Updated Systematic Review of Psychological Interventions With Adoptive Parents to Improve the Well-being of Their Adopted Children

Inclusion Criteria and Literature Search Process

The screening process for the articles included in this replication of Ni Chobhthaigh & Duffy’s (2019) systematic review was conducted with reference to the inclusion criteria detailed in Table 4 of Annex 1, which are based on the study’s delineated focus.

The systematic search for documents, concluded on December 2025, employed the following search string across all databases: (“adopted infant*” OR “adopted child*” OR “adopted adolescen*” OR “adoptee*” OR “adoptive parent*” OR “adoptive carer*” OR “adoptive famil*”) AND (“intervention” OR “therapy” OR “parenting program*” OR “parent* education” OR “parent* support” OR “parent* training”) in title, keywords, and abstract. We used this Population and Intervention (PI) search strategy, which is a modification of the standard PICO strategy (Richardson et al., 1995), to select any kind of intervention study in that population regardless of the type of treatment, comparison or outcome measured.

The following databases were consulted: ASSIA, CINAHL, Social Services Abstracts, and ProQuest (including ERIC, MEDLINE, PsycInfo, and ProQuest Dissertations & Theses Database). Additionally, reference lists of included documents were reviewed, along with two internationally recognised journals in the field: Adoption & Fostering and Adoption Quarterly. In replicating this systematic review, access to the original Embase database was restricted, so the PubMed database was used instead, providing comparable coverage of relevant literature. Similarly, the original grey literature databases, Grey Literature Report and OpenGrey, are now obsolete; instead, the databases Dissertations & Theses Database, Open Dissertations, and Psicodoc were used. Unlike the original review, direct expert consultation was not feasible due to time and resource constraints. To compensate for this deviation, we conducted a manual search of reference lists from all included studies and of two specialised journals in the field (Adoption & Fostering and Adoption Quarterly). Although this approach ensured a comprehensive screening of relevant sources, the absence of expert consultation remains a limitation of the updated review.

Information Organisation

Screening of titles, abstracts, and full texts was conducted by the first author, following the procedure used in the original review. Data extraction was also carried out by the first author. Any uncertainties during screening or extraction were discussed with the second or third author and resolved by consensus (see Table 1). Accordingly, the studies were classified by study population; age of adopted children; the design, type, duration, and theoretical foundation of the interventions; follow-up implementation; and the main instruments used to measure the various psychological variables in the studies. Similarly, information was organised to indicate how the different interventions influenced effect sizes (measured by Cohen’s d) in the three areas of interest: psychological well-being, behavioural functioning, and parent-child relationship.

OPEN IN VIEWER

Table 1.

Classification of Documents Included in the Update

Document and country	Population under study	Age of adopted children	Intervention design	Follow-up of the intervention	Type, duration, and theoretical foundation of the intervention 1	Outcome measurement instruments 2
Barone et al. (2018) Italy	Adoptive mothers with adopted children (N = 80)	1 year 3 months to 6 years (M = 3.6 years)	Randomised controlled trial (pre-post, control group)	No follow-up	VIPP-FC/A Duration not reported attachment theory; positive parenting; emotional availability	EAs (emotional availability scales)
Midgley et al. (2018) United Kingdom	Adoptive families (N = 36)	2 to 17 years (M = 9 years)	Non-randomised trial (pre-post, no control group; mixed methods)	No follow-up	PAFSI 6 months (6 sessions) attachment theory; systemic family theory	BAC (brief assessment checklist)
Opiola & Bratton (2018) United States	Adoptive families (N = 49)	2 to 10 years (M = 5.4 years)	Randomised controlled trial (pre-post, control group)	No follow-up	CPRT 10 weeks (10 sessions) attachment theory; filial therapy; play therapy	CBCL (total problems); MEACI (empathy); PSI-4 (parenting stress)
Barone et al. (2019) Italy	Adoptive families (N = 83)	1 to 6.2 years (M = 3.6 years)	Randomised controlled trial (pre-post, control group)	Follow-up: 6 months	VIPP-FC/A 4 to 6 months attachment theory; positive parenting; emotional availability of caregivers	EAs; CBCL; IF (indiscriminate friendliness)
Swan et al. (2019) United States	Adoptive fathers and mothers (N = 11)	8 to 14 years (M = 10.3 years)	Pilot study (three time points: Baseline, mid, post; no control group)	No follow-up	CPRT 10 weeks (10 × 20 min sessions) attachment theory; filial therapy; play therapy	CBCL (total problems); MEACI (empathy); PSI (parenting stress)
Lind et al. (2020) United States	Adoptive families (N = 124)	1 to 3 years (M = 1.6 years)	Randomised controlled trial (pre-post, control group)	Follow-up: 12, 24, 36, 48, and 60 months	ABC 10 sessions attachment theory; sensitive parenting; filial therapy; play therapy	BITSEA, DB-DOS (examiner present; examiner busy; parent context)
Polek & McCann (2020) United Kingdom	Adoptive couples (N = 51)	5 to 7 years (M not provided)	Non-randomised trial (T1, T2 at 10 weeks, T3 at 20 weeks; no control group)	No follow-up	TLPDI 20 weeks (20 × 50 min sessions) couple skills theory; marital relationships in parenting	SDQ (prosocial scale)
Zeegers et al. (2020) Netherlands	Adoptive families (N = 53)	3 to 11 years (M = 8.12 years)	Non-randomised trial (pre-, post and 6 months follow-up; no control group)	Follow-up: 6 months	Basic trust 8 sessions attachment theory	AISI, CBCL (internalising and externalising problems)
Yarger et al. (2021) United States	Adoptive families (N = 119)	1.5 to 3 years (M = 1.8 years)	Randomised controlled trial (pre-post, control group)	Follow-up: child age 40–60 months	ABC 10 sessions attachment theory; sensitive parenting; filial therapy; play therapy	BITSEA, DB-DOS (examiner present; examiner busy; parent context)
Hunsley (2021) United States, Australia, Canada, China, Macedonia, New Zealand, South Africa, taiwan, Thailand	Adoptive families (N = 95)	12 to 18 years (M = 13.8 years)	Randomised controlled trial (pre-post; waitlist control group)	Follow-up: 3 weeks	PAFSI 7 months Attachment theory; systemic family theory	CPRS-SF, IPPA-R, FACES IV, SDQ
Purrington et al. (2022) United Kingdom	Adoptive families (N = 47)	3 to 16 years (M = 8.7 years)	Non-randomised trial (pre-post, no control group)	No follow-up	Theraplay 10 weeks (10 x 2-h sessions) attachment theory; play therapy; and neurocollaborative therapy	TSCYC, CBCL, BRIEF-2, ACC

Once the final documents were selected, they were analysed based on study location, sample size, participant characteristics, research design, and measurement instruments used. The effect size, Cohen’s d, was subsequently calculated for the outcomes in emotional well-being, behavioural functioning, and parent-child relationships. For studies that did not report effect sizes, Cohen’s d was calculated using the following formula: Cohen’s d = (M1 – M2)/SD; where M1 and M2 are the means of groups 1 and 2, and SD the pooled standard deviation.

Risk of Bias Assessment

This review employed the recommended approach for assessing the risk of bias in studies, as outlined in Cochrane reviews. This tool evaluates five specific domains: sequence generation, allocation concealment, participant and assessor blinding, incomplete outcome data, selective outcome reporting, and other potential biases. These domains are rated as “high risk of bias,” “uncertain risk of bias,” and “low risk of bias” (Higgins & Green, 2011).

To assess the reliability of the risk-of-bias assessment, an external reviewer conducted a blinded evaluation of a random subsample of eight studies (73% of the included studies). The agreement between the author and the external reviewer was measured qualitatively, and any discrepancies were resolved through discussion until consensus was reached. A kappa statistic was not calculated, which represents a minor limitation in the reliability assessment.

Study Characteristics

Figure 1, prepared as a PRISMA flowchart, details the process followed to identify articles and the resulting outcomes. As shown, this procedure concluded with a total of eleven identified documents.

Of the eleven documents analysed, four studies were conducted with adoptive families from the United States (Lind et al., 2020; Opiola & Bratton, 2018; Swan et al., 2019; Yarger et al., 2021), three in the United Kingdom (Midgley et al., 2018; Polek & McCann, 2020; Purrington et al., 2022), two in Italy (Barone et al., 2018, 2019), one in the Netherlands (Zeegers et al., 2020), and one involving adoptive families from various countries (doctoral thesis by Hunsley, 2021).

Regarding participant characteristics, nine studies included adoptive families as the intervention group, whereas the remaining two focused exclusively on adoptive parents. Sample sizes ranged from 11 (Swan et al., 2019) to 124 (Lind et al., 2020). A clear concentration of interventions was observed in early childhood.

Most studies (n = 9 of 11) included infants or young children under the age of six, either exclusively or as part of broader age ranges. Four interventions were directed exclusively at children under six years old (Barone et al., 2018, 2019; Lind et al., 2020; Yarger et al., 2021), while five additional studies included young children alongside school-aged children (Midgley et al., 2018; Opiola & Bratton, 2018; Polek & McCann, 2020; Purrington et al., 2022; Zeegers et al., 2020). Only one study focused exclusively on school-aged children (Swan et al., 2019), and a single study specifically addressed adopted adolescents (Hunsley, 2021). These findings highlight the prevailing emphasis in the literature on early interventions, with limited representation of programmes targeting adolescents.

Across studies, recruitment was typically conducted through adoption agencies, clinical services, or voluntary participation, reflecting convenience sampling procedures.

In terms of intervention design, five studies employed randomised controlled trials with two conditions (experimental and control), using pre-post intervention assessments (Barone et al., 2018, 2019; Lind et al., 2020; Opiola & Bratton, 2018; Yarger et al., 2021). One additional study followed a similar structure but implemented a waitlist control group (Hunsley, 2021). Four studies adopted quasi-experimental, within-subject designs without a control group, including two with pre-post assessments (Midgley et al., 2018; Purrington et al., 2022) and two with multiple time points, incorporating follow-up evaluations (Polek & McCann, 2020; Zeegers et al., 2020). One study was structured as a pilot with a single group and three measurement points (Swan et al., 2019). This variation reflects a notable degree of heterogeneity across the included interventions.

Table 2 provides detailed content descriptions of the interventions analysed. All studies explicitly based their theoretical foundation on attachment theory. Two studies evaluated post-adoption family support interventions (PAFSI) grounded in systemic family theory (Hunsley, 2021; Midgley et al., 2018). Two studies implemented Attachment and Biobehavioural Catch-up (ABC), rooted in sensitive caregiving, filial therapy, and play therapy (Lind et al., 2020; Yarger et al., 2021). The VIPP-FC/A intervention, focusing on parental sensitivity and gentle discipline, was applied in two studies (Barone et al., 2018, 2019).

OPEN IN VIEWER

Table 2.

Effect Sizes of the Results

Study	Comparison	Psychological wellbeing measure	Cohen’s d	Behavioural functioning measure	Cohen’s d	Parent-child relationship measure	Cohen’s d
Barone et al. (2018) VIPP-FC/A	Pre-intervention vs. post-intervention Control vs. intervention	​	​	​	​	EAs	- d = 0.46
Midgley et al. (2018) PAFSI	Pre-intervention vs. post-intervention	​	​	BAC	- d = 0.71	​	​
Opiola and Bratton (2018) CPRT	Pre-intervention vs. post-intervention Control vs. intervention	​	CBCL - Total problems	​	- d = 0.79	- Total stress MEACI - Total empathy	- d = 0.77 - d = 2.30
Barone et al. (2019) VIPP-FC/A	Pre-intervention vs. post-intervention Control vs. intervention	EAs	- d = −0,45	CBCL IF	- d = −0,25 - d = −0,19	​	​
Swan et al. (2019) CPRT	Pre-intervention, mid-intervention, post-intervention	CBCL - Total problems	- d = 0.33	​	​	MEACI - Total empathy PSI - Total stress	- d = 0.97 - d = 0.62
Lind et al. (2020) ABC	Pre-intervention vs. post-intervention Control vs. intervention	BITSEA: - Socio-emotional competence DB-DOS - Examiner present context - Examiner busy context	- d = 0.55 - d = 0.51 - d = 0.42	​	​	​	​
Polek and McCann (2020) TLPDI	T1, T2, T3	​	​	​	​	SDQ - Prosocial scale	- d = 0.05
Zeegers et al. (2020) basic trust	Pre-intervention, post-intervention and follow-up	CBCL - Internalising problems	- d = −0.38	CBCL - externalising problems	- d = −0.53	AISI - Avoidant - Ambivalent - Disorganized	- d = −0.35 - d = -0.33 - d = -0.33
Yarger et al. (2021) ABC	Pre-intervention vs. post-intervention Control vs. intervention	BITSEA - Socio-emotional competence DB-DOS - Examiner present context - Examiner busy context	- d = 0.49 - d = 0.5 - d = 0,23 - d = 0.20	​	​	​	​
Hunsley (2021) PAFSI	Pre-intervention vs. post-intervention Waitlist control vs. intervention	​	​	SDQ - Total difficulties - Conduct problems	- d = −0.10 - d = −0.07	CPRS-SF - Closeness - Conflict FACES-IV - Cohesion - Communication IPPA-R - Father attachment - Mother attachment	- d = 0.21 - d = −0.29 - d = 0.41 - d = 0.22 - d = −0.09 - d = 0.08
Purrington et al. (2022) theraplay	Pre-intervention vs. post-intervention	TSCYC - Anxiety - Anger - Excitement due to post-traumatic stress BRIEF-2 - Inhibition - Emotional regulation	- d = 0.25 - d = 0.40 - d = 0.33 - d = 0.35 - d = 0.29	BRIEF-2 - Behavioural regulation	- d = 0,45	CBCL - Social problems	- d = 0.22

In two studies, the Child Parent Relationship Therapy (CPRT) was used, based on filial therapy and play therapy (Opiola & Bratton, 2018; Swan et al., 2019). Another study focused on the adoptive couple with a time-limited psychodynamic intervention (TLPDI) (Polek & McCann, 2020), which centred on developing couple and marital skills for parenting. Of the two remaining studies, one used Basic Trust Intervention, based on attachment theory (Zeegers et al., 2020); the final intervention, Theraplay, was play-based and emphasised neuro-collaborative therapy (Purrington et al., 2022). Intervention durations ranged from 8 weeks (Zeegers et al., 2020) to 7 months (Hunsley, 2021).

Consistent with the original review, most studies relied on general measures of child functioning (e.g., CBCL, BITSEA, SDQ). Only a small number used adoption-specific tools such as the AISI (Zeegers et al., 2020), the BAC (Midgley et al., 2018), or the ACC (Purrington et al., 2022).

A breakdown of the outcomes assessed shows that psychological wellbeing was examined in 8 of the 11 studies (72.7%), a proportion identical to that observed for behavioural functioning, which was also assessed in 8 studies (72.7%). Aspects of the parent–child relationship were likewise evaluated in 8 studies (72.7%), although only one study assessed all three domains simultaneously (9.1%). Most studies examined two domains (63.6%), while three focused exclusively on a single domain (27.3%).

Across these domains, a wide range of instruments was employed. Psychological wellbeing was primarily assessed through measures of emotional availability, socio-emotional competence and child symptomatology, including the EAs, CBCL, BITSEA, DB-DOS, TSCYC and BRIEF-2. Behavioural functioning was captured through tools targeting behavioural adjustment and regulatory capacities, such as the BAC, CBCL, IF, SDQ and BRIEF-2. Parent–child relationship outcomes were evaluated using measures of emotional availability, parental stress, empathy, attachment and family functioning, including the EAs, PSI-4, PSI, MEACI, SDQ, AISI, CPRS-SF, FACES-IV, IPPA-R and, in one study, the CBCL Social Problems scale.

The following sections summarise the effects of the interventions on psychological wellbeing, behavioural functioning and parent–child relationships. Results are presented by thematic patterns.

Across studies, psychological wellbeing outcomes tended to improve most consistently in interventions grounded in parental sensitivity and attachment-based feedback (e.g., VIPP-FC/A, ABC), which showed medium to large effects on socio-emotional competence and emotional availability. Play-based interventions (CPRT, Theraplay) also demonstrated improvements in children’s emotional expression and regulation, as well as reductions in parental stress. Systemic or family-focused programmes (PAFSI, TLPDI) produced more modest but still positive changes, particularly in family cohesion and relational functioning.

Overall, psychological wellbeing outcomes showed a pattern of medium-to-large improvements, although effect sizes varied substantially across studies depending on intervention type, child age, and measurement tools.

Across studies, behavioural functioning showed improvements primarily in interventions that incorporated structured parent–child interaction components or targeted regulatory capacities (e.g., ABC, CPRT, Theraplay). These programmes tended to reduce externalising behaviours, improve behavioural regulation and decrease indiscriminate friendliness, with effect sizes generally in the small-to-medium range. However, results were more variable than in the psychological wellbeing domain, with some studies reporting limited or domain-specific changes, particularly when baseline behavioural difficulties were mild or sample sizes were small.

Interventions focusing on parental sensitivity, reflective functioning and relational attunement (e.g., VIPP-FC/A, CPRT, PAFSI) consistently reported improvements in parent–child relationship quality. These included increases in emotional availability, parental empathy and perceived attachment security, as well as reductions in parenting stress. Effects were generally medium to large, although the magnitude varied depending on the specific relational construct assessed and the measurement tools used. Overall, relational outcomes showed some of the most robust and consistent improvements across the included interventions.

Across domains, improvements in children were primarily observed in socio-emotional competence, behavioural regulation and emotional availability. In contrast, effects on parents were most evident in reductions in parenting stress, increases in parental sensitivity and empathy, and improvements in perceived relationship quality. These two domains did not always change in parallel, suggesting that child and parent outcomes may respond differently depending on the intervention components.

Risk of Bias in Interventions

Table 3 shows the risk of bias assessment for the eleven included studies. In general, the studies showed a medium-to-high bias risk, particularly in selection bias, reporting bias, and other sources of bias.

OPEN IN VIEWER

Table 3.

Riesgo de Sesgo de Las Intervenciones Analizadas

Study	Generation of random sequences (selection bias)	Allocation concealment (selection bias)	Blinding of participants and personnel (performance bias)	Incomplete outcome data (missing data bias)	Selective reporting of outcomes (reporting bias)	Other potential threats to validity (other biases)
Barone et al. (2018)
Midgley et al. (2018)
Opiola y bratton (2018)
Barone et al. (2019)
Swan et al. (2019)
Lind et al. (2020)
Polek y McCann (2020)
Zeegers et al. (2020)
Hunsley (2021)
Yarger et al. (2021)
Purrington et al. (2022)

Most of the included studies did not provide sufficient information to allow for an adequate assessment of risk of bias related to random sequence generation. Regarding random sequence generation, only three of the eleven studies (27%) used appropriate randomisation procedures and were rated as low risk (Barone et al., 2019; Hunsley, 2021; Lind et al., 2020). The remaining studies were judged as a high risk either due to insufficient information (Barone et al., 2018; Opiola & Bratton, 2018; Yarger et al., 2021) or because no control group was included, making randomisation unnecessary (Midgley et al., 2018; Polek & McCann, 2020; Purrington et al., 2022; Swan et al., 2019; Zeegers et al., 2020).

Allocation concealment was explicitly reported in only two studies, which were therefore rated as low risk (Lind et al., 2020; Opiola & Bratton, 2018). Four studies did not provide sufficient detail and were rated as unclear (Barone et al., 2018, 2019; Hunsley, 2021; Yarger et al., 2021), while the remaining studies were classified as high risk due to the absence of randomisation (Midgley et al., 2018; Polek & McCann, 2020; Purrington et al., 2022; Swan et al., 2019; Zeegers et al., 2020).

Performance bias was low in four studies where blinding procedures were clearly described (Barone et al., 2018, 2019; Lind et al., 2020; Opiola & Bratton, 2018). One study lacked sufficient information and was rated as unclear (Yarger et al., 2021). The remaining studies were judged as high risk due to participant awareness of intervention condition (Hunsley, 2021) or the absence of a control group (Midgley et al., 2018; Polek & McCann, 2020; Purrington et al., 2022; Swan et al., 2019; Zeegers et al., 2020).

Outcome data bias was generally low, as most studies reported no missing data or appropriately addressed missingness (Barone et al., 2018, 2019; Midgley et al., 2018; Opiola & Bratton, 2018; Polek & McCann, 2020; Purrington et al., 2022; Yarger et al., 2021; Zeegers et al., 2020). Three studies did not provide sufficient information and were rated as unclear (Hunsley, 2021; Lind et al., 2020; Swan et al., 2019).

Selective reporting bias was mostly rated as unclear due to limited information on prespecified outcomes (Barone et al., 2018, 2019; Hunsley, 2021; Lind et al., 2020; Midgley et al., 2018; Opiola & Bratton, 2018; Purrington et al., 2022; Swan et al., 2019; Yarger et al., 2021; Zeegers et al., 2020). Only one study was rated as low risk Polek & McCann (2020).

Other potential additional sources of bias were judged as low in four studies (Barone et al., 2018, 2019; Lind et al., 2020; Opiola & Bratton, 2018), while the remaining studies were rated as unclear due to insufficient information (Hunsley, 2021; Midgley et al., 2018; Polek & McCann, 2020; Purrington et al., 2022; Swan et al., 2019; Yarger et al., 2021; Zeegers et al., 2020).

Effects of the Interventions

The results are presented by outcome domain to facilitate comparison across interventions and to highlight common patterns in their effectiveness.

Discussion of Results

This updated systematic review synthesised eleven studies evaluating seven psychological interventions delivered to adoptive parents. Although several interventions reported improvements in children’s psychological well-being, behavioural functioning, or parent–child relationships, the evidence base remains highly heterogeneous. Effect sizes ranged from negligible to very large, and benefits were inconsistent across outcomes, intervention types, and developmental stages. Combined with small samples and high or unclear risk of bias, this variability limits the strength of the conclusions that can be drawn.

These inconsistencies continue the trend already identified by Ni Chobhthaigh and Duffy (2019), and mirror broader findings in the adoption and early adversity literature, where intervention effects often vary depending on baseline risk, developmental timing, and the specific mechanisms targeted (Finet et al., 2020; Gunnar & Reid, 2019; Hornfeck et al., 2019).

The interventions varied substantially in duration, intensity, and theoretical orientation. While all were grounded in attachment theory, many incorporated additional frameworks such as filial therapy, play therapy, systemic approaches, or positive parenting. This reflects broader developments in parent–child intervention models, which increasingly emphasise relational attunement, co-regulation, and interactional processes. Such trends are consistent with contemporary evidence from attachment-based and interaction-focused programmes in foster care and high-risk dyads (Dozier & Bernard, 2019). However, such heterogeneity complicates direct comparison and limits the identification of active components.

Methodological limitations identified in the original review persist. Few studies controlled for placement timing or used adoption-specific outcome measures, despite evidence that adoption-related histories shape symptom presentation. Although over half of the studies employed randomised designs, risks of selection, performance, and reporting bias remain due to limited transparency regarding allocation, blinding, and fidelity procedures.

These methodological weaknesses are consistent with recent systematic reviews highlighting similar challenges in post-adoption and foster-care intervention research (Harris-Waller et al., 2018). To improve clarity, these methodological issues are further elaborated in the subsection ‘Strengths and Limitations’.

Beyond extending the time frame, this review examined whether the quality and scope of the evidence have evolved since 2017. Some improvements are evident: several recent studies used stronger designs, including RCTs with follow-up assessments, and some interventions incorporated elements consistent with trauma-informed practice (e.g., emphasis on safety, predictability, and co-regulation). However, none of the included studies explicitly described their programmes as trauma-informed or evaluated trauma-specific mechanisms. Despite these developments, no new categories of interventions have emerged, and the overall pattern of findings remains similar to the previous review, suggesting slow progress in innovation despite policy and practice developments.

The proportion of studies assessing parent–child relationship outcomes was comparable to the earlier review, indicating continuity rather than expansion in relational measurement. This stability aligns with broader evidence emphasising parental sensitivity and emotional availability as central mechanisms supporting adopted children’s adjustment. Recent work continues to highlight these mechanisms as key drivers of change across diverse caregiving contexts (Meléndez et al., 2024).

Notable gaps remain. Nine of the eleven studies targeted children under six years old, with only one study focusing on school-aged children and one on adolescents. This imbalance limits the generalisability of findings to older adopted populations, whose developmental needs and identity-related challenges may differ substantially. The heterogeneity of designs and outcome measures prevented meaningful comparison of effect sizes by developmental stage. This gap is particularly relevant given emerging evidence that trauma-informed and relational interventions may require adaptation for older children and adolescents (Lo et al., 2024).

The included studies did not consistently report key implementation characteristics such as delivery modality, provider qualifications, or adoption-related contextual factors (e.g., type of adoption, age at placement). This limitation in the systematic reporting of the interventions has been noted in other reviews in the area (Schwochow et al., 2023). This lack of standardised reporting prevented a comparative analysis of intervention components or delivery features. Recruitment was typically based on convenience sampling, introducing potential selection bias. Greater transparency in implementation reporting has been identified as a priority in recent methodological guidance for family-based interventions (Li & Beets, 2023; Pfledderer et al., 2024).

Strengths and Limitations

This review has several strengths. First, it updates and extends previous work by incorporating studies published since 2017, allowing for a more contemporary assessment of the evidence base. Second, the review applies a systematic and transparent methodology, including duplicate screening and structured data extraction, which enhances reliability. Third, the inclusion of diverse intervention models—ranging from attachment-based programmes to systemic and play-based approaches—provides a broad overview of current practices in post-adoption support. Fourth, it offers a detailed and reproducible risk of bias assessment that informs conclusions.

However, several limitations must also be acknowledged. At the study level, methodological weaknesses remain common. Many studies relied on small samples, convenience recruitment, or lacked adoption-specific outcome measures, limiting generalisability. Reporting of key implementation characteristics (e.g., provider qualifications, delivery modality, adoption context) was inconsistent, hindering meaningful comparison across intervention components. Although more than half of the studies used randomised designs, risks of selection, performance, and reporting bias persisted due to limited transparency in allocation procedures, blinding, and fidelity monitoring.

At the review level, limitations include the exclusion of qualitative studies and single-case designs, which may have omitted valuable contextual insights. Language restrictions may have led to the omission of relevant non-English evidence, and the absence of expert consultation may have reduced the identification of grey literature. Additionally, although some studies reported variables such as child age or time in placement, reporting was inconsistent and did not allow for systematic analysis of their influence on outcomes.

Together, these strengths and limitations provide important context for interpreting the findings of this review and highlight priorities for future research, including improved reporting standards, greater methodological rigour, and the development of interventions tailored to the needs of older adopted children and adolescents.

Conclusions

In conclusion, this updated systematic review provides an integrated and contemporary synthesis of psychological interventions delivered to adoptive parents, highlighting both areas of progress and persistent gaps in the evidence base. Although several interventions demonstrated benefits for children’s socio-emotional functioning and parent–child relationships, the overall pattern of findings remains characterised by substantial heterogeneity and methodological limitations, which constrain the strength of the conclusions that can be drawn.

Compared with the previous review, this update offers a clearer understanding of how recent studies align with broader developments in the fields of attachment, early adversity, and parent–child intervention research. The inclusion of newer studies with stronger designs, as well as the identification of emerging elements consistent with trauma-informed practice, provides a more nuanced picture of current approaches to post-adoption support. However, the absence of explicitly trauma-informed frameworks, the limited attention to older adopted children and adolescents, and the inconsistent reporting of implementation characteristics indicate that the field has progressed slowly in terms of conceptual and methodological innovation.

Taken together, the findings underscore the need for more rigorous, theory-driven, and developmentally sensitive research, particularly studies that examine mechanisms of change, incorporate adoption-specific measures, and evaluate interventions tailored to diverse age groups and family contexts. Strengthening reporting standards and expanding the evidence base through mixed-methods and longitudinal designs will be essential for advancing knowledge and informing practice. By clarifying these priorities, this review contributes to shaping future directions for research and intervention in adoptive family support.