leder et al., 2003; Littlewood et al., 2014; Smith & Dannison, 2003; Littlewood et al., 2014; N’zi et al., 2016; Smith et al.,
2003), all studies except Smith and Dannison (2003) stated 2018). Grandchildren’s race/ethnicity was only reported in three
their study objectives specifically, and all identified inter- studies, including White, Black, Hispanic, and other race/
vention populations clearly, assessed outcome measures ac- ethnicity (N’zi et al., 2016; Leder et al., 2003) or Austalian,
curately, and analyzed before- and after-intervention changes Australian-Aboriginal, European, and New Zealand (Dare et al.,
appropriately. However, two of the four studies (Leder et al., 2020). Table 2 includes specific demographic information about
2003; Smith & Dannison, 2003) did not describe their study participants.
services/interventions or delivery methods thoroughly. For The included interventions utilized a variety of formats,
data analysis, two studies only presented descriptive com- such as case management programs (Campbell et al., 2012;
parisons for grandchildren’s outcomes without reporting p- Littlewood et al., 2014), support groups (Dannison & Smith,
values (Campbell et al., 2012; Smith & Dannison, 2003), 2003; Littlewood et al., 2014; Smith & Dannison, 2003),
which limited our confidence in the effectiveness of these residential camps (Dare et al., 2020), psychotherapeutic
interventions. Across these four studies, the participant re- treatment programs (Leder et al., 2003), and evidence-based
tention rate was not reported clearly, nor was the loss of manualized interventions (e.g., child-directed interaction
participants accounted for in the analysis. None utilized a training program, behavioral parent training, and cognitive-
more rigorous research design, such as an interrupted time- behavioral therapy; N’zi et al., 2016; Smith et al., 2018). Some
series design. In addition, all failed to report whether re- programs not only provided support to grandparents, but also
searchers were blinded to participants’ interventions. organized parallel play groups for grandchildren plus inte-
Among two pre-experimental one-shot case design studies, grated group sessions for grandparents and grandchildren
Dannison and Smith (2003) and Dare et al. (2020) employed (e.g., Dannison & Smith, 2003; Smith & Dannison, 2003) or a
qualitative methods to evaluate outcomes of interventions. curriculum for grandchildren (e.g., Littlewood et al., 2014).
Although the assessment tool was not originally developed for Regarding evidence-based manualized interventions, the
intervention research, it was useful in assessing the quality of child-directed interaction training program included parent–
these two studies. Dare et al.’s (2020) study met all evaluation child interaction therapy and parent-directed interaction,
criteria, while Dannison and Smith’s (2003) qualitative data which aimed to enhance the caregiver-child interactions and
were insufficient to evaluate intervention outcomes. reduce children’s behavioral problems (Eyberg et al., 2008;
Funderburk & Eyberg, 2011). Behavioral parenting training
was a modified version of the Triple P-Positive Parenting
Overview of Interventions
Program, which is a parenting support program to reduce
Among these eight studies, these interventions had different emotional and behavioral problems of children via strength-
focuses and targets. The focus of these interventions included ening parenting skills (Sanders, Kirby, Tellegen, & Day,
custodial grandparents only (e.g., Campbell et al., 2012; N’zi 2014). The cognitive-behavioral therapy in Smith et al.’s
et al., 2016; Smith et al., 2018), custodial grandchildren only (2018) study was adapted from the Coping with Caregiving
(e.g., Dare et al., 2020; Leder et al., 2003), and custodial (CWC) program, which was originally for caregivers of
grandparent–grandchild dyads (e.g., Dannison & Smith, 2003; persons with dementia (Gallagher-Thompson et al., 2002).
Littlewood et al., 2014; Smith & Dannison, 2003). For in-
terventions targeting custodial grandchildren and custodial
grandparent–grandchild dyads, the interventions primarily
Effectiveness of Interventions
focused on custodial grandchildren from infancy to middle These intervention studies had intervention outcomes re-
childhood. The length of the interventions varied, ranging garding the well-being of both custodial grandparents and
from 4 days to 14 weeks, and most of the interventions oc- grandchildren, but we did not synthesize grandparents’ out-
curred weekly or biweekly. comes in the present study because examining custodial
In terms of grandparents’ characteristics across these studies, grandparents’ outcomes was not the purpose of our research.
most grandparents were between 50 and 69 years old (Campbell We classified the grandchildren’s outcomes into two primary
et al., 2012; Dare et al., 2020; Leder et al., 2003; Littlewood et al., categories: Socio-emotional, mental, behavioral, and educa-
2014; N’zi et al., 2016; Smith et al., 2018) and were tional outcomes.
Xu et al. 645

Socio-Emotional, Mental, and Behavioral Outcomes identified the positive effects of a psychotherapeutic treat-
ment program on custodial grandchildren’s psychosocial
Two randomized control trials (RCTs) examined the short- outcomes, such as more family interaction, getting along
(i.e., post-test) and long-term (i.e., follow-up test) effects of better with others, better feelings and mood, use of free time,
these interventions on grandchildren’s socio-emotional, fewer problems, and a more positive attitude toward self, by
mental, and behavioral outcomes. N’zi et al. (2016) exam- using paired t-tests to compare pre- and post-test scores for
ined the effectiveness of the child-directed interaction training each outcome.
program on grandchildren’s internalizing and externalizing Similarly, Dannison and Smith’s (2003) qualitative nar-
behavior problems. By comparing pre- and post-test scores, ratives also identified the positive effects of an 8-week edu-
they discovered that grandchildren in the intervention group cational and support group on grandchildren’s social skills and
had fewer externalizing behavior problems than those in the behavioral problems. Particularly, these grandchildren dis-
control group; however, there were no significant differences played more appropriate social behaviors, and teachers also
in children’s internalizing behavior problems between the observed positive changes in grandchildren’s behaviors in the
intervention and the control groups (N’zi et al., 2016). In this classroom setting (Dannison & Smith, 2003; Smith &
study, the control group was a group on a waitlist and received Dannison, 2003). A 4-day overnight leadership residential
no intervention. N’zi et al. (2016) further examined the in- camp for custodial grandchildren in Australia indicated sig-
tervention effects on children’s behavioral problems between nificant psychosocial benefits of the residential camp for
the post-test and a 3-month follow-up assessment. Interest- custodial grandchildren, which was evaluated via qualitative
ingly, N’zi et al. (2016) did not find a significant change in interviews (Dare et al., 2020). Likewise, Littlewood et al.
externalizing behavior problems between post- and follow-up (2014) examined Kin as Teachers (KAT), which consisted of
assessments, while there was a significant improvement in support groups, case management, and a curriculum for
internalizing behavior problems between the post- and follow- children and relatives who cared for children. Littlewood et al.
up assessments. (2014) used two cases to illustrate services provided to two
The other RCT we included in this review compared the grandchildren and their grandparents and the positive effects
effects of three interventions: Behavioral parent training of these services on custodial grandchildren’s behavior.
(BPT), cognitive-behavioral therapy (CBT), and an
information-only control group for custodial grandmothers
(Smith et al., 2018). With a relatively large sample size (total:
Educational Outcomes
n = 343; BPT: n = 115; CBT: n = 128; control group: n = 100), Four studies examined the effect of these interventions on
Smith et al. (2018) found that there were no significant dif- grandchildren’s educational outcomes (Campbell et al., 2012;
ferences in grandchildren’s internalizing and externalizing Leder et al., 2003; Littlewood et al., 2014; Smith & Dannison,
behavior problems between BPT and CBT in pre-, post-, and 2003). Results were mixed across these four studies. Campbell
6-month follow-up tests. A comparison between CBT and the et al. (2012) did not find any positive changes in grand-
control group showed that CBT had a significant effect on children’s educational outcomes among families who par-
decreasing grandchildren’s internalizing and externalizing ticipated in the case management program. More specifically,
behavior problems by analyzing pre- and post-test scores. there was a decrease in activities (e.g., pre: 15.2%; post:
Comparison between BPT, CBT, and the control group 12.1%) and school scores (e.g., pre: 6.1%; post: 3.0%) after
showed that BPT and CBT had a significant positive effect on attending the program. In Campbell et al.’s (2012) study, about
decreasing grandchildren’s internalizing and externalizing 40% of the grandparents suggested their children experienced
behavior problems between the pre- and follow-up tests. In competency problems, particularly in activities and school
other words, receiving either BPT or CBT had positive long- performance. Campbell et al. (2012) discussed the potentially
term effects on custodial grandchildren’s behavioral outcomes negative effects of economic hardship on the grandchildren’s
when compared to receiving no intervention. However, there educational outcomes. However, Campbell et al. (2012) did
was no evidence to show whether BPT worked better than not offer any explanations for children’s decreased educational
CBT. outcomes after attending the case management program.
All pre-experimental design studies indicated that custodial Likewise, Leder et al. (2003) found that there were no
grandchildren improved their socio-emotional, mental, and statistically significant changes in the school performance
behavioral outcomes regardless of evaluation methods. For subscale between pre- and post-treatment scores. Leder et al.
example, Campbell et al.’s (2012) descriptive comparisons (2003) concluded that this was due to the design of the
indicated that fewer children had clinically significant inter- program being less focused on grandchildren’s educational
nalizing and externalizing problems at the end of the case outcomes. In contrast, Littlewood et al. (2014) found that there
management program than at the beginning. Also, case was a significant increase in custodial grandchildren’s
managers reported that grandparents had fewer complaints learning, language development, and academic performance
about grandchildren’s behavioral problems after the program by comparing pre- and post-scores using paired t-tests. Smith
(Campbell et al., 2012). In addition, Leder et al. (2003) and Dannison (2003) also found that the frequency of
646 Research on Social Work Practice 32(6)

grandparents reading to grandchildren had increased after their themselves), logistical challenges (e.g., transportation bar-
grandparents attended the case management program—from riers, childcare, and other demands), and organizational issues
60% who selected “pretty much or a whole lot” before the (e.g., inconvenient hours and locations and lack of technical
program to 87% afterward. One explanation was that the support; Dolbin-MacNab, Roberto, & Finney, 2013). All these
program offered books to grandparents at home, which may result in a small number of grandchildren enrolled in
eliminated a barrier to grandparents reading books to their interventional programs, limiting the potential for more rig-
grandchildren and increased their reading frequency (Smith & orous research designs with more advanced statistical
Dannison, 2003). analyses.
In conclusion, most interventions we identified in this
review primarily focused on improving grandchildren’s socio-
Limitations of Current Intervention Research
emotional, mental, and behavioral outcomes, and these in-
terventions, in general, had positive effects on these grand- We identified several methodological limitations present in the
children. Despite much fewer interventions or programs being studies included in the review. Out of eight included studies,
designed to improve custodial grandchildren’s educational only two utilized randomized control trial designs (N’zi et al.,
outcomes, some positive effects of educational-related ac- 2016; Smith et al., 2018). Those studies that utilized pre-
tivities on grandchildren’s educational outcomes were ob- experimental designs were largely descriptive or qualitative in
served in some interventions. nature and simply presented changes over time. An additional
limitation present in the studies was a lack of clarity around
how the outcomes were measured, including whether re-
Discussion
searchers and participants were double-blinded. Finally, all
We conducted the first systematic review of the literature on included studies did not explicitly discuss the theory of change
interventions for custodial grandchildren targeting socio- for their interventions, limiting our theoretical understanding
emotional, mental, behavioral, and educational outcomes. of how interventions affect grandchildren’s changes.
This review contributed to an existing gap in the literature as it We further conducted a quality assessment of the articles
is the first to synthesize the results of interventions that target included in the review using the NIH Quality Assessment
improved outcomes for custodial grandchildren. Findings Tools for pre-experimental designs: Randomized Control
across studies included in the review differed depending on Trials and Pre-Post Design Without a Control Group (National
the type and target of the intervention. The studies included in Heart, Lung, and Blood Institute, n.d.), and a qualitative study
this review showed that the interventions targeting socio- assessment tool derived from a Mixed Methods Appraisal
emotional, mental, and behavioral outcomes decreased in- Tool (MMAT; Hong et al., 2018). Only two studies met all the
ternalizing and externalizing behavior problems and improved quality criteria for controlled intervention (Smith et al., 2018)
social skills and self-confidence. Studies on interventions and pre-experimental qualitative (Dare et al., 2020) designs.
targeting educational outcomes, however, were less consis- This further suggests that there are significant methodological
tent. Two studies found that support group interventions for concerns (e.g., study design and data analysis) present in the
both grandparents and grandchildren increased reading fre- existing literature on interventions for custodial
quency and improved scores on one measure of language, grandchildren.
academic, and learning outcomes (Littlewood et al., 2014;
Smith & Dannison, 2003). The other two studies that targeted
educational outcomes via a case management program or
Future Directions for Research
psychotherapeutic treatment for grandchildren alone found Given the limitations in current intervention studies, we
either no significant changes (Leder et al., 2003) or a decrease identified several potential directions for future research based
in outcome measures (Campbell et al., 2012). on the results of this review. Future studies on interventions for
Given few studies used rigorous designs to evaluate custodial grandchildren should be guided by high standards
grandparents’ programs, it is important to note some chal- for quality intervention research. Further, there is a need for
lenges in implementing and evaluating such interventions in more studies on interventions for custodial grandchildren that
the field. First, recruiting grandchildren to participate in in- utilize randomized control, quasi-experimental designs with a
tervention research is not easy. These challenges may relate to control group, or interrupted time-series design. Longitudinal
the difficulty in identifying grandchildren raised by grand- studies are needed to best track change and maintenance of
parents in the community and obtaining assent from grand- change over time. Moreover, outcomes measured using reli-
children’s legal guardians due to grandparents’ lack of a legal able, consistent measures would allow for comparison of
relationship to their grandchildren (Dolbin-MacNab, 2020). In effectiveness between studies. Also, more rigorous statistical
addition to these, there are some other barriers related to the analyses, including reporting p-values and effect sizes should
implementation of interventions among this population, such be employed in these intervention studies. Finally, it is im-
as grandfamilies’ interpersonal factors (e.g., unaware of po- portant to consider participants’ retention issues and include
tential interventions and the desire to resolve things by intention-to-treat analysis. Future research that explores the
Xu et al. 647

impact of the interventions included in this review by utilizing Second, practitioners working with custodial grandchildren
a more rigorous design could provide greater insight into should develop and seek out interventions based on a theo-
whether and how these interventions benefit custodial retical framework. For example, Strong et al. (2010) proposed
grandchildren and their families. Implementing and evaluating a model for treatment that utilizes a trauma-informed lens and
more evidence-based interventions targeting custodial family systems theory to improve attachment and repair re-
grandchildren’s socio, emotional, behavioral, and educational lationships within custodial grandfamilies. Edwards and Ray
outcomes would improve the well-being of grandchildren in (2008) incorporated attachment theory into a proposed plan
the short- and long term. For instance, it would buffer the for both prevention and intervention with custodial grand-
negative effects of trauma on grandchildren’s well-being and families. In addition, a positive youth development framework
promote grandchildren’s healthy development. would be a good reference for practitioners who are interested
Of particular concern is that additional research is needed on in targeting interventions to custodial grandchildren (Edwards
the effectiveness of interventions that are developed within a & Taub, 2009). Programs that are informed by a positive youth
theoretical framework. We call for future intervention researchers development framework target the “five C’s”: competence,
to consider integrating the theory of change when designing their confidence, caring/compassion, character, and connection
interventions for custodial grandchildren and grandfamilies. (Edwards & Taub, 2009). Although these still are conceptual
Several theories have been proposed in relation to the problems models, they provide some theoretical/conceptual directions
facing custodial grandfamilies and how to address them. For for practitioners to consider.
example, both Edwards and Ray (2008) and Poehlmann (2003)
have proposed attachment theory as a framework for viewing
Limitations of the Current Review
custodial grandfamily interactions. Similarly, Strong, Bean, and
Feinauer (2010), Fruhauf et al. (2020), and Choi et al. (2016) all It is important to consider the limitations of the review when
incorporated family therapy frameworks into their conceptual interpreting and applying the findings. A primary limitation of
models of custodial grandparent interactions. To our knowledge, the review is publication bias. This review was limited to
these proposed theoretical or conceptual models for interventions articles published in peer-reviewed journals in English and
among custodial grandchildren and grandfamilies have not been therefore may have excluded case studies, studies published in
tested empirically, but these conceptual models are an important other languages, and studies that appeared in non-peer-
resource for those hoping to develop theory-driven interventions reviewed journals. In addition, registries and citations of
for custodial grandchildren and grandfamilies. Future research candidate articles were not searched; thus, we may have
could focus on testing the application of these conceptual models excluded some gray literature in this review. An additional
with custodial grandchildren and custodial grandfamilies, de- limitation is that the search was limited to interventions that
veloping new interventions for them within sound theoretical measured socio-emotional, mental, behavioral, or educational
frameworks, and testing the effectiveness of the developed outcomes specifically for custodial grandchildren. Studies that
interventions. targeted other outcomes, or that measured outcomes for
custodial grandfamilies as a whole, may have been excluded
from the review. Furthermore, all but one of the included
Future Directions for Practice
studies were conducted in the United States. Additional
We also propose some future directions for practitioners who studies from other countries may have a different cultural
work with custodial grandchildren and their grandparents. viewpoint that would add valuable knowledge to our un-
First, practitioners hoping to improve socio-emotional, derstanding of how to support custodial grandchildren and
mental, behavioral, and educational outcomes for custodial grandfamilies. Finally, there may be potential threats to the
grandchildren should take a holistic approach. The holistic internal validity of the findings of the included studies due to
approach refers to engaging multiple actors (e.g., school their research designs. These threats to internal validity may
personnel, grandparents, biological parents, and community include the use of the same instruments in pre- and post-
partners) involved in raising grandchildren and implementing surveys, multiple tests, and grandparents’ and grandchildren’s
interventions in multiple settings (e.g., community, school, maturation and attrition (Dunbar-Jacob, 2018). Also, the
and home). Findings from this review suggest that inter- generalizability and precision of this review’s findings are
ventions that focus solely on providing information to cus- limited as most included studies utilized pre-experimental
todial grandparents are less effective at addressing custodial designs.
grandchildren’s needs than those that take a holistic approach
(Campbell et al., 2012; N’zi et al., 2016; Smith et al., 2018).
Conclusion
Similarly, interventions that combine multiple forms of de-
livery and engage educators in addition to custodial grand- This review has provided insight into the effectiveness of
families seem to be most effective in improving custodial interventions that target improved socio-emotional, mental,
grandchildren’s educational outcomes (Littlewood et al., and behavioral, and educational outcomes for custodial
2014; Smith & Dannison, 2003). grandchildren. Findings of the review suggest that there is a
648 Research on Social Work Practice 32(6)

lack of rigorous empirical research on interventions for this Work Practice, 29(6), 607–617. [Link]
population, and that existing interventions lack grounding in a 1049731518798470
clear theory of change. The review further suggests that inter- Choi, M., Sprang, G., & Eslinger, J. G. (2016). Grandparents raising
ventions that take a holistic approach and those that engage grandchildren. Family & Community Health, 39(2), 120–128.
custodial grandparents and school personnel are most effective at [Link]
improving outcomes for custodial grandchildren. Researchers Conway, K. S., & Li, M. (2012). Family structure and child out-
and practitioners invested in improving the socio-emotional, comes: A high definition, wide angle “snapshot”. Review of
mental, behavioral, and educational outcomes of custodial Economics of the Household, 10(3), 345–374. [Link]
grandchildren should invest time in developing and testing in- 10.1007/s11150-011-9121-x
terventions that are grounded in a solid theoretical framework, Cross, S. L., & Day, A. G. (2008). American Indian grand families:
such as positive youth development, attachment, and family Eight adolescent and grandparent dyads share perceptions on
systems theories. These interventions should be evaluated using various aspects of the kinship care relationship. Journal of
rigorous experimental or quasi-experimental designs with ad- Ethnic & Cultural Diversity in Social Work, 17(1), 82–100.
vanced statistical methods and in-depth qualitative methods. [Link]
*Dannison, L. L., & Smith, A. B. (2003). Custodial grandparents
Declaration of Conflicting Interests community support program: Lessons learned. Children &
Schools, 25(2), 87–95. [Link]
The authors declared no potential conflicts of interest with respect to
*Dare, J., Marquis, R., Wenden, E., Gopi, S., & Coall, D. A. (2020).
the research, authorship, and/or publication of this article.
The impact of a residential camp on grandchildren raised by
Funding grandparents: Grandparents’ perspectives. Children and Youth
Services Review, 108, 104535. [Link]
The authors disclosed receipt of the following financial support for
childyouth.2019.104535
the research, authorship, and/or publication of this article: This re-
Dolbin-MacNab, M. L., Roberto, K. A., & Finney, J. (2013). Formal
search was supported by the University of South Carolina the Office
social support: Promoting resilience in grandparents parenting
of the Vice President Research Racial Justice and Equity Research
grandchildren. In Hayslip B. Jr, & Smith G. C. (Eds.), Resilient
Fund (Grant number: USCIP 80004043).
grandparent caregivers: A strengths-based perspective (pp.
134–151). New York: Routledge. [Link]
ORCID iD
9780203803905
Yanfeng Xu  [Link] Dolbin-MacNab, M. L. (2020). Interventions to support grandparents
raising grandchildren. In Wampler K.S., & McWey L. M. (Eds.),
References The handbook of systemic family therapy (pp. 479–501).
* indicates studies included in the systematic review Hoboken, New Jersey: John Wiley & Sons Ltd. [Link]
Achenbach, T. M. (1991). Manual for the youth self-report and 1991 10.1002/9781119438519.ch52
profile. Burlington, VT: Department of Psychiatry, University of Dunbar-Jacob, J. (2018). Minimizing threats to internal validity. In
Vermont. Melnyk B. M., & Morrison-Beedy D. (Eds.), Intervention re-
Achenbach, T. M., & Rescorla, L. A. (2001). Manual for the ASEBA search and evidence-based quality improvement: Designing,
school-age form and profiles. Burlington, VT, USA: University conducting, analyzing, and funding (pp. 129–143). New York,
of Vermont, Research Centre for Children, Youth and Families. NY: Springer Publishing Company.
Bernedo, I. M., Fuentes, M. J., & Fernández, M. (2008). Behavioral Edwards, O., & Ray, S. (2008). An attachment and school satisfaction
problems in adolescents raised by their grandparents. The framework for helping children raised by grandparents. School
Spanish Journal of Psychology, 11(2), 453–463. [Link] Psychology Quarterly, 23(1), 125–138. [Link]
10.1017/s1138741600004467 1045-3830.23.1.125
Blackburn, M. L. (2000). American’s grandchildren living in Edwards, O. W., & Taub, G. E. (2009). A conceptual pathways model
grandparent households. Journal of Family and Consumer to promote positive youth development in children raised by
Sciences: From Research to Practice. 92(2), 30–36. their grandparents. School Psychology Quarterly, 24(3), 160.
Caldwell, B. M., & Bradley, R. H. (2001). HOME inventory and [Link]
rd
administration manual (3 ). University of Arkansas for Medical Eyberg, S. M., Nelson, M. M., & Boggs, S. R. (2008). Evidence-based
Sciences and University of Arkansas at Little Rock. psychosocial treatments for children and adolescents with disrup-
*Campbell, L., Carthron, D. L., Miles, M. S., & Brown, L. (2012). tive behavior. Journal of Clinical Child & Adolescent Psychology,
Examining the effectiveness of a case management program for 37(1), 215–237. [Link]
custodial grandparent families. Nursing Research and Practice, Fruhauf, C. A., Yancura, L. A., Greenwood-Junkermeier, H., Riggs,
2012. [Link] N. R., Fox, A. L., Mendoza, A. N., & Ooki, N. (2020). The
Chan, K. L., Chen, M., Lo, K. M. C., Chen, Q., Kelley, S. J., & Ip, P. importance of family-focused and strengths-based approaches to
(2019). The effectiveness of interventions for grandparents interventions for grandfamilies. Journal of Family Theory &
raising grandchildren: A meta-analysis. Research on Social Review, 12(4), 478–491. [Link]
Xu et al. 649

Funderburk, B. W., & Eyberg, S. (2011). Parent–child interaction caregiver distress, family resources, and the home environment.
therapy. In Norcross J. C., VandenBos G. R., & Freedheim D. K. Children and Youth Services Review, 33(11), 2138–2145.
(Eds.), History of psychotherapy: Continuity and change (pp. [Link]
415–420). American Psychological Association. [Link] *Leder, S., Grinstead, L. N., Jensen, S., & Bond, L. (2003). Psycho-
10.1037/12353-021 therapeutic treatment outcomes in grandparent-raised children.
Gallagher-Thompson, D., Solano, N., McGee, J., Krisztal, E., Kaye, Journal of Child and Adolescent Psychiatric Nursing, 16(1), 5–14.
J., Coon, D., & Thompson, L. (2002). Coping with care giving: [Link]
Reducing stress and improving the quality of your life (Leader Lee, Y., & Blitz, L. (2020). Understanding experiences of grand-
Version). Stanford University Press. parents raising grandchildren in Malawi. International Social
Generations United. (2021). Resources: Grandfamilies: Strengths Work, 1–14. [Link]
and challenges fact sheet. [Link] *Littlewood, K. A., Strozier, A. L., & Whittington, D. (2014). Kin
grandfamilies-strengths-and-challenges-fact-sheet/ as Teachers: An early childhood education and support in-
Goodman, R. (2001). Psychometric properties of the strengths and tervention for kinship families. Children and Youth Services
difficulties questionnaire. Journal of the American Academy of Review, 38, 1–9. [Link]
Child and Adolescent Psychiatry, 40(11), 1337–1345. https:// 026
[Link]/10.1097/00004583-200111000-00015 McCarthy, L. P. (2021). Attention to gender and race in interventions
Goodman, C. C. (2012). Caregiving grandmothers and their for custodial grandparents: A scoping review. Journal of
grandchildren: Well-being nine years later. Children and Youth Feminist Family Therapy, 33(4), 1–20. [Link]
Services Review, 34(4), 648–654. [Link] 08952833.2021.1880186
childyouth.2011.12.009 McLaughlin, B., Ryder, D., & Taylor, M. (2017). Effectiveness of
Goodman, C. C., Potts, M., Pasztor, E. M., & Scorzo, D. (2004). interventions for grandparent caregivers: A systematic review.
Grandmothers as kinship caregivers: Private arrangements Marriage & Family Review, 53, 509–531. [Link]
compared to public child welfare oversight. Children and Youth 1080/01494929.2016.1177631
Services Review, 26(3), 287–305. [Link] Nadorff, D. K., Scott, R., & Griffin, R. (2021). Custodial grand-
childyouth.2004.01.002 children and foster children: A school performance comparison.
Harter, S., & Pike, R. (1984). The pictorial scale of perceived Children and Youth Services Review, 131, 106253. [Link]
competence and social acceptance for young children. Child org/10.1016/[Link].2021.106253
Development, 1969-1982. [Link] National Heart, Lung, and Blood Institute. (n.d.). Study quality as-
Hayslip, B. Jr., Shore, R. J., Henderson, C. E., & Lambert, P. L. sessment tools. Retrieved from [Link]
(1998). Custodial grandparenting and the impact of grand- health-topics/study-quality-assessment-tools
children with problems on role satisfaction and role meaning. *N’zi, A. M., Stevens, M. L., & Eyberg, S. M. (2016). Child directed
The Journals of Gerontology Series B: Psychological Sciences interaction training for young children in kinship care: A pilot
and Social Sciences, 53(3), S164–S173. [Link] study. Child Abuse & Neglect, 55, 81–91. [Link]
geronb/53B.3.S164 1016/[Link].2016.03.001
Hong, Q. N., Fàbregues, S., Bartlett, G., Boardman, F., Cargo, M., Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I., Hoffmann,
Dagenais, P., & Pluye, P. (2018). The mixed methods appraisal T. C., Mulrow, C. D., & Moher, D. (2021). The PRISMA 2020
tool (MMAT) version 2018 for information professionals and statement: An updated guideline for reporting systematic re-
researchers. Education for Information, 34(4), 285–291. https:// views. BMJ, 372, 1–9. [Link]
[Link]/10.3233/EFI180221 Park, M., & Lee, C. (2016). The effects of self-esteem and life
Ihilevich, D., Gleser, G. C., Gritter, G. W., Kroman, L. J., & Watson, satisfaction on problem behaviors among adolescents in
A. S. (1982). The progress evaluation scales: A system for grandparent-grandchild families: The mediating effect of social
assessing child and adolescent programs. Professional Psy- support. Indian Journal of Science and Technology, 9(46), 1–9.
chology, 13(3), 470–478. [Link] [Link]
13.3.470 Peterson, T. L., Scott, C. B., Ombayo, B., Davis, T., & Sullivan, D.
Ingersoll-Dayton, B., Tangchonlatip, K., & Punpuing, S. (2020). A (2019). Biggest concerns of school personnel about students
confluence of worries: Grandparents in skipped-generation raised by grandparents. Children and Youth Services Review,
households in Thailand. Journal of Family Issues, 41(2), 102, 201–209. [Link]
135–157. [Link] 004
Kelch-Oliver, K. (2011). African American grandchildren Pilkauskas, N. V., & Dunifon, R. E. (2016). Understanding grand-
raised in grandparent-headed families: An exploratory families: Characteristics of grandparents, nonresident parents,
study. Family Journal: Counseling and Therapy for Cou- and children. Journal of Marriage and Family, 78(3), 623–633.
ples and Families, 19(4), 396–406. [Link] [Link]
1177/1066480711417235 Poehlmann, J. (2003). An attachment perspective on grandparents
Kelley, S. J., Whitley, D. M., & Campos, P. E. (2011). Behavior raising their very young grandchildren: Implications for inter-
problems in children raised by grandmothers: The role of vention and research. Infant Mental Health Journal: Official
650 Research on Social Work Practice 32(6)

Publication of the World Association for Infant Mental Health, Society: The Journal of Contemporary Social Services, 90(2),
24(2), 149–173. [Link] 167–175. [Link]
Poehlmann, J., Park, J., Bouffiou, L., Abrahams, J., Shlafer, R., & Smith, G. C., & Hancock, G. R. (2010). Custodial grandmother-
Hahn, E. (2008). Representations of family relationships in grandfather dyads: Pathways among marital distress, grandparent
children living with custodial grandparents. Attachment & dysphoria, parenting practice, and grandchild adjustment. Family
Human Development, 10(2), 165–188. [Link] Relations, 59(1), 45–59. [Link]
14616730802113695 2009.00585.x
Sanders, M. R., Kirby, J. N., Tellegen, C. L., & Day, J. J. (2014). The *Smith, G. C., Hayslip, B. Jr., Hancock, G. R., Strieder, F. H., &
triple P-positive parenting program: A systematic review and Montoro-Rodriguez, J. (2018). A randomized clinical trial of
meta-analysis of a multi-level system of parenting support. interventions for improving well-being in custodial grand-
Clinical Psychology Review, 34(4), 337–357. [Link] families. Journal of Family Psychology, 32(6), 816. [Link]
1016/[Link].2014.04.003 org/10.1037/fam0000457
Schaefer, E. S., & Edgerton, M. D. (1978). A method and a model for Strong, D. D., Bean, R. A., & Feinauer, L. L. (2010). Trauma, at-
describing competence and adjustment: A preschool version of tachment, and family therapy with grandfamilies: A model for
the classroom behavior inventory. In The 86th Annual Meeting treatment. Children and Youth Services Review, 32(1), 44–50.
of the American Psychological Association, Toronto, Canada, [Link]
28 August–1 September 1978. [Link] Sumo, J. N., Wilbur, J., Julion, W., Buchholz, S., & Schoeny, M.
ED183262 (2018). Interventions to improve grandparent caregivers’ mental
Smith, G. C., Cichy, K. E., & Montoro-Rodriguez, J. (2015). Impact and physical health: An integrative review. Western Journal of
of coping resources on the well-being of custodial grandmothers Nursing Research, 40(8), 1236–1264. [Link]
and grandchildren. Family Relations, 64(3), 378–392. https:// 0193945917705376
[Link]/10.1111/fare.12121 U.S. Census Bureau. (2018). Table S10002- grandparents, 2018
*Smith, A. B., & Dannison, L. L. (2003). Grandparent-headed American community survey 1-year estimates. Retrieved from
families in the United States: Programming to meet unique https ://[Link] [Link]/c edsci/ tabl e?q=S1002&ti d=
needs. Journal of Intergenerational Relationships, 1(3), 35–47. ACSST1Y2 018 .S1 002
[Link] Veritas Health Innovation. (n.d.). Covidence systematic review
Smith, G. C., Palmieri, P. A., Hancock, G. R., & Richardson, R. A. software, Melbourne, Australia. Retrieved from www.
(2008). Custodial grandmothers’ psychological distress, dys- [Link]
functional parenting, and grandchildren’s adjustment. The In- Yorgason, J. B., Gavazzi, S. M., Kamp-Dush, C., Yarcheck, C. M.,
ternational Journal of Aging and Human Developmen, 67(4), Chang, S. L., & Stockdale, L. (2014). Mental health and dis-
327–357. [Link] rupted family processes in the lives of court-involved African
Smithgall, C., Mason, S., Michels, L., LiCalsi, C., & Goerge, R. American and Caucasian youth being raised by grandparents.
(2009). Intergenerational and interconnected: mental health and Journal of Family Issues, 35(10), 1279–1297. [Link]
well-being in grandparent caregiver families. Families in 10.1177/0192513X13481944