The promotion of speedy, permanent outcomes for foster children is a central child welfare policy goal. However, while children with intellectual disability (ID) are at greater risk for child welfare involvement, little is known about their case outcomes. This cross-sectional national study explores between-group foster care outcomes. Foster children with intellectual disability were more likely to have experienced an adoption disruption or dissolution but less likely to be reunified with a parent, primary caretaker or other family member. Implications for interagency collaboration in support of pre and post-foster care discharge support services are discussed.
That children with disabilities are a concern for the child welfare system is well documented in the literature (Berrick, 1999; Fuchs, Burnside, Marchenski & Mudry, 2010; Hanley, 2002; Hill, et al., 1990; Hughes & Rycus, 1998; Jonson-Reid, Drake, Kim, Porterfield, & Han, 2004; Kistin, et al., 2016; LaLiberte & Lightfoot, 2013; Lightfoot, Hill & LaLiberte, 2011; Mitchell, Turbiville & Turnbull, 1999; Shannon & Agorastou, 2006; Shannon & Tappan, 2011; Sullivan & Knutson, 2000). Children with intellectual disability (ID) are specifically identified as a population at higher risk for child welfare system involvement due to increased risk of maltreatment in the form of abuse or neglect and by specific forms of abuse and neglect (Slayter & Springer, 2011; Vig & Kaminer, 2002; Westat, Inc., 1993). Crosse, Elyze and Ratnofsky (1993) found that children with ID were two times more likely to experience physical and sexual abuse. In a national study on children with ID in the child welfare system, Slayter and Springer (2011) found that children were more likely to be removed from family members due to relinquishment, parental death, their disability, sexual abuse, or physical abuse.
Looking at what we know of child welfare placements for children with ID, they were more likely to be placed in restrictive foster placements when compared to children without ID (Schmidt, Cunningham, Dalton, Powers, Geenen and Orozco, 2013). Such placements are not ideal, given disability policy goals related to the need for community inclusion and normalization. Although placement practices should be governed by the dictates of the Adoption and Safe Families Act of 1997—namely, that placement decisions are based on the least restrictive setting possible that will meet a child's safety needs, promote stability and support family reunification—specific disparities are noted. Slayter and Springer (2011) note that children with ID experienced more placement instability (number of settings) than did their counterparts, and were more likely to be placed out-of-state. Furthermore, despite the fact that the ideal goal for most foster youth is family reunification, children with intellectual disability were 49% less likely to be placed with family in a trial home visit. Although kinship foster care is often regarded as a good placement, children with intellectual disability were 61% less to live in this type of setting.
Children in foster care may experience trauma due to a variety of causes—the maltreatment that leads to system involvement, child protection investigations, their removal from families and/or caregivers, their experiences in foster care and the inevitable challenges that youth face when transitioning out of care (Hill, Lightfoot & Kimball, 2010; Slayter & Springer, 2011). These potential sources of trauma, along with the impact of living in the child welfare system, are detrimental (Strickler, 2001; Weaver et al., 2005). Having a lack of a sense of belonging, disrupted family identity, attachment disorders, emotional distress, and the stigma of living in foster care may all be experienced by children with ID (Barahal, Waterman, & Martin, 1981; Mallon & McCartt Hess, 2005; Slayter & Springer, 2011). Research also suggests that children who spend time in the child welfare system may be less likely to establish the personal connections that will guide them into adulthood upon “aging out” of foster care (Renne & Mallon, 2005). For children with ID, the transitional age period is already fraught with disability-specific challenges in finding and securing disability-related services in the “turning 22” time period, even without child welfare involvement (Hill, Lightfoot & Kimball, 2010).
All of these potentially negative experiences may function together to create cumulative risk for negative life outcomes, including mental illness, lower levels of education, and limited employment—all of which present threats to community inclusion, a central disability policy goal (Pecora et al., 2007; Slayter & Springer, 2011). All of the risk factors associated with foster care placement in general are likely to be equally present for children with intellectual disability (Slayter & Springer, 2011). It is also possible that foster children with intellectual disability are actually especially vulnerable to the potential socioeconomic and psychological stressors associated with foster care that may particularly impede community inclusion (Slayter & Springer, 2011).
In order to add to our knowledge of this population's experience in the child welfare system, this study will explore three areas related to foster care outcomes. First, an analysis of the rate of previous disrupted or dissolved adoptions among youth with and without ID will be presented. Second, an analysis of the demographic differences of youth with ID who were and were not discharged from care will be presented in order to assess for demographic or disability-related disparities. Third, an analysis of the foster care outcomes of youth with and without intellectual disability will be presented. Exploration of these data will assist policymakers and professionals in the intellectual and developmental disabilities sector in identifying areas for collaboration with child welfare professionals towards the goal of improving the case outcomes—and lives—of foster youth with ID.
Adoption Disruption and Dissolution
One of the central concerns for all foster children regardless of ID relates to the need for “permanency planning.” This term originated as part of the Adoption Assistance and Child Welfare Act of 1980 (Public Law 96-272) and emphasizes a child's right to a life lived in a nurturing (ideally) family setting that has the potential for a life-long relationship. The increased availability of funds to support special needs adoptions has spurred on an increase towards more special needs adoptions (Child Welfare Information Gateway, 2012). Unfortunately, this has been accompanied by an increased rate in post-adoptive problems, including an increase in disrupted and dissolved adoptions (Reilly & Platz, 2004). Failed permanent placements, such as those involving adoption disruptions or dissolutions, will likely have negative impact on a child. For example, in addition to the experience of a disrupted relationship with a caregiver, children face changes to their living environment, community, school, and peer group relations. Adoption disruptions or dissolutions may be an important determinant of other psychosocial outcomes. Therefore, examining the rate at which this phenomenon exists is a well justified first step in understanding more about the long-term impact of a failed permanent placement.
A small percentage of adoptions disrupt (end before legal finalization) or are dissolved (ended after legal finalization). A range of studies based in the United States consistently report adoption disruption rates between 10% and 25%—depending on the population, the length of the study, and other factors (Goerge, Howard, Yu, & Radomsky, 1997; Festinger, 2002; Festinger, 2005). However, using the same data set as the present study, Festinger and Maza (2009) report that of all the children who entered foster care for the first time and who then exited the foster care system in FY 2005, 0.5 percent had previously dissolved adoptions. The present study will add to this literature by exploring how this phenomenon impacts youth with intellectual disability in comparison to their peers without disabilities.
Little is known about the factors associated with adoption disruption or dissolution (Child Welfare Information Gateway, 2012). One study determined that the rate of children returning to the custody of the State from an adoption increased with the age of the child (at adoption) and was more likely for male or non-Hispanic/Latino children (Goerge et al., 1997). Other research suggests that families who adopt children with special needs from the child welfare system face enormous struggles and limited access to needed services, such as respite care (Festinger, 2002; Reilly & Platz, 2004). Further, research with adoptive families suggests that the two most common barriers to a successful adoption are a lack of information about how to access services and the cost of those services (Festinger, 2002; Soderlund, Epstein, Quinn, Cumblad, & Petersen, 1995). The provision of resources for these service needs is the responsibility of the state entities that govern adoption procedures and the contract agencies providing direct services. While the present study cannot examine these factors, it can establish whether there may be a disability-related disparity in the experience of this phenomenon.
Demographic Differences in Foster Care Exit
Existing research suggests that child welfare experiences differ for children based on some of their demographic characteristics. For example, research on adoption preferences suggest that girls are favored over boys, that younger children are favored over older children, and that African American/Black children are not as favored as their counterparts (Baccara, Collard-Wexler, Felli & Yariv, 2014; McKendry, 2013). Most notably, much attention has been paid to the presence of racial and ethnic disproportionality and disparity as a feature of the child welfare (Tilbury & Thoburn, 2009). Indeed, Barth and Needell (1997, p. 2) indicate that “race is perhaps the most powerful background factor in all our child welfare services research.” For example, African American/Black youth are reported to be more likely to be placed in out-of-home care, and more likely to stay longer (Courtney et al., 1996; Hill, 2006; Needell, Brookhart, & Lee, 2003). A range of factors are related to racial disproportionality and disparity in the child welfare system, including the discriminatory practices of mandated reporters (such as law enforcement professionals, educators and social workers); institutional racism and system biases such as a lack of cross-cultural competence; culturally inappropriate or inaccessible service delivery; and discriminatory practices of child welfare workers (Cross, 2008; Hines, Lemon, Wyatt, & Merdinger, 2004; Tilbury & Thoburn, 2009). Other factors include poverty, language barriers, childrearing differences, and discrimination in child and family assessments (Chand & Thoburn, 2006; Thoburn et al., 2005). The identification of any demographic differences in foster care exit will assist practitioners in targeting disparity-reducing interventions toward potential adoption resources and systems of care.
Foster Care Outcomes
Long-term placement in foster care is generally not considered a good outcome for any child. While concerns about the impact of permanency planning efforts for children with disability at the transitional age stage are noted (Hill, Lightfoot & Kimball, 2010), little is known about the outcomes of permanency planning for all foster children with ID (Martin, Barbee, Antle & Sar, 2002, March). Youth exiting foster care before age 18 have access to a range of child-focused services that are not available to those exiting after the age of 18. Given that the social welfare system approaches the provision of services to these two population groups differently, the present study will examine foster care outcomes for each group. The ways in which children with ID exit the child welfare system is a matter of concern for professionals in the field of ID services who may be supporting children in child welfare custody (Simms, 1991). These data will assist practitioners in understanding where to target resources for youth with ID who are aging out of foster care. If most youth are emancipated, for example, the focus might need to be on the provision of independent living services, as opposed to respite services for guardians.
While little is known about the foster care outcomes of children with ID under the age of 18, a bit more is known about the transitional outcomes of child welfare cases involving transitional-aged youth with ID (Hill, Lightfoot & Kimball, 2010; Hill, 2009; Pecora, et al., 2007). Youth in the transitional-aged population are undergoing what Jeffrey Arnett (Arnett, 2000) refers to as the “emerging adulthood” phase of development. This is a particularly important phase for foster children as age eighteen is the age at which children must either “sign on” for continued foster care services or “sign out” of foster care as an adult. It is apparent that youth with disabilities in general have been identified as a priority for assistance at the transitional stage (Hogansen, Powers, Geenen, Gil-Kashiwabara & Powers, 2008). Hill, Lightfoot & Kimball (2010) surveyed independent living coordinators in child welfare, and found that although children with disabilities are being served in transition programs, there were few targeted services, a lack of collaboration with community agencies and a need for more cross-system information sharing. Evidence of gender disparities in the life outcomes (i.e. employment, wage level, job type) of girls with disabilities have also been noted (Coutinho, Oswald & Best, 2006; Doren & Benz, 2001; Hazazi, Johnson, Gordon & Hull, 1989). The literature establishes that children who are both in foster care and in special education services (i.e., children with disabilities) are at a “compounded disadvantage” during the phase between high school and young adult life (Powers et al., 2012).
In order to set a foundation for further development of practice guidelines and inter-system collaboration around foster children with ID, this study focuses in on the child welfare outcomes of foster children with ID. Specifically, this study has four aims. First, this study aims to determine the prevalence of disrupted adoptions among current foster children with and without ID. Second, this study aims to explore the demographic differences of foster children with ID who were and were not discharged from foster care within 2012 in order to determine whether certain demographic characteristics are associated with foster care exit. Third, this study aims to explore case outcomes among foster children aged 0-17 with and without ID. Fourth, this study aims to explore case outcomes among foster youth who are transitional-aged youth aged 18+ with and without ID.
This cross-sectional, exploratory study drew on data from the Adoption and Foster Care Reporting System (AFCARS) for foster youth in 50 states, the District of Columbia, and Puerto Rico. AFCARS includes case level data entered by child welfare caseworkers on all foster youth in the custody of state child protective services either via institutional placement, family foster care or supervision. AFCARS was developed to collect uniform, reliable information on youth who are under the responsibility of the child welfare system for placement, care, or supervision. The collection of AFCARS data is mandated by section 479 of the Social Security Act. According to the Children's Bureau, “the requirements for AFCARS are codified in Federal regulation at 45 CFR 1355.40” (Children's Bureau, 2009: 1). Further:
AFCARS was established to provide data that would assist in policy development and program management. Data can be used by policymakers at the Federal, Tribal, and State levels to assess the reasons why children are in foster care and to develop strategies to prevent their unnecessary placement into foster care. Specifically, the data include information about foster care placements, adoptive parents, and length of time in foster care, and make it possible to identify trends in particular geographic areas. (Children's Bureau, 2009: 1)
When relying on administrative data, a type of data source derived for administrative versus research purposes, construct validity and reliability may be limited. While the potential challenges to the validity and reliability of the variables in the AFCARS data file are noted to include state by state differences in reporting and the validity of diagnostic conditions, a number of national efforts were initiated to address this concern (Slayter & Springer, 2011). Efforts to improve the data file's reliability and validity were supported through financial support for improvements to each state's statewide automated child welfare information systems (SACWIS) (Government Accounting Office, 2003). In 2009, the Children's Bureau developed the AFCARS Assessment Review process, which targets the improvement of these data (Children's Bureau, 2009) as follows:
The Children's Bureau created the AFCARS Assessment Review process to ensure the accuracy and reliability of the foster care and adoption data. During these reviews, the Federal review team assesses the efficiency and effectiveness of a title IV-E agency's data collection, extraction, and reporting processes, and provides intensive technical assistance to title IV-E agency staff responsible for those processes. (Administration for Children and Families (ACF), 2012)
Documentation provided by the ACF states that “the AFCARS data submissions are subject to a minimal number of edit checks, as listed in Appendix E of 45 CFR 1355. These edit checks are able to determine substantial compliance only for the timely submission of the data files; the timeliness of data entry of certain data elements; and whether the data meets a 90 percent level of tolerance for missing data and internal consistency” (ACF, 2012). Further, the Assessment Review guidance document indicates that “these edit checks are an important first step to ensuring the collection of quality data…there may still be a need to improve its AFCARS data for accuracy, reliability, and validity. The edit checks are not able to determine whether …an agency is submitting accurate and reliable data to the Children's Bureau that meet all of the AFCARS requirements” (ACF, 2012) The results of the AFCARS Assessment Review process do not appear to be publically available data. Given the above, a statement about the general reliability and validity of data collected in AFCARS is impossible to make.
Drawing on AFCARS data about children with and without diagnosed clinical disabilities, but excluding those whose disabilities were not yet determined, a group of 527,461 children were identified. The sample of youth with ID made up 2.2% of the group (n = 11,420) and a comparison group of youth without ID comprised 97.8% of the group (n = 516,041). A second stage of data analysis eliminated any youth without ID who had another diagnosed disability (i.e. emotional disturbance, physical disability, visual/hearing disability or other medical condition, described below) in order to create a fair comparison group. This approach resulted in the identification of a sample of youth with ID (n = 11,420, 3.1% of the group) and a comparison group of youth without ID or any other disabilities (n = 359,353 96.9% of the group). Using these groups, univariate and bivariate analyses were conducted. These analyses included odds ratios and independent samples t tests, and, where appropriate adjusted odds ratios. Both unadjusted and adjusted odds ratios were derived from logistic regression analysis. Age and gender-adjusting were used for select outcome variables given existing knowledge about the impact of age and/or gender on foster care-related outcomes (Guterman & Taylor, 2005). Given the large size of the AFCARS data set, odds ratios are reported as a standard measure of effect (Haddock, Rindskopf, & Shadish, 1998). As Chen, Cohen, & Chen (2010) note, “the odds ratio (OR) is probably the most widely used index of effect size in epidemiological studies” (p. 860). Also, as Haddock, Rindskopf & Shadish (1998) state “there is widespread consensus among statisticians that … the most appropriate measure of effect size from a four-fold table is the odds ratio” (p. 340). Therefore, both absolute and relative measures of effect are reported in this manuscript.
Children included in the sample were categorized as having a clinical diagnosis of “mental retardation” as documented in AFCARS, referred to hereafter as ID. In this data file, mental retardation was defined as ‘‘significantly sub-average general cognitive and motor functioning existing concurrently with deficits in adaptive behavior manifested during the developmental period that adversely affect a child's/youth's socialization and learning'' (Adoption and Foster Care, 2012, p. 5). Further, AFCARS included children in the category if they had a clinically validated diagnosis of Down syndrome, borderline intellectual functioning, hydrocephalus, microcephaly, or mental retardation (mild, moderate, severe, profound), although individual counts of these conditions were not accessible given the way in which the data were reported in AFCARS. The notation of having one of the above-described diagnoses originally occurs when a child welfare caseworker makes such a note in his or her state's case management data system. This notation is supposed to be made once documentation of the condition is obtained from a medical doctor or other appropriate source (e.g., an Individual Education Plan or neurological report). Each state reports data on the children in their care to the federal government on an annual basis for inclusion in the AFCARS data file (Administration for Children & Families, 2012).
We examined two sets of variables: demographic and foster care case outcome-related data. Demographic characteristics measured in the study were age, gender, race, and Hispanic/Latino/a ethnicity. The variable for age was operationalized by AFCARS as “age of child at the end of the federal fiscal year or at exit from foster care.” Other comorbid disabilities (e.g., emotional disturbance, physical impairment, or a combined variable for visual/hearing impairment) and ‘‘other medical conditions'' (including, for example, acquired immune deficiency syndrome, diabetes, and epilepsy) were also measured among youth with ID. The diagnostic conditions included in each of these categories are listed in detail by the Administration for Children and Families (Administration for Children and Families, 2012).
Demographic data and clinical data on comorbid disabilities are reported in Table 1. AFCARS variables followed standard reporting conventions for race and ethnicity with the exception of the use of separate categories for “Asian” and “Hawaiian/Pacific Islander.” These variables were used to create a variable for Asian/Pacific Islander in keeping with reporting norms. Foster children with ID were, on average, older than their counterparts by three to four years and were less likely to be female. Further, foster children with ID were slightly less likely to be Asian/Pacific Islander or Latino/a, but essentially equal vis-à-vis racial categories of American Indian/Alaskan Native, White or Black/African American. 23.9% of youth with ID had no other clinically diagnosed disability, but 76.1% of this group did have another clinically diagnosed disability.
AFCARS includes four other disability-related variables. First, AFCARS operationalizes physical disability as a situation in which a youth has “a physical condition that adversely affects the youth's day-to-day motor functioning, such as cerebral palsy, spina bifida, multiple sclerosis, orthopedic impairments, and other physical disabilities” (ACF, 2007, 5). Second, AFCARS groups visual and hearing impairments together, and operationalizes them as “having a visual impairment that may significantly affect educational performance or development; or a hearing impairment, whether permanent or fluctuating, that adversely affects educational performance” (ACF, 2007, 5). In the present study, the term “visual and hearing disabilities” is used. Third, AFCARS includes a variable for emotional disturbance which is categorized as “a condition exhibiting one or more of the following characteristics over a long period of time and to a marked degree: An inability to build or maintain satisfactory interpersonal relationships; inappropriate types of behavior or feelings under normal circumstances; a general pervasive mood of unhappiness or depression; or a tendency to develop physical symptoms or fears associated with personal problems.” The term includes persons who are schizophrenic or autistic. The term does not include persons who are socially maladjusted, unless it is determined that they are also seriously emotionally disturbed (AFCARS, 2007, 5-6). Fourth, AFCARS includes a variable for “other medically diagnosed condition requiring special care” which is operationalized as “conditions other than those noted above which require special medical care such as chronic illnesses. Included are children diagnosed as HIV positive or with AIDS” (AFCARS, 2007, 6).
Foster care case outcome characteristics included two groupings of variables: foster care experience and foster care case outcome data. First, foster care experience variables included “total days stay in foster care, all epidodes (LifetimeLOS),” and “child ever adopted” (Administration for Children and Families, 2012). Variables measuring the foster care case outcomes of children living in foster care included eight outcome types. The definitions that follow are all from the AFCARS Foster Care File Codebook (2012) data documentation. Children either remained in foster care at the end of 2012, or were discharged. The first discharge type is framed as family reunification, in which a child returns to the parent or primary caregiver from whom they were removed or who voluntarily put them in foster care. The second discharge type involved a child being discharged to a situation in which they were living with other relatives. As is indicated by the AFCARS Codebook, “It is important to note that the discharge reason of living with other relatives only applies to children discharged from foster care to the care of a relative. It does not apply to children in the foster care system in out-of-home placements with relatives, and it does not apply to children discharged from foster care to a legal guardian, even if the legal guardian is a relative.”
The third discharge type was adoption, in which a child is adopted from the child welfare system. AFCARS also includes two other variables that are relevant to adoption as a foster care outcome. One variable indicates whether an adopted child receives Title IV-E Adoption Assistance funds – funds that are often considered an adoption incentive. Another variable measures the number of days a child waited between their mother's termination of parental rights date and the day of their exit from foster care to adoption. Both of these variables are reported upon in the present study as they are contextually relevant. The fourth type of discharge was emancipation, in which a child is emancipated to their own custody either at age 18 or via a legal emancipation process at an earlier age. The fifth type of discharge involved being released to a legal guardian – which might or might not involve a family member. The sixth type of discharge involved a transfer from the child welfare authority to another state agency, such as a Department of Developmental Services. The seventh type of discharge involved a situation in which the child was noted to have run away at the end of 2012. The eighth type of discharge was a situation in which the child died during 2012.
Children with ID constituted 3.2% of the population of youth in foster care for whom confirmed disability diagnoses were available. Foster children with ID spent, on average, 1,631.6 days in care over their lifetime (SD = 709.4.) as compared to foster children without ID, who spent, on average 609.0 days in care (SD = 1,484.1) [See Table 2].
Our first study aim sought to explore information about adoption disruption – in other words, the prevalence of children who experienced a failed previous child welfare placement via adoption. Findings related to a history of adoption disruption before the present foster care episode suggest that 6.70% of all foster children with ID (n = 765) were reported to have been adopted previously (i.e. returned to foster care after a disrupted adoption) as compared to 1.90% of foster children without ID (n = 6,700). Foster children with ID were 3.77 times more likely to experience this reality when compared to foster children without ID. However, when age and gender were included as adjusters, this odds ratio dropped to 1.93 (confidence intervals: 1.78–2.08, p < .001). [See Table 2].
Demographic differences by outcome status
Our second study aim sought to learn about demographic differences between foster children with ID who were and were not discharged from the child welfare system in 2012 in order to assess for the presence of racial or gender disparities. 29.8% of children with ID exited foster care in 2012 (n = 3,407) and 70.2% remained in foster care (n = 8,013). Findings are presented in Table 3. Children with ID who exited care were essentially equal in age to those who remained in care, and were equally likely to be female. With respect to race and ethnicity, children with ID who exited foster care in 2012 were equally likely to be American Indian/Alaskan Native and Asian/Pacific Islander. Those who exited foster care in 2012 were also almost equally likely to be White or Latino but were 17% less likely to be African American/Black as compared to children who remained in foster care, suggesting a very slight disparity for African American children. With respect to comorbid disability or illness, children with ID who exited foster care in 2012 were slightly less likely to have an emotional disturbance, and essentially equally likely to have a visual/hearing impairment, a physical impairment, or another diagnosed medical condition.
Foster care outcomes for youth aged 0–17
Our third study aim sought information about foster care outcomes for foster children aged 0–17 with and without ID who exited the system in 2012. Among foster children with ID aged 0-17, 23.4% (n = 2,147) were discharged from foster care in 2012 as compared to 36.9% of their counterparts without ID (n = 126,174). Foster children with ID were 46.00% less likely to exit foster care as compared to their counterparts without ID – a finding that remained essentially the same when age and gender adjusters were added to the analysis (AOR = 0.53, confidence intervals 0.51–0.56, p < .001). Among the foster children who remained in the child welfare system, 30.70% were the subject children on a completed termination of parental rights case as compared to 14.50% of their counterparts (OR = 2.60, confidence intervals 2.47–2.74, p < .001). To be the subject child on such a case indicates that the child is technically free to be adopted (among other outcomes).
Findings related to the foster care outcomes for children with and without ID that exited the child welfare system in 2012 are presented in Table 4 and data discussed below refer to adjusted odds ratio findings. Family reunification was an outcome for 42.80% of foster children with ID, who were 48.00% less likely to have this outcome as compared to their counterparts. Another type of family reunification is discharge to the custody of a relative other than a parent or primary caretaker (e.g. in a situation other than guardianship). This was an outcome for 6.50% of foster children with ID, but this group was 32.00% less likely to experience this outcome as compared to their counterparts.
Thirty six percent of foster children with ID were adopted in 2012, and were almost two times more likely to be adopted. On average, foster children with ID waited 709.84 days (SD = 772.09) between the termination of their mother's parental rights and their exit to adoption—compared to 405.53 days (SD = 475.74) for their counterparts. Foster children with ID who were adopted out of the child welfare system were slightly more likely to receive Title IV-E Adoption Assistance funds (OR = 1.22, p < .01). 32.70% of foster children with ID who were adopted in 2012 received these funds, as opposed to 28.40% of their counterparts without ID.
Only two percent of foster children with ID aged 0–17 were emancipated in 2012—but this was almost three times the rate of their counterparts without ID. Just over 8 percent of foster children with ID were discharged into a situation where they were being cared for by a guardian, but no statistically significant differences were noted between groups.
Having an outcome which involved being transferred from the child welfare authority to another agency was equally likely for children with and without an ID (just over 1% for the sample). Less than a third of one percent of foster children with ID were reported as runaways at the end of 2012, and there was no statistically significant difference noted between groups. Finally, while just over one percent of foster youth with ID died while in foster care in 2012 (n = 30), they were over 15 times more likely to die than were their counterparts without ID (0.10%, n = 143).
Foster care outcomes for transitional-aged youth
Our fourth study aim sought information about foster care outcomes for foster children in the transitional-aged range of 18+ with and without ID who exited the system in 2012. 2,227 foster youth were aged 18 or higher. Although there were very small numbers of people aged 23 to 47 (n = 49), it is postulated that people over aged 22 who were listed in as being in foster care may have been outliers who remained in the child welfare system in lieu of transfer to another system. Among foster children with ID aged 18+, 56.60% (n = 1,260) were discharged from foster care in 2012 as compared to 64.20% of their counterparts without ID (n = 11.308). Foster children with ID were 28% less likely to exit foster care as compared to their counterparts without ID (OR = 0.72, confidence intervals 0.66–0.79). This finding remained essentially the same when age and gender adjusters were added to the analysis (AOR = 0.78, confidence intervals 0.71–0.85, p < .001). Findings on foster care outcomes for transitional-aged youth with and without ID are presented in Table 5. Foster children with ID were 51.00% less likely to be reunified with family as compared to their counterparts without ID—although reunification is a less expected option in this age range. No between group differences were noted with respect to the outcomes of living with another relative, being adopted, being emancipated or having run away. However, foster children with ID were almost 4 times more likely to be discharged into a guardianship situation or to be transferred to another agency as compared to their counterparts. Finally, foster children with ID were five times more likely to have died in foster care as compared to their counterparts without ID, although the n was less than or equal to five children in each group.
Findings reported upon in the present study provide a helpful set of baseline data about the foster care case outcomes of child welfare-involved children with ID that can inform child welfare and disability policymakers and practitioners in their work to improve the chances of positive transitions and further community inclusion for this population.
Before discussing the implications of the major findings of this study, a note on the prevalence of ID in the AFCARS data set is warranted. As reported above, 2.2% (n = 11,420) of all youth in this data set were noted to have a confirmed diagnosis of ID. This is slightly below Slayter and Springer's (2012) findings which indicated that Youth with ID constituted 2.6% of the entire foster care population represented in AFCARS. While, there are no national estimates for the prevalence of children with ID living in foster care (Mitchell, Keller, & Loyek, 2005), other local population-based research gives some indication of what the prevalence range should be. Previous findings reported in a study of foster youth in the care of the child welfare system (Geenen and Powers, 2006) found that 4.0% of their sample (aged 13–21) had a diagnosis of ID. As this study provided a good estimate for comparison by comparing foster care records and/or school records for youth with developmental disability, a methodology that suggests a potential undercount in the present study.
Increased Likelihood of Previous Adoption Disruption and Dissolution
Findings related to our first study aim suggest that foster children with ID are 1.9 times more likely to have experienced an adoption disruption prior to the current foster care episode. Although less than 10% of this group experienced this condition, these findings support existing research about the experiences of children with disabilities (Festinger, 2005). Adoption disruption may have an especially problematic impact on youth with ID vis-à-vis their ability to form healthy attachments with their caretakers (Festinger, 2005). Previously adopted foster youth with ID may actually be at increased risk of poor outcomes while in the care of the child welfare system and in transitional-age placements. Adoption disruption can result from anything in between a poor placement fit to the death of an adoptive parent, and can have an impact on the possibility of healthy attachment to longer-term caregivers for foster children with ID (Ciccetti & Toth, 2005; Festinger, 2005). Research (ideally in the form of case record reviews) about the reasons for adoption disruption with children with ID specifically is needed in order to determine implications for both adoptive parent training programs and post-adoption supports that would function as disruption prevention mechanisms. The development of such training and support programs are areas in which there is ample opportunity for expertise from both the child welfare system and developmental disability service providers would be ideal.
Potential Demographic Risk Factors for Remaining in Foster Care
Several findings from our second study aim suggest the presence of demographic risk factors for remaining in care. For example, children with ID who exited foster care in 2012 were less likely to be American Indian/Alaskan Native, Asian/Pacific Islander and White, but roughly equally likely to be Latino/a or African American/Black as compared to children who remained in foster care. Drawing on our understanding of racial disparities in foster care outcomes, researchers should consider approaching future disparities research with an intersectional lens vis-à-vis the presence of disability (Gourdine, 2013). In future research, intersectionality may also be an important lens to consider with respect to the ways in which comorbid disability or illness impacted the exit of children with ID from foster care. A more in-depth examination of the factors associated with these potential exit disparities is warranted in order to develop intervention strategies designed to equal the playing field for children in these racial and ethnic groups.
Mixed Results on Foster Care Discharge Outcomes for Younger Youth
Our third study aim explored foster care outcomes for youth aged 0–17. The ideal goal of the majority of child welfare cases is the reunification of children with their family. Unfortunately, youth with ID were 41% less likely to have this outcome. These findings have implications for developmental disabilities service providers working to extend original circles of support beyond immediate family members before and during a foster care episode in order to create options for placement if a child welfare case is initiated. This could be accomplished, possibly, by targeting extended family members for the provision of respite services to parents in need, especially for families who already have involvement with the child welfare system. Additionally, given that children with ID were less likely to be discharged to live with non-parent family members than were their counterparts, caseworkers should engage in a reconsideration of how and whether formalized ‘kinship care' could be facilitated more often. Future research should examine the reasons why reunification with other relatives occurred at lower rates for children with ID would also be helpful in developing strategies to reunify foster children with ID with their families more often.
For youth whose parents' rights have been terminated, adoption is usually an ideal goal. Analyses suggest that foster children with ID were almost two times more likely to be adopted than were their counterparts without ID. This supports existing research related to the pattern of adoption of children with disabilities (Rosenthal, Groze & Aguilar, 1991). Given Hill and Koester's (2015) research, this bodes well for children with ID as those who were adopted were shown to have accurate and complete disability identification in their IEPs, more opportunities to build community and peer relationships, and higher-quality IEPs than they did after their adoption. Child welfare researchers should consider studying the characteristics of children with ID that were adopted in order to identify protective factors for adoption. This research would ideally look at both children's clinical characteristics and the types of supports present in the child's life at exit from care. Future research could also examine the relationship between adoptions of this population of children and the availability of adoption subsidies in order to identify promising approaches to the use of financial incentives. However, longitudinal research to track the long-term well being of children with ID in adoptive homes that receive subsidies would also be warranted, given the presence of higher risk of adoption disruption and dissolution.
This study has two limitations that should be considered in order to contextualize study findings. First, the definition of “mental retardation” used by the Adoption and Foster Care Analysis and Reporting System (AFCARS) is limited in that it does not allow for specification of findings by level of ID. This measure may also include diagnoses other than ID, as noted above. Further, prevalence data derived from the use of this variable are likely to be an undercount of the actual prevalence of ID, as child welfare case workers who enter data about their clients might not enter this diagnostic condition for all of their clients. Additionally, it is also possible that the underreporting of this condition may be a function of caseworkers' choosing a label for the disability type they view as most dominant vis-à-vis case needs, as opposed to underrepresentation. Second, as the data set used to conduct this study was a child welfare data set, important variables related to the use of disability services such as details about a person's functional capacity, services received or contextual information as noted in an Individual Education Plan for children or an Individual Support Plan for adults.
Future Research on Self-determination in Transition Planning
Our fourth aim explored foster care outcomes for transitional aged youth—i.e. youth aged 18 and above. It is generally expected that youth in this age group will be more likely to be discharged into an independent living situation, such as emancipation, and this was the case for approximately 80% of youth with ID. Given that the vast majority of transitional aged youth with ID were emancipated circa age 18 (81%), there is a need to understand what happens to these at-risk youth in the longer term—and how self-determination-oriented training efforts may help (Geenen, Powers, Hogansen & Pittman, 2007).
With the development of the disability rights and self-advocacy movements along with the passage of disability civil rights laws, people with disabilities have often had more of a say about their lives. While people with ID still experience many obstacles to full community inclusion, there are often many more options available to people with disabilities and services available to support them. Whether foster children with ID are routinely afforded such options during their permanency planning is unknown.
Hill, Lightfoot & Kimball (2010) document that youth transitioning out of foster care appeared in one state to be less likely to be referred to community-based disability service agencies. Whether youth with ID are receiving disability-appropriate referrals to assist in their transition remains unclear as well. Future intervention and research should build on the models for transition support that have been developed by those exploring the lives of children with all disabilities – specifically as it relates to the role of fostering self-determination. For example, Powers, Geenen, Powers, Pommier-Satya, Turner, Dalton, Drummond & Swank (2012) found that an intervention focused on the enhancement of self-determination skills had a positive impact on transition outcomes as well as quality of life.
Study findings suggest the presence of disability-related disparities. This suggests the need for inter-system collaboration between child welfare and developmental disability service providers in order to reduce adoption disruptions and dissolutions and to improve foster care outcome disparities, especially as they relate to family reunification. Foster children with ID who are facing foster care exit are reliant on both the child welfare and disability service systems—systems which differ greatly. While child welfare is usually an involuntary, time-limited intervention that is focused on surveillance with the goal of safety, permanence and well being, disability services are voluntary, and focus on the provision of supports across the lifespan. Despite attention to the presence of children with disabilities in the child welfare system within the research literature, these systems do not appear to have a strong history of collaboration, which may in part be understood as a function of their differences (Hill, 2009).
With a wide gap between the underlying assumptions and practice approaches in each system, the development of locality-specific model collaboration initiatives is called for. At the micro practice level, communication between the “lead” caseworker in both the child welfare and disability systems needs to be established in order to streamline casework. On a broader scale, non-governmental collaboratives can be established as a resource exchange mechanism. For example, in Minnesota, a Disability Child Welfare Collaborative was established in 2011 with the goal of bringing together practitioners and researchers in the areas of child welfare, disability and education. This collaborative works to raise awareness of the needs of children with disabilities in each system, acts as a central resource to all three types of providers, and fosters dialogue among providers around how to promote positive outcomes (Center for the Advanced Study of Child Welfare, 2016). Most prominent in the description of this collaborative's work is the attention paid to the need for ongoing cross-training between systems in an effort to build disability and child welfare competence—and the sense of who to turn to when unanticipated practice questions arise. Members of a collaborative such as this could be come the point-person in their own agency, functioning in a consulting capacity with other caseworkers, who, in turn, should be encouraged by supervisors to seek help that is specific to the child with whom they are working (Coyle, 2014). On the most basic level, child welfare caseworkers faced with a child with ID should understand that such children will have varying needs, even if they have the same diagnosis. A very young child who is diagnosed with failure-to-thrive may need early intervention services that can be provided in foster homes, which will result in training for foster parents vis-à-vis the child's care. However, a child with Down Syndrome might have more specialized medical needs, such as cardiac problems. Such varying needs may impact the ways in which child welfare workers work towards positive outcomes and, ultimately, exit from foster care.
In addition to a focus on fostering substantive collaboration between child welfare, education and disability providers, there is also a need to foster disability competence among foster and pre-adoptive parents—especially given the fact that most foster placements are emergency vs. planned placements (Coyle, 2014). Trainings on disability competence can include discussions on promoting the inclusion of children with ID into everyday life and helping others see such children as valued and unique individuals. Such trainings could also work to increase parent knowledge of child development and disability-specific issues, such as communication techniques and tutoring in the use of the relevant technologies, all of which can reduce frustration and foster attachment (CWLA, 2012).
In addition to adding modules on basic disability competency to foster parent training programs, child welfare providers can work to develop peer-to-peer mentoring and support groups for new foster and pre-adoptive parents who have taken in children with ID, those who work with the children the most closely can build their own competence and confidence in their care work. Especially important in these groups would be discussions and skill-building activities related to the need for foster and adoptive parents to develop leadership skills in advocating for their children in the schools and in the community. Additionally, parent advocacy trainings provided by state non-profits such as The Arc might be targeted towards foster parents and potential foster parents.
Finally, with regard to the large number of transitional aged youth who are exiting foster care to emancipation, there is a need to target the development of self-determination skills as part of independent living training (which is required for all youth who are aging out of care). For example, one study examined the impact of a self-determination enhancement project on the transition of this population, finding that the project had a positive impact on outcomes. Specifically, children receiving this intervention were more likely to finish high school, be employed, use community transition services and engage in independent living activities (Powers, Geenen, Powers, Pommier-Satya, Turner, Dalton, Drummond &Swank, 2012). Models of this nature should be drawn upon by child welfare professionals, ideally in collaboration with developmental disability service providers.
While all children in foster care are vulnerable, findings suggest reasons to consider the special vulnerabilities of children with ID vis-a-vis foster care outcomes. This study creates a foundation for continued qualitative research on this topic, with particular emphasis on the exploration of reasons for adoption disruption as well as the nature of foster care outcomes that do not involve family reunification or community-based placements. A better understanding of the life outcomes of children with ID after age 18 is also needed. Additionally, given Hill's (2009) research showing a lack of true inter-system collaboration for youth in transition, more research is needed on the ways in which child welfare and disability service systems do and do not support youth with disabilities who are transitioning out of care. This research would ideally include formal case studies that examine state or county-specific systems and approaches. There is a critical need for both further examination of these youths' outcomes and the identification of effective support strategies, particularly with regard to optimizing their transition to productive and inclusive adult lives. Finally, while there is a nascent body of research that is inclusive of the voices of children with disabilities vis-à-vis their views about foster care exit and transition, more attention must be paid to obtaining perspectives from children with ID about their experiences in the child welfare system (Blakeslee, Quest, Powers, Powers, Geenen, Neslon & McHugh, 2013; Quest, Fullerton, Geenen & Powers, 2012).