[ Main page of Report | Contents of Report ]
Considering the rapid and ongoing development of the field of post-adoption services, none of the inquiries conducted within this study can be considered definitive. Many promising programs have yet to be documented in the literature, and our case studies included only a few of the well-regarded programs currently in operation. Much work remains to be done before program and administrative data achieve their potential for describing the experience of adoptive families.
Based on this incomplete evidence, we can nevertheless make some early observations on what is known with respect to the study’s research questions:
From this base, we suggest a framework of strategies to move the field forward. These represent research, policy, and practice directions for consideration at the federal, state, and program levels.
A general caveat to these observations is that available data have rarely addressed the experiences of families with subsidized guardianship arrangements. As states expand their use of subsidized guardianship, these families will become an important part of the population of adoptive families. To the extent that their needs differ from other adoptive families, PAS programs may need to develop new approaches to reaching and serving these families.
[Go to Contents]
| What patterns of disruption, dissolution, and displacement indicate potential need for PAS? |
What we know: Disruptions, dissolutions, and displacement are extreme indicators of difficulties within adoptions, suggesting the existence of many more families in less dire need who could use PAS. There is little published data in this area, but more could potentially be mined from state administrative data systems. Concerns about confidentiality, as well as data system inadequacies, have hindered progress to date in this area.
Four studies using different methods and samples estimate the rate of disruptions for special needs children at between 10 and 16 percent of adoptions (Barth and Berry, 1988; Goerge, Howard, and Yu, 1996; Partridge, Hornby, and McDonald, 1986; Urban Systems Research and Engineering Inc. [USR&E], 1985). Several published studies agree that older children, and those with behavioral or emotional problems, are more likely to disrupt (Barth and Berry, 1991; Partridge, Hornby, and McDonald, 1986; Smith and Howard, 1994; Smith, Howard, and Monroe, 1998). In addition, better-educated parents, particularly mothers, are more likely to have troubled placements and are more likely to disrupt (Barth et al., 1988; Barth and Brooks, in press; Boyne et al., 1984).
Two prior studies (Goerge, Howard, and Yu, 1996; Festinger, 2002) have arrived at similar adoption dissolution rates of between 2 and 6 percent — figures consistent with those generated in our analysis of North Carolina data. Analysis of administrative data from North Carolina indicates that half of the dissolutions identified occurred within the first three years of adoption, and that the risk of dissolution was higher for minority children and those adopted after age five.
Patterns of out-of-home care for adopted children are doubtless affected by variations among states in how residential treatment, a common form of care, is funded. Although data are again sparse, a recent analysis of adopted families in California found about 6 percent of children adopted between 1988 and 1989 had used residential care (Allphin, 2000). Our analysis of California administrative data found that transition to residential care was typically preceded by three or more changes in adoption subsidies, suggesting that families had increased their use of services prior to out-of-home care. Increased age at time of adoption was associated with greater use of residential care, as was moderate family income.
Moving forward: Better information about patterns of disruption, dissolution, and displacement could help PAS programs reach out to, and tailor their services for, those families in which the adoption may be at risk. Of particular interest would be questions such as the following:
| Which adoptive families are most likely to want PAS? |
What we know: Data on families seeking PAS in New England, Massachusetts, Illinois, and Oregon indicate that PAS programs serve a variety of adoptive families, including those formed by guardianship, adoptions from public child welfare agencies, and private and international adoptions; and many have more than one adopted child. Median age of the focus child at the time of case opening was noted between 10 and 11 in two studies, and children are roughly evenly divided by gender. Families often perceived their needs to be acute at the time that services were sought: 21 percent of families in Massachusetts reported a severe or extreme risk to the adoption. Behavioral problems were reported by more than half of families in both Illinois and Oregon. In Illinois, 55 percent of children served have one or more diagnosed disability, and 27 percent had been placed out of the home for some time since adoption (Gibbs, Barth, and Lenerz, 2000; Hudson et al., 2002; Fine, 2000; Howard and Smith, 2001). In the absence of population-based data on adoptive families, we cannot specify the ways in which families seeking services differ from other adoptive families.
Moving forward: Most PAS programs report collecting data on the children and families they serve. If this information were compiled in a way that captured comparable data where possible, a picture might emerge. Development of a minimum program data set would facilitate both the data collection start-up and data consistency. Questions of particular relevance to PAS program planning include:
| What services are most needed by different types of adoptive families? |
What we know: Assessments of needs among adoptive families are commonly conducted by states. Approximately half of the states responding to the ILSU survey reported they were currently conducting a needs assessment or had done so within the past 10 years. Although their usefulness is often limited by low response rates and inconsistent methods, these surveys offer a general picture of services desired by adoptive families. These include educational and information resources, clinical services, material help in the form of subsidies and help with child care, and support networks. A sizeable minority of families — 30 percent in an Oregon needs assessment — report that no services are needed at this time. When families receive service, they prefer working with providers who understand the long-term impact of early trauma and the unique dynamic of adoptive families.
These surveys typically use the most complete sampling frame available, the list of families receiving adoption subsidies, but one that is only a portion of all eligible families. We know less about the service needs and experiences of those families who do not receive subsidies. Although it is possible that they do not pursue subsidies because they have few service needs, they may also be underserved by the subsidy program, with fewer resources with which to address their needs. Most needs assessments also exclude families who have adopted privately or from other countries.
Moving forward: Our understanding of the needs of adoptive families could be substantially improved by two measures. First, development of a standardized needs assessment instrument for states’ voluntary use could reduce design costs, increase the comparability of data over time and across states, and — in some cases — improve data quality. Second, a population-based survey of adoptive families, with adequate follow-up to ensure high response rates, would provide needed data on topics such as
| What part do adoption subsidies play in supporting post-adoption services? |
What we know: Adoption subsidies are a sizable and growing federal and state expenditure. We know only the broad outlines of subsidy distribution. Nationally, 88 percent of children adopted in 2000 received subsidies (DHHS, 2001c). The number of children receiving subsidies appears to be rising in tandem with increases in the number of adoptions; preliminary AFCARS data showed a 3 percent increase between 1998 and 2001 (Penelope Maza, personal communication, August 26, 2002). The proportion of adoptive families receiving subsidies may be lower for adoptions in the program’s earlier years. A 1993 analysis found that substantial numbers of families who could have qualified for subsidies did not receive them (Sedlak and Broadhurst, 1993). This is significant in light of data indicating that families receiving subsidies are less likely to experience disruption (Berry and Barth, 1990).
State and county subsidy policies vary widely. A review by the North American Council on Adoptable Children found that states varied in both the generosity and flexibility of their subsidy policies. The review documented differences in the definition of special needs, basic and specialized subsidy rates, availability of deferred subsidy agreements and subsidized guardianships, and payment for residential and respite care (Bower and Law, 2002). Although there is not yet data with which to rigorously evaluate the relationship between adoption subsidy provision and the success of adoptions, the authors may be correct in asserting that the underuse and misuse of subsidies may discourage would-be adoptive parents and contribute to the incidence of disruptions. Even when state policies govern subsidy provision, variation in implementation may produce vastly different results. Between-county variability in subsidy provision may be nearly as great as state-to-state variability (Avery, 1998), and have similar consequences.
Very little has been done to examine changes in subsidies over time. These changes may be essential purchase post-adoption services and supports that help them respond to changing needs. Our analysis of administrative data from California and North Carolina found large differences between the two states in the likelihood that families will experience a change in subsidy amount at some point. Although most subsidy increases appear to be associated with periodic reviews or with administrative cost-of-living adjustments, an important proportion of changes are unrelated to review schedule. These changes appear to be associated, in part, with changes in children’s behavioral problems. A pattern of increasing frequency of subsidy changes often preceded initiation of residential care, suggesting that these changes could serve as a marker for families in need of supportive services. The likelihood and amount of subsidy changes were also found to be associated with parental characteristics (income and maternal education) as well as child characteristics. Higher income and more educated households received a disproportionate share of larger increases. This suggests that families’ ability to negotiate needed increases may play as substantial a part in subsidy determinations as financial need.
Moving forward: Better understanding is needed of both the determinants of state policy and their role within adoptive families. The substantial variations among states with respect to subsidy policies suggest that impacts on the adoption of children from foster care and the level of services available to them must certainly exist. More needs to be understood about which factors influence state policies on subsidies and other supports, and how federal reimbursement policies shape state decision-making. Given better administrative data, it would also be possible to more fully examine the relationship between subsidy levels and changes and time to permanency, adoption disruption, and displacement.
A variety of changes in adoption subsidy policies have been proposed in recent years, such as allowing subsidies to be higher than foster care payments or allowing subsidies to be used to pay for temporary residential treatment. These would be challenging to assess because they are governed by state statutory requirements. Waivers from standard Title IV-E funding schemes could be granted to test other cost-neutral approaches to delivering subsidies, but this would be very difficult to achieve, and the outcomes would probably be quite subtle.
At the family level, we can work to understand how subsidy increases are being used, whether they are being adjusted in ways that appear to be appropriate to family circumstances, the implications of current approaches to subsidy determination and re-authorization, and the interface between subsidies and PAS programs. Among the questions to be answered:
| How satisfied are adoptive parents with the services they have received? |
What we know: Adoptive parents’ satisfaction with services available to them — either from PAS programs or other community resources — is a subjective indicator of the need for new or additional PAS. Many programs assess the extent to which participating families are satisfied with services provided by the PAS program (see, for example, Fine, 2000 and Hudson et al., 2002). Client satisfaction studies appear to generally indicate strong satisfaction with PAS program offerings. However, there are several concerns about the quality of their data. Studies vary widely in design sophistication, ranging from simple ratings to detailed assessments. Unless substantial effort is invested in follow-up, response rates may be so low as to compromise data validity. Response bias is also likely, particularly because few PAS programs contract with external evaluators, and much PAS evaluation data is collected by agency staff who may be associated with PAS delivery.
Several barriers may limit the usefulness of community services, including limited availability, cost, restricted access and poor match with family needs (Fine, 2000). Adoptive parents participating in focus groups as part of Casey Family Services’ evaluation described community providers as often lacking the necessary understanding and skills to address their families’ issues (Gibbs, Barth, and Lenerz, 2000). Taken together, these findings suggest that families may be unable to access the specific services they need in their communities, and that services available to them may fall short of being adoption-competent.
Moving forward: Satisfaction studies ideally should address the universe of services used by adoptive families in the community and through PAS programs. In this way, programs would not only know how satisfied families were with the PAS program services they received, but also whether they were providing the services needed. Such studies would also identify areas in which the PAS program could work toward advocacy and system change by increasing awareness of adoption issues within schools, or training mental health providers in adoption issues. Research in a limited number of communities with varying PAS availability, conducted by independent researchers, using both qualitative and quantitative methods, could illuminate such questions as the following:
[Go to Contents]
| What program models are being used to provide PAS? |
What we know: In recent years, states have established new statewide PAS programs and expanded existing smaller scale projects. Most of these programs go beyond provision of a clearly defined intervention to families with similar needs, instead offering an array of interventions with different objectives and activities to serve a broad range of adoptive families. PAS programs in the five case-study states are similar in three fundamental ways. First, the PAS programs are contracted out rather than provided by state child welfare staff, to control costs and increase administrative flexibility. Second, these programs offer services statewide, using various organizational structures. Finally, the programs share goals of keeping adoptive families together and creating a consumer-driven program.
These states did not report drawing on proven models or documented best practices to develop their PAS programs. Instead, they relied primarily on related experience and advocates’ input. Only one state reported using a publication about post-adoption services (Spaulding for Children, 1996) to help design their PAS program.
Better understanding of mix of subsidies, agency contracts, or private providers to be used for post-adoption services is certainly feasible. The relationship between this payment strategy and program outcomes, however, will be far more difficult to test, until we have methods to gauge the effectiveness of the purchased services.
Moving forward: The experience of the five states participating in the case studies is not sufficient to provide guidance to others considering developing or expanding programs. The ILSU interviews provide limited information about service availability, but lack detail and do not differentiate time-limited services provided by child welfare agencies from PAS programs. As more states consider establishing PAS programs, the availability of well-documented and evaluated program models will be useful in garnering support for funding and implementation. Specific questions include the following:
| Which adoptive families are eligible to be served by PAS programs? |
What we know: Most programs identified in the literature review provided PAS to families that have completed adoptions from the public child welfare system. The ILSU interviews suggest that these families remain the priority population for state-sponsored PAS programs. Children adopted internationally or privately may be excluded from receiving PAS. Among the five case-study states, however, two served all adoptive families regardless of adoption type. We do not know how likely it is that families with children adopted privately or from overseas will require services. However, data suggest that the needs they identify when they do seek services are similar to those of families who adopted from public child welfare systems (Gibbs, Barth, and Lenerz, 2000). One of the case-study states reported that its PAS program was open to pre-finalization adoptive families as well, an important target population given the challenges that families often face in the early phases of adoption (Pinderhughes, 1996).
Moving forward: State adoption agencies naturally place priority on services to children adopted from their foster care population. However, if PAS programs are effective in preserving adoptive families, removing eligibility restrictions could reduce displacement and dissolution among children adopted privately or internationally, and reduce disruptions among children from public child welfare agencies. Even modest achievements of either goal might make broader eligibility cost effective. Evaluation data from programs serving these families could answer questions including:
| Which funding streams are being used to provide PAS? |
What we know: PAS programs have been funded by a variety of sources, including foundations, Adoption Opportunities grants, state general revenue, Adoption Incentive Program bonuses, and, more recently, Title IV-B Subpart 2 funds (Promoting Safe and Stable Families program). A recent review shows that 15 percent of Title IV-B Subpart 2 funds were used for adoption support and preservation, with state allocations for those purposes ranging from 0 to 25 percent of the funds received. However, it is not possible to identify the extent to which those funds were spent on post-adoption services rather than facilitating adoptions (James Bell Associates, 2001). The five case-study states used a mix of state and Title IV-B Subpart 2 funds for their PAS programs.
Moving forward: Little is known about states’ decision-making process when it comes to funding PAS. Specifically, we do not know how the availability of federal funding support influences states’ consideration of developing PAS programs, how states choose among available funding streams. Given states’ recent achievements in increasing adoptions, expansion and sustainability may become critical issues in PAS funding.
Specific questions related to PAS funding include:
| What services are being offered by PAS programs? |
What we know: Although the field has proposed an optimal continuum of care for adoptive families (Howard and Smith, 1997), the provision of post-adoption services can best be described as patchy rather than comprehensive. Services offered range from information and referral networks to support for residential treatment. Yet there is little uniformity in provision of services across, and sometimes within, states.
Although each of the states responding to the ILSU interviews reported providing a range of PAS, open-ended descriptions of these services revealed that many states relied on adoption subsidy workers to provide limited information and referral services and/or had one-time grants or time-limited programs designed to provide various PAS. Some states considered adoption subsidies to be a PAS, since families can use these funds to purchase needed services.
Our case studies and available written descriptions of state-level programs suggest that information and referral, parent and professional training, advocacy, and support groups are the most commonly offered post-adoption services. Adoption-sensitive counselors were frequently offered as well. Although respite care was commonly mentioned by families as a significant service need, funding for, or provision of, respite care was relatively uncommon.
Many sites do not offer one or more of these basic elements, and available services are often not linked to each other. In a few places (e.g., Minnesota, Oregon, Massachusetts), we did find statewide information and referral resource interfaces that identify adoption support groups that operated in local communities. In some places, training of clinicians about adoption was paired with information mechanisms to inform adoptive families about which professionals were trained, but this pairing was more the exception than the rule.
Moving forward: At a basic level, we can identify effective strategies for putting families in touch with the resources that exist in local communities. Web-based resources may be of value, including interactive systems that enable parents to search for adoption-competent therapists and websites with information on services provided, eligibility, support group locations, and links to other resources.
PAS program developers have to date drawn on a patchwork of professional recommendations, state-level needs assessment and practice wisdom to design their service menu. Absent from this mix is systematic compilation of the experience of states that have developed programs, at a level of detail that could offer practical guidance for other states, particularly those at earlier stages of program development. Useful information would answer the following questions:
| What treatment models are most effective in meeting the needs of adoptive families? |
What we know: Among the many interventions offered to adoptive families, there is no accumulated evidence regarding which approaches are useful, at what dose or desirable duration, and for which types of families. Disappointing results from early efforts to apply short-term, intensive family preservation models (Barth, 1991) suggest that service systems designed for biological families may not be effective with adoptive families. Only a few evidence-based interventions, such as parent-mediated training approaches (e.g., Fischer and Stormshak, 2000) and clinical models (Hoagwood et al., 2001; Weersing and Weisz, 2002), have been adapted for adoptive families.
Developing evidence-based approaches to treatment of adoptive families is fraught with unknowns (Schoenwald and Hoagwood, 2001). The greatest unknown is how adoptive families differ from other families that have benefited from evidence-based clinical approaches. Although the adoption field has long concluded that adoptive families are quite different from those formed by birth (Kirk, 1981), the importance of these differences to effective service delivery needs to be better understood. Although there is also much diversity among adoptive families, some have argued (e.g., Groza and Rosenberg, 2001) that adoptive families — whether created via international, independent, or public agency adoptions — are more clinically alike than different.
Moving forward: We need to develop a disciplined and strategic approach to advancing the quality and effectiveness of post-adoption counseling services. Promising approaches to development and dissemination of treatment innovations that begin with intensive training and supervision of homogeneous adoption populations and moves toward more naturalistic conditions of intervention with more heterogeneous adoption populations (Schoenwald and Hoagwood, 2001). Burns and Hoagwood (2002) have provided an important resource with their description of child and family services at various stages of testing. Yet we do not know how effective these efficacious interventions would be in the PAS context.
Questions that could be addressed by an expert panel and through an intensive review of existing research include:
| How are PAS programs working to change the larger service delivery system? |
What we know: In addition to services provided by PAS programs, families may be involved with contract agencies, private clinicians contracting with the public agency, mental health clinics, or private-sector providers. Many PAS programs are attempting to improve the adoption competence of community providers who interact with or treat adoptive families.
Training may be designed to increase the sensitivity of professionals such as school or medical personnel who come into contact with adopted children and families. More intensive efforts involve training mental health workers to increase their competence in working with adoptive families. The University of Washington, in collaboration with the state, offers a postgraduate certificate for mental health professionals working with adoptive families, and Casey Family Services provides extensive training to professionals in New England states. Other states offer training in specific therapeutic approaches, such as Georgia’s training for mental health professionals in attachment therapy. However, public agencies may hesitate to train private providers whose fees are not reimbursable by Medicaid. Some may also question whether turnover in county mental health agencies limits the impact of staff training.
States may also influence the service delivery system by their choice of which services are delivered through the PAS program. Oregon has elected not to provide counseling through its PAS program, in part because county mental health offices are intended to be the point of entry for mental health services.
Moving forward: To inform system change efforts, information is needed on how families experience services they receive from community resources, in general, and from clinicians with experience and training in adoption issues. Understanding the stability of efforts to sensitize communities and mental health systems will be necessary in assessing the impact of training efforts. Questions in this area include the following:
[Go to Contents]
Evaluations of PAS programs are fairly rare in the published literature, and many well-regarded programs are strikingly limited in their evaluation activities. Among the reasons for this deficit are the relative newness of the field, the ongoing evolution of treatment models and service classifications, program models that tailor services to family needs rather than following predefined protocols, and a lack of evaluation experience among program staff. Three basic strategies could substantially advance evaluation development.
| How can the field begin to build a foundation of evidence-based practice for PAS? |
Moving forward: PAS programs have not yet developed an evidence base for treatment of adoptive families. They are proceeding down the path of other services — e.g., family preservation and SAMHSA’s child and adolescent services system program (CASSP) — which became well-established before they were evaluated and are now struggling to document their effectiveness (Schoenwald and Hoagwood, 2001). The field needs a systematic approach to rigorously analyzing PAS.
A systematic approach can support development of clinical treatment models for use with adoptive families. There are existing paradigms that may guide this process well (e.g., Hoagwood, Burns, and Weisz, 2002; Rothman and Thomas, 1994). These approaches could be adapted to the current context of the PAS field to plan a long-term approach for the development of evidence-based services. A 10-year plan for the development of post-adoption service research could address selection of interventions to be tested, identification of the subpopulations of adoptive families for initial trials, dissemination of promising approaches, and subsequent program refinement.
| What strategies can improve evaluation quality and reduce its burden for PAS programs? |
Moving forward: We know that the diversity and complexity of the PAS program offerings make it extremely challenging to evaluate “post-adoption services.” While the field works to determine which programs are most effective for those families who need substantial amounts of clinical services, evaluation efforts are needed to document and assess other commonly delivered services, and their delivery as part of multifaceted PAS programs. Although it may not be feasible to precisely quantify the specific impact of less intensive services such as information and referral and support groups, there is much to be learned about issues such as which families use these services, in what quantities and patterns, and how they work together to address the needs that families identify.
Because the staff time required is a major barrier to evaluation, providing a ready resource of PAS evaluation tools and strategies would substantially facilitate the process. A minimum data set along with electronic data collection and scoring tools would help make the evaluation enterprise less onerous. Technical assistance in evaluation could build capacity among service providers with little or no experience in evaluation.
| What incentives could be used to encourage evaluation of PAS programs? |
Moving forward: Incentives can facilitate evaluation by addressing both the resource and goodwill costs of evaluation activities. We do not know what kind of incentives would open the doors to using evaluation approaches with adoptive families. Some PAS providers have concerns about asking families to participate in evaluations, feeling that data collection activities are at odds with their effort to move from a clinical atmosphere to one of support. The introduction of random assignment to services, on top of rigorous data collection, is likely to be met with still greater levels of concern.
It is not clear what mix of incentives and technical assistance might be the most powerful accelerant for evaluation efforts. Because funding for post-adoption services is so diverse, incentivizing evaluations will be difficult. Funds to pay for additional services — and not just for evaluation activities — would be needed in order to gain the commitment of the necessary number of PAS program sites.
For providers to commit to evaluation, evaluation must be sensitive to their concerns about family sensitivities. Families may in fact be quite receptive to participating in evaluations that can help secure funding for services.
Funding agencies can promote evaluation as a tool for program improvement, building on the amply demonstrated commitment of PAS coordinators and providers to delivering the best possible support to adoptive families. This will involve structuring evaluation methods that can actually provide timely and useful input on questions of interest to program staff, and providing technical assistance to enable staff to interpret and apply evaluation findings.
Promoting Safe and Stable Families funds are often used for post-adoption services and could be used for evaluation. However, some states use these funds for recruitment of families and purchase of services to expedite adoptions; they use adoption bonuses or state funds to purchase post-adoption services. The Adoption Opportunity program could serve as a vehicle for soliciting competitive proposals for rigorous evaluations. State agencies funding PAS can also increase their support for evaluation. Such approaches may be needed but not sufficient to increase the level of evaluation activity. The field also needs to examine ways to add new resources for both evaluation and services. Just paying the ticket for evaluation — without a companion ticket for additional services — is unlikely to entice states and counties to attend the evaluation dance.
[Go to Contents]
PAS has been something of a handcrafted field, in which dedicated advocates and managers developed creative approaches to families’ needs. Sustaining these programs over the long term, however, would require a solid knowledge base that complements the dedication and creativity that have fueled it to date. The questions and tasks outlined in the previous sections, while ambitious and broad-ranging, could be addressed with a fairly modest set of interrelated research and evaluation activities:
Main Page of Report | Contents of Report
Home Pages:
Human Services Policy (HSP)
Assistant Secretary for Planning and Evaluation
(ASPE)
U.S. Department of Health and Human Services
(HHS)