Adoption Breakdown: Concept,Research,and Implications

Child-Related Factors

Regarding child characteristics associated with adoption breakdown, an older age at placement stands out in all reviews (Child Welfare Information Gateway, 2012; Coakley & Berrick, 2008; Evan B. Donaldson Adoption Institute, 2004; Faulkner et al., 2017; Festinger, 2014; Rushton, 2004; Smith, 2014). The child’s age at placement was a factor identified in the first studies (Kadushin & Seidl, 1971) as well as in more recent research (e.g., Selwyn et al., 2014). It has been highlighted in both intercountry (e.g., Palacios, Sánchez-Sandoval, & León, 2005) and domestic (e.g., Selwyn et al., 2014) adoptions, as well as both in special needs (e.g., Rosenthal et al., 1988) and nonspecial needs studies (e.g., Goerge, Howard, Yu, & Radomsky, 1997; see also Paniagua, Palacios, Jiménez-Morago, & Rivera, this issue).

Growing up in very adverse circumstances, an older age implies a longer exposure to adversity. Persistent maltreatment and toxic stress alter stress reactivity, brain functions, development, and behavior (Turecki, Ota, Balangero, Jackowski, & Kaufman, 2014) and the children involved tend to develop a view of the world and people as dangerous and unpredictable (Gibb, 2002), with negative consequences for mood and behavior (Heim & Nemeroff, 2001), as well as for self-regulation and interpersonal relationship capacities (Tarren-Sweeney & Vetere, 2013). Age at placement is a proxy for accumulated adversity whose harmful impact goes well beyond the moment of placement.

In the adoption breakdown literature, the child’s age is usually mentioned only in terms of placement, with age at breakdown being less considered. Recent research has consistently found that early adolescence is when more youngsters leave their adoptive homes either temporarily or permanently. Maza (2014), Palacios et al. (2015), Rolock and White (2016), and Selwyn et al. (2014) report an average age of 13–14 years when the adoptees leave their families, normally several years after placement. The average time since placement until breakdown was 6 years in Palacios et al. (2015), while in Selwyn et al. (2014), 57% left their home 5 years after the adoption order, and only 14% within 2 years. As Palacios and colleagues (2015) indicated, more than the consequence of a temporary crisis, adoption breakdown tends to happen after several years of stress and difficulties (see also Paniagua et al., this issue).

The other child-related factor systematically associated with breakdown in all the reviews is the child’s behavioral and emotional problems (Child Welfare Information Gateway, 2012; Coakley & Berrick, 2008; Evan B. Donaldson Adoption Institute, 2004; Faulkner et al., 2017; Festinger, 2014; Maza, 2014; Rosenthal, 1993; Rosnati et al., 2017; Rushton, 2004; White, 2016). As an example, Selwyn et al. (2014) provide a good summary of the problems that the adoptive parents in their study reported as being difficult to handle, the number and severity of such problems not being similar for all children. In this study, the most common situation was that the difficulties started soon after placement (80% of cases), while for the rest, the escalation of problems began at around puberty, similar to the proportion reported by Palacios et al. (2015). Difficulties with early onset included troubles forming close attachments, manipulation and control, anger and aggression, mood and self-esteem problems, inappropriate sexual behaviors, and thinking and learning problems. The difficulties starting or escalating during adolescence included anger, aggression, violence within the adoptive home, oppositional behavior and running away, criminal offences, and sexualized behaviors (Selwyn et al., 2014). Of particular relevance was the presence of violence in the family, very often (but not exclusively) from children to parents. Similar to the percentage in Palacios et al. (2015), in almost two thirds of their cases, Selwyn et al. (2014) found a connection between child violence and breakdown. Testa, Snyder, Wu, Rolock, and Liao (2014) have shown that as the adoptive parents’ rating of the difficulties increases so do their thoughts about ending the relationships, which, mediated by the level of parental commitment, is a predictor of breakdown.

Regarding emotional problems, attachment disturbances at placement were often described by parents of children who experienced breakdown in the study by Selwyn et al. (2014). In 60% of the breakdown cases in Palacios et al. (2015), attachment problems were identified, mostly from the very beginning of placement. Since attachment problems are relational, this aspect will be returned to when discussing parent-related factors.

Both earlier (Barth, Berry, Yoshikami, Goodfield, & Carson, 1988) and more recent studies (Rolock & White, 2016; Selwyn et al., 2014) are consistent in showing that children with more moves in care (including previous experiences of adoption breakdown) prior to their new adoptive placement have a higher risk of new breakdown experiences (Rolock et al., this issue). Lack of trust toward adults and fear of new emotional commitments are most likely involved, as well as difficult behavioral issues and lack of training and support for the parents.

Findings for other child aspects have been more mixed. Some studies have identified specific types of preadoption maltreatment associated with breakdown. But the type of maltreatment varies from one study to another: neglect and emotional abuse in Smith et al. (2006); sexual abuse in Nalavany, Ryan, Howard, and Smith (2008) and in Smith and Howard (1994); and exposure to domestic violence in Selwyn et al. (2014).

Contrasting findings have also been reported for the placement of siblings. While in Smith et al. (2006), the placement of siblings was associated with a higher risk of breakdown (with the exception of four or more placed together), in Rolock and White (2016), it was associated with a lower probability. In her review on sibling placement outcomes, Hegar (2005) concluded that most studies suggest that these placements are as stable as, or more stable than, placements of single children or separated siblings. Reflecting on these contradictory findings, Festinger (2014) suggests that other factors besides sibling placement per se might be involved, such as the composition of the household where the siblings are placed. According to the Evan B. Donaldson Adoption Institute (2004) review, some of the aspects to consider are the presence of biological children in the adoptive home, the number of children present, the age and age order of all the children involved (see also Selwyn, this issue).

Poorer placement outcomes have been reported for “preferential rejection” cases where a child is adopted while other siblings remain with their birth parents. This may occur because the preferentially rejected child develops a negative image of self and adults that may work against the formation of a secure attachment to the adopters (Rushton & Dance, 2003). A similar phenomenon of preferential rejection has been identified in breakdown cases involving siblings when only one of them leaves the adoptive home while the other(s) remain, the oldest child in the sibling group being involved in 70% of the cases and the youngest one only in 18% of the cases in Palacios et al. (2015).

Research has also explored the role of other child factors, such as gender or race. Most studies have found these characteristics not to be associated with breakdown once all the other factors (particularly age at placement and behavior/emotional problems) are considered. When these factors have been identified to play a role, males were overrepresented compared to females (Barth et al., 1988), as well as African American compared to White children (Rolock et al., this issue; Rolock & White, 2016; Smith, 2014; Smith et al., 2006).

Other child-related factors potentially relevant for the breakdown experience have not been empirically explored, or are only in an initial stage of study, due to the difficulty in accessing the relevant information. Genetic and constitutional characteristics, genes–environment interactions, exposure to prebirth or very early adversities not documented or very poorly reported in the child records can carry long-term negative consequences including serious difficulties in interpersonal relationships (e.g., Litrownik, Proctor, & Yeh, 2018).

Parent-Related Factors

While a few child-related factors have been unanimously associated with adoption breakdown, this is not the case for parent-related factors. This in part has to do with the fact that different researchers consider different characteristics. Palacios et al. (2015), for instance, analyzed the motivation to adopt, concluding that when the main motivation was the satisfaction of adults’ needs (e.g., desire to have children, need to give love) the risk of breakdown was higher. Testa et al. (2014) analyzed parents’ commitment once serious problems were perceived as exceeding their capacity, finding this to be a mediator between problems and breakdown. But the variables in these two examples (motivation, commitment) are not included in most adoption breakdown studies. Also, even when a given parental characteristic is found to be relevant in several studies, its association with breakdown tends to be at best modest, particularly when considered in isolation.

Although in these circumstances it is more difficult to provide a synthesis of the main research findings, the characteristics of intact adoptive families described by McRoy (1999) for special needs adoptions may be a good guide for our summary. Following her conclusions, parents who are able to avoid the escalation of problems leading to breakdown have more stability in their marriages, realistic expectations, more flexibility, more experience in parenting special needs children, greater commitment to the child, and more willingness to seek help when needed.

The stability of the couple relationships mentioned by McRoy (1999) has also been identified in breakdown studies (Westhues & Cohen, 1990). Longer marriages, and hence adopters’ older age, have been mentioned as relevant for placement stability in some publications (Berry & Barth, 1990; Groze, 1986; Kadushin & Seidl, 1971). But studies using a matched design of intact versus disrupted placements (Barbosa-Ducharne & Marinho, this issue; Marinho et al., 2012; Rosenthal et al., 1988; Smith & Howard, 1991) have not identified this variable to be relevant. As with many other issues, the length of a marriage or the age of the adopters per se is probably unrelated to the breakdown experience.

Adopters’ unrealistic expectations have also been highlighted in some studies (Palacios et al., 2015; Paniagua et al., this issue; Randall, 2013; Reilly & Platz, 2003; Rosenthal, Groze, & Curiel, 1990; Smith, 2014). These expectations typically involve idealized views of the child or of their own abilities as parents (Schmidt, Rosenthal, & Bombek, 1988) and might be related with two aspects identified in some other studies. First, Barth and Miller (2000) reported more unrealistic expectations in parents with a higher education. Although some studies (Rosenthal et al., 1988) found more stable placements among parents with lower education, Rosenthal’s (1993) review suggests that for each study showing the relevance of parents’ education for adoption breakdown, there is another one where such association is not found. Second, idealistic expectations might be facilitated by lack of accurate information provided to parents about the child at the time of placement (Schmidt et al., 1988; Selwyn et al., 2014), a point to be discussed later.

Parenting abilities to address problems and be flexible in their adaptation to the child’s characteristics, particularly in the absence of appropriate and targeted support services, have been identified in some adoption breakdown studies. Partridge, Hornby, and McDonald (1986) reported that some parental characteristics associated with adoption stability were parents’ ability to emotionally distance themselves from the child’s problems, their capacity for advocating for services for the child, for altering their expectations in line with the child’s abilities, and their flexibility. Similarly, Marinho, Barbosa-Ducharne, and McRoy (2012) found that flexibility toward the child needs and self-confidence in their parenting role were two of the adoptive parents’ characteristics that distinguished the intact from the disrupted group. Dance and Rushton (2005) also reported that warmth, involvement with the child, and parental sensitivity were significantly associated with placement outcome and stability. Parental inability to cope with children’s problematic behavior was also a factor in the postadoption placements studied by Maza (2014).

In the analysis by Dance and Rushton (2005), one of the key issues in adoptive parents’ commitment to the child was their perception of problems in the attachment relationships of the child to them. Frequently, adoptive parents who experienced breakdown referred to their children’s inability to attach to them as one of the main factors leading to serious difficulties (Schmidt et al., 1988). Relatedly, in the study by Rushton, Mayes, Dance, and Quinton (2003), when adoptive mothers felt that they were getting something back from the child they were more likely to develop attachment and to remain committed despite the difficulties.

The combination of several of the previous characteristics might help to understand one of the relatively more consistent findings regarding parents’ characteristics associated with adoption breakdown: A relationship with the child previous to the adoption plan has been associated with placement stability in several studies. When parents and child had been living together for some time before considering adoption (typically, in a foster care placement converted in adoption), the expectations may be better adjusted to the child’s characteristics, and the parents are committed to continue parenting the child. This might help to explain why the adoption by known parents has been found to be more stable than the adoption by “new” parents in several studies (Festinger, 1986; McRoy, 1999; Rosenthal et al., 1988), but not in all (Reilly & Platz, 2003; Selwyn et al., 2014). The finding that adoption by relatives has more stability (Festinger, 1986; Smith et al., 2006) goes probably along similar lines.

To end with the roster of parents’ characteristics summarized by McRoy (1999), their willingness to seek help when in trouble needs a brief mention. Parents can receive support from the family, their social network, or the available professional services. Since this second aspect will be addressed in the following section, we only refer here to the first one. In keeping with what is known in the nonadoption literature, several studies have identified a social support network (family, friends, resources in the community, and other adoptive parents) to be associated with placement stability (Berry, 1997; Leung & Erich, 2002; Marinho et al., 2012; McRoy, 1999; Randall, 2013; Rosenthal et al., 1990). Enhancing community networking is an integral part of the goals of effective support services (Barth & Miller, 2000).

In their study of adoption breakdown in Spain, Palacios et al. (2015) reported that while 45% of the adoptive parents experiencing serious difficulties in their adoption tried to find solutions to their problems and sought help, 55% went to child protection services to unexpectedly communicate that the placement of the child should end. This could reflect their emotional distance to the child, their isolation from support networks, or their perception that professional services were not available or were ineffective. Most likely, this is a multifaceted problem in which the parents’ despair with the long-lasting difficulties could be coupled with the lack of support to be discussed below.

In summary, research findings concerning adoptive parents’ characteristics associated with breakdown are inconsistent. This does not mean that those characteristics are irrelevant and has probably more to do with the limitations of studies that consider some but not all the relevant factors and that analyze characteristics in isolation more than their interaction. It is important to consider not only interaction with each other of the parents’ characteristics, but also with the child’s characteristics discussed previously and with those of support and services to be discussed now.

Support and Service Factors

All reviews on the adoption breakdown literature refer to the importance of what is referred to as “systemic factors” (Evan B. Donaldson Adoption Institute, 2004), “service characteristics” (Festinger, 2014), and “agency factors” (Faulkner et al., 2017). The specific aspects considered in each review vary as much as these labels. Once again, the unanimity regarding some of the children’s characteristics relevant for adoption breakdown disappears when it comes to the other aspects. Nevertheless, some similarities between different studies exist and they can be grouped in pre- and postplacement professional activities around adoption.

Suitability assessment; preparation of both parents and children; and matching, sharing, and explaining information about the to-be-placed children are the main preplacement professional activities involved. Although research about these different services is quite scarce, the conclusions of a few studies are worth considering. Even if it is mandatory in all jurisdictions, there is very little empirical evidence about the effectiveness of adoption suitability assessment and the research evidence in adoption breakdown cases seems nonexistent. The little we know about the content of the suitability reports suggests that they deserve more research attention. For example, in their study of assessment reports in Sweden (not related to adoption breakdown), Lind and Lindgren (2017) report that risk factors were never mentioned in the conclusions, which probably reflects the fact that the risk of breakdown is not contemplated in the assessments. The three areas suggested by Selwyn (2015) for prospective adopters’ assessment are fully coherent with the previous section in this review: parental hopes and expectations, parental sensitivity, and management of stress and support networks.

Something similar happens with parents’ preparation for adoption, which is now mandatory in most jurisdictions. All prospective adopters receive it, but very little is known about its quality and content or its connection with breakdown. In their study of intercountry adoption disruptions in Spain, Palacios et al. (2005) identified lack of preparation as a risk factor for parents who had not received training at a time when this was not yet compulsory. Its importance is highlighted by the fact that, according to Wind, Brooks, and Barth (2007), parents comprehensively prepared were more likely to use postadoption services if needed than those less well prepared. Similar to what happens with their parents-to-be, the preparation of children for adoption has not received consideration in most studies of the breakdown experience or has been found unrelated with the adoption outcome (Festinger, 2014).

Some information exists about the importance of parent–child matching, an issue whose interest is emphasized in the Evan B. Donaldson Adoption Institute (2004) and Quinton (2012) reviews. As indicated there, the context of the problem is often the delicate exercise of “stretching” parents’ initial preferences to accommodate the characteristics of children that otherwise would have no available family. Although stretching makes the placement of the child feasible in the short-term, it may also create the context for serious long-term difficulties. In fact, in McRoy’s (1999) study, a large proportion of the breakdown cases (87% in prelegalized adoptions and 76% in postlegalized ones) were considered poor matches. Similarly, Berry (1997) found that disruptions were more likely when, for instance, parents did not intend to adopt a child with emotional disturbances and later realized that the child placed was emotionally very difficult. In Marinho et al.’s (2012) study, lack of consideration of the birth order in the matching of children into adoptive families which already contained children was also found to be a problem.

Matching errors could be linked to two opposite attitudes. On the one hand, “picky” people who listed a large number of preferences for the child they are ready to adopt (Festinger, 2014): “such pickiness at the start did not augur well in the long run” (p. 445). On the other hand, applicants who were ready to adopt children with characteristics (such as a late placement age) that in the long run they could not deal with successfully (Palacios et al., 2015). This optimism can be coupled with adoption professionals’ misjudgment of the adopters’ capacity to meet high levels of children’s problems (Randall, 2013).

Finally, failure to share and explain information about the child has been identified as a risk factor in some studies that have emphasized the importance of honest and accurate preplacement information about the child (Barbosa-Ducharne & Marinho, this issue; Berry, 1997; Randall, 2013; Rosenthal, 1993). Often, parents learn about relevant information (e.g., child’s sexual abuse) only when the child is with them (Barth & Berry, 1988). Although it could happen that the caseworkers were not aware of some of the child’s problems, some parents feel that they were deliberately misinformed or deceived (Berry, 1997). In these circumstances, there is a frequent mismatch between the adoptive parents’ and adoption workers’ perceptions, as the latter more often thought that they had given all the information to the parents and explained its meaning (Barth & Berry, 1988; Selwyn et al., 2014). Communication problems between what is told and what is heard seem to be involved in some cases, perhaps with adoption professionals not transmitting a complete and accurate picture of the child, and with prospective adopters downplaying the problems they are told about (Palacios, 2012). Communication issues may also be associated with the timing of the receipt of information. Information received about a child at the time of placement may not be relevant to the parent until issues start to develop later in life, at a time when initial information may be forgotten or not fully understood.

Aspects of the postplacement professional intervention related to breakdown include the early identification of problems and the support provided to ameliorate them. The role played by some organizational characteristics of the services provided, often in connection with the two previous problems, has also been analyzed. All these issues will be considered now.

In the breakdown cases where problems started soon after placement, early detection is of critical importance. If these problems do not receive adequate attention, it facilitates the “mounting of unresolved conflicts” identified in the early studies of adoption breakdown (Donley, 1978, p. 36). Limited identification of problems may have different reasons. Adoption professionals could be busy with other activities, or parents could not be fully aware of the problems or prefer to end contact with the caseworkers (Selwyn et al., 2014). Adoptive parents could believe that help seeking reflects parenting failure, worry about being blamed for the child’s problems, or be frustrated by their perception of a lack of recognition of the importance of their difficulties (Rushton, 2004). Also, professionals can misinterpret problems as “typical of the adaptation stage,” expecting a “normalization” with time that does not happen (Palacios et al., 2015).

Research has identified several problems in the support services and activities to address the needs of troubled families. Postadoption services are in place “rarely and irregularly,” to use Barth and Miller’s (2000, p. 449) expression. Paradoxically, for those adopting out of the foster care system, the level of support after adoption is lower than what was available while in foster care (Festinger, 2002). Also, parents have less contact with adoption support over time (Houston & Kramer, 2008). Although many families report that what is available is adequate to meet the needs of the children they had adopted or assumed guardianship of (Fuller et al., 2006), others report significant difficulties in accessing needed services, with a clear mismatch between the problems faced and the amount and quality of the support provided, often limited to well-intended advice and counseling (Palacios et al., 2015).

The organization and provision of support services receive frequent mention in adoption breakdown research. The Evan B. Donaldson Adoption Institute (2004) review identifies the fragmentation of responsibilities and division of services as one of the main issues (see also Barbosa-Ducharne & Marinho, this issue). Staff discontinuities (Festinger, 1986) and problems with the availability and quality of mental health services for families in crisis (Randall, 2013; Selwyn et al., 2014) are part of the same picture. Also, Palacios et al. (2015) observed that, when in serious trouble, adoptive families sought private clinical assistance from professionals with no expertise in adoption issues, even when free of charge specialized public services were available. This was interpreted as lack of information about the available services, but also in terms of the parents’ attempts to escape from the blaming and control by adoption services. As shown by Smith et al. (2006), the use of professionals with more experience in the field is related to a lower breakdown risk, which highlights the importance of adoption-competent clinical services (Brodzinsky, 2013) as stressed in Brodzinsky and Smith (this issue).

The duration of the offered professional support deserves a final mention. Frequently, postadoption services are provided only in the months following placement. But, even if the problems start soon after placement, many adoptive parents do not seek support until several years afterward (Faulkner et al., 2017), which creates just one more mismatch between what is needed and what is offered. In this regard, it is worth mentioning that Berry, Propp, and Martens (2007) found that the importance of the available postadoption support for achieving permanence increased with time after placement.

Some Legal and Policy Basic Requirements

The legal regulation of adoption and the organization of adoption-related services vary significantly from one jurisdiction to another. The policy implications of what is known about adoption breakdown need to be analyzed with this diversity in mind. Considering this, our analysis here will be limited to two general issues: placement age and organization of adoption-related services.

Our review of factors associated with adoption disruption highlighted the critical importance of an elevated age at placement as a key risk factor for adoption breakdown. As indicated above, age at placement is a proxy for accumulated adversity, which finds expression in the child’s emotional and behavioral problems identified as the other more relevant risk factor for adoption stability and permanence. The implication is that all that can be done to reduce the exposure to adversities and to speed up adoption placements will be beneficial to prevent adoption breakdown.

Legal changes in different countries have stressed the importance of reducing waiting times and promoting the stability that adoption can afford. Actions to consider include improving early identification of child abuse and neglect, imposing serious restrictions to the institutionalization of young children where this is still an option, reducing the number of transitions from one care context to another, preserving family connections through suitable kin caregivers, using concurrent planning (placement with a family who will become the adoptive family if reunification is unsuccessful and the court makes the necessary order), and facilitating adoption from foster care when this is in the best interest of the child and without a reduction of child and family support. Furthermore, for children adopted at an older age, the provision of supports and services after placement is particularly important to sustain parent–child relationships and prevent adoption breakdown.

The organization of adoption-related services must acknowledge the fact that the time when assessing adoption suitability and matching children and parents were the main (if not the only) professional activities is long past. Thorough assessment of adoption suitability that identifies both strengths and weaknesses of parents and children, comprehensive adoption preparation of parents and children, identifying criteria for good enough parenting and effective matching, and provision of quality postplacement follow-up and support services are some of the areas that cannot be entirely left at the discretion of each agency or professional. Developing standards of good practice in all these areas and ensuring its implementation should be part of a policy oriented toward the promotion of placement stability.

Beyond the family level, the organization of community educational and mental health services must also be considered. These are critical components of the ecology of adoption located outside the family, but with a relevant influence on what happens inside (Palacios, 2009). At school, many adopted children face learning difficulties, but some also encounter problems in managing relationships, facing stigma associated with adoption, and adapting to the requirements of the school context (Bomber, 2010; Soares, Barbosa-Ducharne, Palacios, & Fonseca, 2017). As for the need that many adoptive families have regarding counseling and clinical services, a thorough analysis of the barriers encountered and the way to overcome them has been presented by Brodzinsky (2013). As indicated in his analysis, services lacking adoption awareness and specialization can at times do more harm than good. The availability and specialization of educational and mental health services ready to take account of adopted persons and adoptive families in need of support are a key component to promote stability and prevent breakdown. This may need to begin with the training of adoption-related professionals for them to understand the unique strengths and needs of families formed through adoption (see Brodzinsky & Smith, this issue).

Selwyn (2017) has reviewed the support needs of adoptive families and the evidence for effective interventions, rating them at three different levels: interventions effective and proven by research, interventions supported by research, and promising practices. But for the more problematic cases, even the best clinical treatment may not be enough. Severely troubled youngsters need to increase the number and quality of positive interactions outside the clinical context, and according to Perry (2006), the approach must include the process of creating a “therapeutic web” in which any healthy and invested people in the child’s life (parents, siblings, teachers, kin, and neighbors) are needed in providing therapeutic opportunities in their daily life interactions. The multisystemic therapeutic approach discussed by Brodzinsky and Smith (this issue) is a good example of this type of intervention.

Adoption Practice

Even when a wide array of services are available, the final day-to-day work occurs in the families. In the best of circumstances, these families are supported by informal networks and professional staff. For staff working with adoptive children and families, this activity requires professionals who are able to manage the complexity of adoptive families. This may include a shift from work where the main goal was helping families to find children, to a long-term view that includes finding, assessing, preparing, and supporting families to ensure the stability and permanence for children with early adversities.

The complexity of the work is highlighted in a Child Welfare Information Gateway (2006) bulletin outlining “the basics of adoption practice.” The list of the “basics” includes family and child assessment, birth parents’ involvement, openness in adoption, matching families and children, sharing information with families, involving youth in their placements, placing children with families, postplacement services to facilitate adjustment and deal with attachment issues, and support services once the adoption has been legalized. Other relevant basics to consider are preparation of both parents and children, advocacy and liaison roles with the health and education systems, and supporting the adoptees’ search for their origins. The list of standards for child welfare professionals put forward by the National Association of Social Workers (2013) is no less intimidating in variety and complexity.

The diversity and complexity of the tasks and the skills required to fulfill them makes team work and specialization inevitable. Selwyn et al. (2014) outlined specific recommendations for team and service managers as well as for adoption professionals. Ensuring adequate training, supervision and support of adoption workers, dealing with caseload distribution and individual case management, putting forward effective ways to respond to the varied needs of children and families, and promoting team reflective practice are some of the activities to be added to their long list of recommendations for adoption teams or agencies. For adoption professionals, some of the recommendations additional to the ones outlined by Selwyn et al. (2014) include capacity to work under stress and uncertainty, skills for team work, abilities to relate to parents and children, comfort in multidisciplinary work, and continuous professional improvement.

Research Needs

When Kadushin and Seidl (1971) studied “adoption failures” several decades ago, their review of the literature included eight studies and was summarized in a table. Fifteen years later, in her insightful analysis of the “necessary risks” that adoption professionals need to take when placing children in families, Festinger’s (1986) review of the literature about adoption breakdown occupied little more than three pages. If we now consider the extension and list of references of more recent reviews, including this one, it is evident that we have accumulated a rich body of empirical knowledge about adoption breakdown. However, significant gaps and unanswered questions still remain. Two will be mentioned here regarding, respectively, basic and applied research.

In its current state, studies of adoption breakdown incidence are a jumble of samples, designs, methods, and figures. New studies will probably add more diversity of approaches and findings. It is in this context that a meta-analytic exercise was suggested above as a necessary step forward.

Although a substantial body of knowledge about factors associated with breakdown has accumulated in recent years, the findings lack cohesion due to the diversity of methods and factors in each study. Approaches considering variables jointly rather than in isolation are a significant step forward and should to be pursued in future research. Matched designs of intact and disrupted adoptions have been used only scarcely (Barbosa-Ducharne & Marinho, this issue; Marinho et al., 2012; Rosenthal et al., 1988; Smith & Howard, 1991; Wijedasa & Selwyn, 2017), but are likely to be a fruitful direction for research. Also, as Festinger (2014) has indicated, some children experience an adoption that ends and are later successfully adopted in other families, and this could provide for another interesting avenue of new research designs.

Finally, while we have limited knowledge about the postbreakdown adopters’ perspective (Schmidt et al., 1988; Selwyn et al., 2014), even less is known about postbreakdown children’s perspectives. Selwyn et al. (2014) interviewed 12 youngsters who had experienced an adoption breakdown and this needs to be seen as just an initial effort in a promising direction. In residential facilities or in new families after breakdown, adopted youngsters are a very rich source of information and their voice needs to be heard and understood more often.

The needs in applied research regarding adoption breakdown and the strengths that allow some families to thrive are perhaps more fundamental. As Rushton (2004) indicated, the main problem is our lack of knowledge about what works best for whom and about how to replicate successful interventions to achieve similar outcomes. Given the complexity of the problems involved and the intricate dynamics of unhappy adoptive families, research is needed on the use of multidimensional treatments like those developed for troubled youngsters at risk of out-of-home placement (Brodzinsky & Smith, this issue; Hengglerer & Schaeffer, 2016) or placed in foster families or residential homes (Fisher & Gilliam, 2012; Sinclair et al., 2016).

Among the many needs that these children and families experience, two stand out for their singular importance: establishing and maintaining attachment bonds and effective ways to handle problem behaviors such as violence within the family. Zeanah, Berlin, and Boris (2011) have analyzed some of the main attachment issues in working with children and parents, including assessment and effective interventions, with a particular emphasis on children in care and adopted after institutional experiences. A better knowledge of how these tools can be used to strengthen parent–child emotional bonds would be essential to facilitate permanence and stability in problematic placements, as well as to increase adoptive parents’ commitment with their new child.

Finally, with regard to research about effective ways to handle problem behavior and violence, much can be learned from the multidimensional approaches mentioned above and developed to avoid out-of-home placements or to increase stability in at-risk foster care placements. At least two research questions deserve consideration. First, the early identification of difficulties that may hamper placement stability. Screening tools like the one researched by Hurlburt, Chamberlain, DeGarmo, Zhang, and Price (2010) for foster care placements at risk of disruption need to be developed and studied also in the context of risky adoptive placements. Second, effective ways for working with children and parents in contexts where the relationship is jeopardized by problems and violence need to be studied in problematic adoptions. The guidelines for pediatricians working with children with maltreatment experiences, including the analysis of evidence-based treatments and trauma-informed practices for children in foster care (Sege & Amaya-Jackson, 2017), is a good example of the way to follow and of the need of sound research about the viability and efficacy of different treatment approaches and methods.