Effectiveness of Social Skills Interventions in young people aged 12 to 25 years with Autism Spectrum Disorder: Systematic Review and Meta-analysis

doi:10.21203/rs.3.rs-5113751/v1

Download PDF

Systematic Review

Effectiveness of Social Skills Interventions in young people aged 12 to 25 years with Autism Spectrum Disorder: Systematic Review and Meta-analysis

https://doi.org/10.21203/rs.3.rs-5113751/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

The aim of this review to examine the available evidence on the efficacy of social skills interventions (SSI) on core autistic traits among youth diagnosed with autism aged 12-25 years. The review included 15 studies (N=1244) where the meta-analysis findings revealed that participants receiving SSIs resulted in improvements in social skills, based on observation (SMD=0.49, 95% CI: 0.20, 0.78), adolescent (social knowledge (SMD=2.03, 95% CI: 1.15; 2.91) and social performance (SMD=0.45, 95% CI: 0.10; 0.81), and parent (SMD=0.42, 95% CI: 0.13, 0.71) reports. However, when SSIs were compared to an active control group, none of the reported measures showed a statistically significant effect of the SSI. Future research should prioritise enhancing study quality, particularly using active control conditions.

Developmental Neuroscience

Autism

social skills intervention

meta-analysis

adolescent

Autism spectrum disorder (ASD) is a neurodevelopmental condition that emerges early in life and persists into adulthood. The core symptoms of ASD, as defined by the American psychiatric association (APA) are social communication deficits and restricted, repetitive behaviours, interests, or activities, with associated functional impairment (American Psychiatric Association, 2013). The prevalence of ASD in Australia is estimated to be between 0.4 to 1.3% (Australian Bureau of Statistics, 2018), with an estimated male-to-female ratio of 3:1 (Loomes et al., 2017). The global prevalence of ASD has increased from less than 0.05% in the 1970s, to 0.8% in 2010, and to 2.3% in 2018 (Nielsen et al., 2013). The rising prevalence is likely multi-factorial, including increased awareness and acceptance of ASD, changing diagnostic criteria, and increases in diagnostic and support services (Hansen et al., 2015).

As children transition into adolescence and the social expectations increase, non-verbal communication and social reciprocity may worsen in some cases, making adolescence an especially challenging period (Shattuck et al., 2007). Changes in ASD symptom expression in an individual, may be further compounded by the lag in emotional, cognitive and social maturation and changes in their environment (e.g. transition into higher education, with increased social demands, or changes in family dynamics) (Seltzer et al., 2003). Further challenges may arise in adolescence due to changes in social networks, increased academic expectations, and introduction into the workplace. Teenagers with ASD face more adverse social outcomes, such as bullying, fewer friends, and limited understanding from peers, resulting in isolation. Within the school setting, ASD students tend to have lower levels of self-direction and reduced social interaction, when compared to their peers (Liu et al., 2018), and teachers feel under-prepared to meet their needs (Knight et al., 2019). Parent reports indicate increased concern in areas of social skills leading to worries about future independence and functioning (Vuattoux et al., 2021). Young adults (18–25 years old) may face persistent social skills impairments, which could be masked by improving intelligence, but can still lead to problems with daily living tasks, employment, and relationships (Gantman et al., 2012).

Current behavioural interventions for adolescents and young people include cognitive behavioural therapy (CBT), social skills training, self-regulation interventions, and transition planning programs. This systematic review specifically focused on the effects of social skills interventions (SSI) in adolescents and young adults, within the age range of 12 to 25 years. SSIs primarily focus on the core autism traits, rather than on co-morbid mental illness (e.g. depression and anxiety), emotional dysregulation, or employment attainment. Hence, other behavioural interventions were excluded as they report on outcome measures outside the scope of this review. Since social skills assessment can vary between subjective and objective reports, this review included multiple reports (e.g. observed, parent, teacher, and adolescent reports) to capture the true effect of SSI. In addition, the variation in reporting can reveal differences in symptom expression and perception of its effects (Gates et al., 2017).

Past meta-analysis on SSIs among youth with ASD included: (i) different age range (5–35 years old) (Bellini et al., 2007; Gates et al., 2017; Hirvikoski et al., 2015; Reichow et al., 2012; Wang and Spillane, 2009; Wolstencroft et al., 2018); (ii) non-RCTs (Babb et al., 2021; Bellini et al., 2007; Wang and Spillane, 2009); (iii) greater variability in intervention length (5–12 sessions) (Gates et al., 2017); (iv) different types of intervention based on age (Babb et al., 2021; Gates et al., 2017; Wang and Spillane, 2009); and (v) only studies with specific outcome measures (Wolstencroft et al., 2018). In general, SSI resulted in improved social skills, with variable effect sizes. However, because all the meta-analyses included children younger than 12 years old, the results are not necessarily generalisable to adolescents. This is due to differing emotional and social development, increased social demands and expectations in adolescence, and varied social settings (Hirvikoski et al., 2015). In addition, interventions for pre-schoolers and primary school-aged children target simple social tasks, such as greetings, sharing play and or toys, and joining a peer group. Whilst adolescents undergo complex training, such as social cognition, conflict resolution tasks, and learning about differing perspectives of their peers (Hirvikoski et al., 2015). Hence, better understanding of the effects of SSI among adolescents could guide future research and implementation of evidence-based interventions specific for adolescents within the home and school settings. With rising rates of adolescents with ASD, there is greater importance for evidence-based interventions targeting the needs of adolescents (Lounds Taylor et al., 2012).

To address this knowledge gap, this study aimed to (i) conduct a systematic review and meta-analysis to examine the efficacy of SSIs for youth diagnosed with ASD aged 12–25 years and (ii) assess whether the effects differed based on the source of the report (i.e., observed, self, parent, and teacher reports).

This review was registered with PROSPERO (CRD42022303329) and conformed to the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) recommendations which were used as a framework (Moher et al., 2015) (Fig. 1).

Search strategy

A search was conducted to identify relevant studies in APA PsychINFO, Academic Search Complete, CINAHL Plus with Full Text, Education Source, Psychology and Behavioral Sciences Collection, ERIC, Pubmed, Ovid MEDLINE, Educational administration Abstracts and SocINDEX with Full Text.

The following search terms and sequences were used to identify potentially eligible studies: “Autis” OR ‘Autism Spectrum Disorder” OR “Pervasive Developmental Disorder” OR “Aspergers” AND “Intervention” OR “therapy” OR “teach” OR “treat” OR “program” OR “package” AND “assign* OR “control group” OR “business as usual” OR “wait list” OR “randomised controlled trial” OR “randomized controlled trial” OR random* OR quasi OR pragmatic OR “treatment group” OR “intervention group” OR “group design” OR “trial”.

Eligibility criteria

Studies were considered for inclusion based on the following criteria: (i) Original quantitative research article published in a peer reviewed journal in English, from 2000 onwards, as majority of randomised controlled trials (RCT) in autism were published after 2010 (French); (ii) Participant age (between 12 to 25 years), (iii) Diagnosis of ASD, or pervasive developmental disorder (PDD), (iv) RCT or quasi-experimental design studies, with either waitlist, treatment as usual, or active control groups, (v) Studies focusing on SSIs; (vi) Outcome measures linked to core ASD traits; and (vii) Studies that included ASD and comorbid diagnosis (e.g. intellectual disability) were included if the intervention targeted social skills.

Studies were excluded if: (i) The age inclusion was too wide, despite being inclusive of the 12–25 year group; (ii) The intervention was a transition program from primary to high school, or high school to University; (iii) If outcome measures did not include core ASD traits; and (iv) Studies reported using study designs other than RCT or quasi-experimental design studies.

Study selection

Rayyan, a web-based screening tool for systematic reviews, was used to screen titles and abstracts (Ouzzani et al., 2016). All the articles were reviewed by two independent reviewers (R.V. and WT.W. ), where their screening process was blinded to the other reviewer. Discrepancies between the reviewers were resolved via discussion to reach consensus, and where consensus could not be reached, this was resolved by a third reviewer (P. IL.).

Data extraction

Data extraction was carried out by one reviewer (R.V.), summarising the following: author details, country of the study, baseline characteristics, study design, comparison groups, and outcome measurement tools. Authors of studies with missing data were contacted, of which one provided additional data.

Quality assessment and Risk of Bias

Two reviewers (R.V. and WT.W.) evaluated the methodological validity of the included studies, using the RCT Joanna Briggs Institute (JBI) critical appraisal checklist (Peters et al., 2015). Quality of studies were rated as good (≥ 8), fair (6–7), or poor (≤ 5) based on the summary scores. Additionally, the quality of evidence was determined by applying the Grading of Recommendations Assessment, Development and Evaluation (GRADE) criteria (Guyatt et al., 2008). The overall GRADE quality was based on quality of studies, heterogeneity between the studies, and number of studies contributing to each outcome measure.

Outcome measures

The focus of this review was the effect of SSIs on core symptoms of autism. It is to be noted that the studies included in this review primarily focussed on social skills, using a number of scales as below (also see Supplementary Table S1):

Observed measures: Contextual Assessment of Social Skills (CASS) (Ratto et al., 2011), Autism Diagnostic Observation Schedule (ADOS) social score (Lord et al., 2000), and Ko’s mutual engagement rating (Ko et al., 2019)

Adolescent social knowledge measures: Test of Adolescent Social skills Knowledge (TASSK) (Laugeson et al., 2010), TASSK-R (revised, unpublished version), and Test of Young Adult Social Skills Knowledge (TYASSK) (Laugeson, 2017)

Adolescent social performance measures: Social Skills Improvement System (SSIS) (Gresham and Elliott, 2008), Quality of socialisation questionnaire (QSQ) (Laugeson, 2017), Chou’s social problem-solving measure (Chou, 2020), and Goal Attainment Scale (GAS) (Kiresuk et al., 1994)

Parent measures: SSIS (Gresham and Elliott, 2008), Quality of Play Questionnaire (QPQ) (Frankel and Myatt, 2013), and Social Skills Rating System (SSRS) (Gresham and Elliott, 1990)

Teacher measures: Social Responsiveness Scale (SRS) (Constantino and Gruber, 2012), SSIS, GAS, and Chou’s social problem-solving measure.

Data analysis

Previously reported data from various studies were synthesised. The standardised mean difference (SMD), between the groups, was calculated as the effect size for the outcome measure (Viechtbauer, 2010), for continuous data and different scale measures. In addition, the inverse-variance method of random effects model was used due to variability between interventions and populations (Higgins and Thompson, 2002). Statistical heterogeneity of the studies was calculated using I². All analyses were conducted using RevMan version 5.4 (The Cochrane Collaboration, 2020), and SPSS (IBM Corp, 2007).

Search results

The search strategy retrieved 6232 articles, of which 2599 were duplicates. Two reviewers (R.V. and WT.W.) screened articles by reading the titles and abstracts, resulting in 56 articles determined to be eligible for full-text assessment. Finally, 15 studies met the eligibility criteria and were included in the systematic review (Fig. 1).

(Insert Fig. 1)

Study characteristics

Characteristics of the included studies (Table 1), showed that only one (Laugeson et al., 2014) out of the 15 studies was a non-randomised study, whilst the others were all RCTs. Collectively, there were 1244 participants enrolled, of which around 80% were males. Sample sizes recruited into the studies ranged from 20 to 110 participants. Studies were conducted in the United states of America (Gantman et al., 2012; Hume et al., 2022; Ko et al., 2019; Laugeson et al., 2009; Laugeson et al., 2014; Laugeson et al., 2015; Matthews et al., 2018; Matthews et al., 2020; Vernon et al., 2018), Israel (Rabin et al., 2018, Rabin et al., 2020), Korea (Yoo et al., 2014), Taiwan (Chou, 2020), Australia (Afsharnejad et al., 2022) and Netherlands (Idris et al., 2022). Twelve studies included youth aged between 12 to 18 years (Afsharnejad et al., 2022; Chou, 2020; Idris et al., 2022; Ko et al., 2019; Laugeson et al., 2009; Laugeson et al., 2014; Matthews et al., 2018; Matthews et al., 2020; Rabin et al., 2018; Rabin et al., 2020; Vernon et al., 2018; Yoo et al., 2014), two studies included youth aged between 18 to 23 years old (Gantman et al., 2012; Laugeson et al., 2015), and one study only reported a mean age of 16.2 years (Hume et al., 2022). One study was clinician-led (Afsharnejad et al., 2022), three were teacher-led (Chou, 2020; Laugeson et al., 2014; Hume et al., 2022), nine were parent-led (Gantman et al., 2012; Idris et al., 2022; Laugeson et al., 2009; Laugeson et al., 2015; Matthews et al., 2018; Matthews et al., 2020; Rabin et al., 2018; Rabin et al., 2020; Yoo et al., 2014), and two were peer-led (Ko et al., 2019; Vernon et al., 2018).

(Insert Table 1)

All participants had a diagnosis of ASD. Eight studies (Afsharnejad et al., 2022; Idris et al., 2022; Ko et al., 2019; Laugeson et al., 2014; Matthews et al., 2018; Matthews et al., 2020; Rabin et al., 2018; Rabin et al., 2020) noted a formal diagnosis of ASD as per Diagnostic Statistical Manual-IV/5 (APA) or the ADOS, two studies (Chou, 2020; Hume et al., 2022) included participants who were receiving education and/or services for ASD, four studies (Gantman et al., 2012; Laugeson et al., 2009; Laugeson et al., 2015; Yoo et al., 2014) included participants who had a previous diagnosis of ASD by a reliable mental health clinician, and one study (Vernon et al., 2018) didn’t elaborate on how a diagnosis was obtained. Seven studies (Afsharnejad et al., 2022; Gantman et al., 2012; Laugeson et al., 2015; Matthews et al., 2020; Rabin et al., 2018; Rabin et al., 2020; Yoo et al., 2014) validated the ASD diagnosis by administering the ADOS or the Autism-spectrum quotient (AQ) (Baron-Cohen S et al., 2001) upon entering the study.

All the studies were included in the quantitative analysis, were separated into studies that compared the intervention with a waitlist control group (or services as usual group) and those that compared the intervention with an active control group. An active control involved participants being randomly assigned to a manualised structured group, which controlled for the exposure to a social group.

The average duration of the intervention was 21.1 weeks (SD = 23, ranging from 12 to 104 weeks). And the average duration of the weekly sessions was 92.3 minutes per week (SD = 29.91, ranging from 30 to 180min/week). All included studies targeted social skills as the primary outcome measure, and none targeted restricted interests or behaviours. A summary of findings is presented as Supplementary Table S2.

Quality assessment and Risk of Bias

Quality assessment and risk of bias for each study is outlined in Tables 2 and 3. Most studies ranged from fair (Chou, 2020; Gantman et al., 2012; Hume et al., 2022; Ko et al., 2019; Laugeson et al., 2015; Rabin et al., 2018; Yoo et al., 2014) to good (Afsharnejad et al., 2022; Idris et al., 2022; Laugeson et al., 2014; Rabin et al., 2020), however four were assessed as poor quality (Laugeson et al., 2015; Matthews et al., 2018; Matthews et al., 2020; Vernon et al., 2018). The quality of the evidence was assessed by the GRADE tool (Supplementary Table S2), where the outcome measures faired low to very low.

Table 2

Quality assessment of randomised controlled studies using Joanna Briggs Institute (JBI) critical appraisal checklist
Author and Year	Q1	Q2	Q3	Q4	Q5	Q6	Q7	Q8	Q9	Q10	Q11	Q12	Q13	Quality
Afsharnejad 2022	Y	U	N	NA	NA	Y	N	Y	Y	Y	Y	Y	Y	Good
Chou 2020	Y	U	N	NA	NA	NA	Y	U	Y	Y	Y	Y	Y	Fair
Gantman 2012	Y	U	N	NA	NA	NA	Y	Y	Y	Y	N	Y	Y	Fair
Hume 2022	Y	U	N	NA	NA	NA	Y	N	N	Y	Y	Y	Y	Fair
Idris 2022	Y	U	Y	NA	NA	Y	N	Y	N	Y	Y	Y	Y	Good
Ko 2018	Y	U	U	NA	NA	Y	Y	N	N	Y	Y	Y	Y	Fair
Laugeson 2009	U	U	N	NA	NA	NA	Y	N	Y	Y	N	Y	Y	Fair
Laugeson 2015	U	U	Y	NA	NA	NA	Y	N	N	Y	N	Y	Y	Poor
Matthews 2018	U	U	N	NA	NA	N	Y	N	N	Y	N	Y	Y	Poor
Matthews 2020	U	U	N	NA	NA	N	N	Y	Y	Y	N	Y	Y	Poor
Rabin 2018	U	U	N	NA	NA	Y	Y	N	N	Y	Y	Y	Y	Fair
Rabin 2020	Y	Y	Y	NA	NA	N	Y	Y	N	Y	Y	Y	Y	Good
Vernon 2018	U	U	N	NA	NA	NA	Y	Y	N	Y	N	Y	Y	Poor
Yoo 2014	Y	U	N	NA	NA	Y	Y	N	N	Y	Y	Y	Y	Fair

Table 3

Quality assessment of non-randomised controlled studies using JBI critical appraisal checklist
Author and Year	Q1	Q2	Q3	Q4	Q5	Q6	Q7	Q8	Q9	Quality
Laugeson 2014	Y	Y	Y	Y	N	Y	Y	N	Y	Good

(Insert Table 2)

(Insert Table 3)

Effects of intervention

Meta-analysis of four studies (Ko et al., 2019; Rabin et al., 2018; Rabin et al., 2020; Yoo et al., 2014) showed that youth with ASD who underwent a SSI demonstrated a significant observed improvement in social skills, as compared to the waitlist group, with a standardised mean difference (SMD) of 0.49 (95% CI 0.20, 0.78; p-value = 0.0008) (Fig. 2). The GRADE of evidence for the observed measure was classified as low (Supplementary Table S2), due to limited confidence in the precision of the evidence, secondary to small sample sizes.

(Insert Fig. 2)

Meta-analysis of seven studies (Gantman et al., 2012; Laugeson et al., 2009; Laugeson et al., 2015; Matthews et al., 2018; Rabin et al., 2018; Rabin et al., 2020; Yoo et al., 2014), showed outcome in favour of SSI, based on adolescent reports of their social knowledge (SMD = 2.03, 95% CI [1.15; 2.91], p-value = 0.0001) (Fig. 3). However, there was substantial heterogeneity among the included studies (I² = 86%). During sensitivity analysis, no single cause (such as age, gender distribution, geographical location, or outcome measure) could completely account for the heterogeneity. Given this degree of heterogeneity, the GRADE of evidence was classified as very low quality (Supplementary Table S2).

(Insert Fig. 3)

An additional meta-analysis was conducted on seven studies (Chou, 2020; Laugeson et al., 2009; Laugeson et al., 2015; Matthews et al., 2018; Rabin et al., 2018; Vernon et al., 2018; Yoo et al., 2014) based on adolescent reports of their social performance, which showed in favour of the intervention group (SMD = 0.56, 95% CI [0.19; 0.93], p-value = 0.003) (Fig. 4).

(Insert Fig. 4)

Seven studies (Gantman et al., 2012; Laugeson et al., 2009; Laugeson et al., 2015; Matthews et al., 2018; Rabin et al., 2020; Vernon et al., 2018; Yoo et al., 2014) reported on parent observations, which showed significant improvements in social engagement in the intervention group (SMD = 0.42, 95% CI [0.13, 0.71], p-value = 0.004) (Fig. 5). The GRADE of evidence was rated as low, due to risk of bias of the included studies (secondary to unclear randomisation and concealment processes, lack of blinding, and per-protocol analysis) (Supplementary Table S2).

(Insert Fig. 5)

The pooled estimate of three studies (Chou, 2020; Hume et al., 2022; Rabin et al., 2020) showed that teacher reports were in favour of SSI, however they failed to reach significance (Supplementary Figure S1). In addition, there was substantial heterogeneity between the studies, secondary to differences in geographical locations, type of interventions (in addition, two were teacher-led, and one was parent-led), and outcome measures. Hence, the GRADE of evidence was rated as very low quality (Supplementary Table S2).

Four studies compared the SSI with an active control group (Afsharnejad et al., 2022; Idris et al., 2022; Laugeson et al., 2014; Matthews et al., 2020). One study (Idris et al., 2022) reported a non-significant effect, not in favour of the intervention group, based on an observed measure. The adolescent social knowledge (Laugeson et al., 2014; Matthews et al., 2020), social-performance (Afsharnejad et al., 2022; Idris et al., 2022; Laugeson et al., 2014; Matthews et al., 2020) and parent reports (Idris et al., 2022; Matthews et al., 2020) showed in favour of the social intervention, however failing to reach statistical significance (Supplementary Figures S2, S3 and S4). And, the teacher report, from two studies (Idris et al., 2022; Laugeson et al., 2014) did not favour the intervention group (Supplementary Figure S5).

Follow-up measure

Seven studies reported on a follow-up measure, but only two studies (Afsharnejad et al., 2022; Idris et al., 2022) permitted an analysis. The waitlist control studies compared the follow up measure with the baseline measure of all participants, but not between groups. This wasn’t feasible, as the waitlist control group also received the intervention after a wait time. The two included studies in this analysis were those with active controls. Adolescent social-performance measure showed an effect in favour of the SSI, however this failed to reach significance (Supplementary Figure S6).

Publication bias

Outcome measures with six or more studies were inspected for publication bias visually on the funnel plot and statistically from Egger’s test of asymmetry (Egger et al., 1997). The funnel plots do not indicate obvious signs of asymmetry (Supplementary Figures S7, S8 and S9), and this was supported by the Egger’s test, where all outcomes had a p > 0.05, suggestive of funnel plot symmetry (Supplementary Table S2).

Implementation fidelity

Thirteen studies commented on implementation fidelity. Fidelity was measured on either a daily or weekly check, dependent on the structure of the study. Adherence was measured via a variety of methods, such as observed role-play scenarios during weekly group sessions (Gantman et al., 2012; Laugeson et al., 2015; Rabin et al., 2020; Yoo et al., 2014) checklists (Afsharnejad et al., 2022; Hume et al., 2022; Idris et al., 2022; Laugeson et al., 2009; Matthews et al., 2018; Matthews et al., 2020), and self-assessment (Laugeson et al., 2014). Three studies (Idris et al., 2022; Ko et al., 2019; Vernon et al., 2018) commented on the implementation, which indicated that between 91 to 97% of the time, procedures were accurately implemented.

Summary of main results

This systematic review comprehensively summarised current evidence on the effectiveness of SSIs on core ASD traits among youth diagnosed with ASD aged 12–25 years. There were eleven studies that compared the effects of SSI to a waitlist control group, and four studies that compared with an active control group. The results showed that SSIs led to improvements in core social skills outcomes, based on observed, adolescent, and parent reports, although not on teacher reports. When compared to the active control group, none of the reported measures showed a statistically significant effect of the SSI.

Quality of evidence

Quality of the evidence, using GRADE (Supplementary Table S2), was jeopardised due to poor quality of studies, heterogeneity in population samples, and small sample sizes. The nature of the intervention limited the ability to blind participants, intervention providers, and outcome assessors (i.e., teachers, parents, or adolescents). In addition, the provision of an observed and blinded outcome measure was limited due to financial constraints. Furthermore, the use of teacher and parent reports are not considered reliable due to high risk of bias. Another factor to consider is that the majority of the included studies were of the UCLA PEERS intervention (Gantman et al., 2012; Idris et al., 2022; Laugeson et al., 2009; Laugeson et al., 2014; Laugeson et al., 2015; Matthews et al., 2018; Matthews et al., 2020; Rabin et al., 2018; Rabin et al., 2020; Vernon et al., 2018), which involved the developer of the intervention, who received financial benefits from the sales of the training manual.

Clinical relevance

This review showed a large effect size for self-report, small effect size for parent and observer reports, and non-significant effect size for teacher reports. It is hard to discern the true effect size from the intervention given the varying effect sizes between observed (0.49), parent (0.42) and adolescent (2.03 and 0.56) reports.

The adolescent social knowledge measure (TASSK (Laugeson et al., 2010)) effect size was much larger than the parent or lab-observed measure, which could be due to participants overestimating their social functioning (Lerner et al., 2012) or due to other reports failing to capture the social improvements. However, the meta-analysis on adolescent report of social performance (0.56) had a similar effect size to parent report (0.42). This could be secondary to adolescents being unable to apply the gained knowledge or due to gained knowledge taking time to implement, which may have been better assessed by a follow-up measure.

The parental report, though positive, is a biased measure and possibly other factors contributed to the results, such as the expectancy effect. This is where time and emotional investment, and faith in the intervention, may lead to positive parent outcome reporting. Another important consideration is whether the impact of the intervention on the parent and the parent-child relationship influenced parent outcome reporting. This was noted by Rabin et al, who identified that parents who had reported lower levels of support for their child’s distress at the start of the SSI made greater social gains (Rabin et al., 2020).

Teacher outcome reports failed to reach significance, in both the waitlist and active control studies. Factors that could have contributed to this finding may be the poor sensitivity of the measures used or despite social gains, the gains were not detected in schools due to greater symptom expression (as opposed to home), because of increased social demands and negative bias from peers (Rabin et al., 2020).

Sensitivity analysis showed that teacher reports in teacher-led studies (Chou, 2020; Hume et al., 2022; Laugeson et al., 2014) had a greater effect size (when omitting the parent-led study (Rabin et al., 2020)), but still failed to reach significance. The greater effect size could be explained by parent-led interventions lacking generalisability in the school setting or an over-estimation of effect related to expectancy bias in the teacher measures and lack of blinding of outcome measures. Interventions with role-playing and interactive sessions were more effective than instructive interventions (January et al., 2011). Hence, school-based interventions may have an advantage, as teachers and peers can give active social coaching and feedback on observed interactions. The school setting may also be more accessible to families and serve as a better setting to assess for maintenance of effect and dose-response relationship of SSIs.

There has been concern about the ethics of assigning motivated participants to a waitlist control group, and whether waitlist groups over-estimate the treatment outcome, due to a placebo effect (Cunningham et al., 2013). Hence, active control groups are used to provide immediate intervention and to account for the confounding effect of access to a socially supportive group. Any shared-activity interventions (i.e. participant and teacher/parent/peer/clinician) tend to lead to improved social outcomes in the short term (Afsharnejad et al., 2022), due to indirectly influencing the youth’s social motivation, interaction, and problem-solving skills. Nonetheless, in all the studies that utilised an active control group, the active control group did not include any social skills training or teaching and served to identify treatment effects specific to the SSI. Despite this, analysis showed no difference between the SSI and active control groups. This could be secondary to limited power, heterogeneity between the studies, and variation in outcome measures.

Strengths and Limitations

This study provides the most up to date evidence on the efficacy of SSIs on core ASD traits among youth diagnosed with ASD aged 12–25 years. Other strengths include a robust, peer-reviewed search strategy with the use of nine databases to maximise the identification of the eligible studies. However, this review also has several limitations. First, non-English language studies and grey literature sources were not included in the search, due to possible erroneous conclusions secondary to improper translations, and lack of peer-review process, respectively. Second, the quality of the included studies was compromised due to majority of them failing to outline randomisation and concealment processes and being unable to blind participants (as well as parents and teachers) due to the nature of the intervention. Third, there was substantial heterogeneity in population demographics (such as the lack of consistent validation of ASD/PDD diagnosis, a continuum of intellectual abilities, location of studies), the type of intervention, and outcome measures, which compromises the generalisability of the results. Ideally outcomes should be objectively measured by blinded, trained assessors, but this is an expensive and labour-intensive process, hence is often omitted. Generalisability is further limited, as only five studies (Idris et al., 2022; Matthews et al., 2020; Rabin et al., 2018; Rabin et al., 2020; Vernon et al., 2018) obtained third-party outcome measures, to assess the validity of the SSI in a different setting.

Fourth, the included studies lacked consistent reporting on implementation fidelity and its role in moderating the relationship between the SSI and outcomes. Studies that have shown high implementation fidelity had improved efficacy, as opposed to studies with little to no monitoring of fidelity (Carroll et al., 2007). Fifth, moderator analysis couldn’t be effectively conducted in this review, due to limited power and lack of patient-level data, contributing to the review’s methodological limitation. Sensitivity analysis was conducted, accounting for intervention type (SSI and who conducted the intervention), age group, location, and quality of the study. However, this did not alter the overall result of the meta-analysis. Sixth, this review was unable to provide an understanding on the maintenance effects of SSIs. A waitlist control study restricts the ability to truly gauge the effects of the maintenance effects of the SSI, as both groups receive the intervention. However, an active control group or treatment as usual control is more capable of assessing the maintenance effects of the SSI. Two of the four active control studies provided follow-up data, however due to limited data it was not possible to conclude whether SSI resulted in maintaining social skills improvement. Data was limited due to small sample size (further complicated by attrition in the follow up period), and non-comparable follow up periods and outcome measure.

Lastly, the use of social competence as the only outcome measure, may not be representative of other ASD domains and symptoms. The included studies did not have consistent nor comparable secondary outcome measures. In addition, use of questionnaires is a limitation, due to variable interpretations of the questions based on who is completing, reporting on only observed behaviour in a particular setting (e.g. home or school), and use of unblinded reporters (i.e., parents or teachers involved in the study). Despite these limitations, the scales used included a range of constructs (e.g. communication, assertion, engagement, problematic behaviours, empathy), which captures a range of social domains. Furthermore, they are cost-effective and show different perspectives through the use of parallel versions.

Future directions

This review has implications for clinical practice, as it reinforces past meta-analysis (Bellini et al., 2007; Gates et al., 2017; Hirvikoski et al., 2015; Reichow et al., 2012; Wang and Spillane, 2009; Wolstencroft et al., 2018), however this review is specific for youths with ASD, as opposed to children. It shows promising trends in the administration of SSI for young people with ASD, as it targets core symptoms and is crucial for ongoing emotional and social maturation. However, in order to improve efficacy, recommendations for clinicians include strict adherence to intervention protocols and monitoring of implementation fidelity to achieve consistent and desired outcomes. And the need to assess the maintenance of intervention effects over time, in order to understand the lasting impact on the participant’s social skills development.

Recommendations for future investigators include, improving study quality with appropriate randomisation and concealment methods, standardised critical intervention components and duration, and blinded outcome measures. In addition, future research should consider the implementation of active control groups and assessing the maintenance of effects over time with the use of standardised follow-up measures. Future RCTs should also consistently report on implementation fidelity to better understand its moderating influence on the relationship between the SSI and outcomes. Lastly, the benefits of the school setting should be explored, as it can increase sample size and enhance accessibility of the intervention.

The main findings of the meta-analysis reveal that SSIs result in significant improvement in social skills, when compared to a waitlist control group, at least based on observed objective measure, two subjective measures of self, and parent reports. However, the applicability of the evidence is limited due to the low quality of the studies, source of reporting and heterogeneity between the studies. Future recommendations include improving study quality, use of follow up measures, considering schools as an intervention setting, and monitoring of implementation fidelity.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship and/or publication of this article.

Author contributions

Conceptualisation: Valsamma Eapen and Ping-I Lin; Methodology: James Rufus John, Ping-I Lin; Formal analysis and investigation: Ramya Velalagan, Weng Tong Wu, James Rufus John; Writing - original draft preparation: Ramya Velalagan;Writing - review and editing: Ramya Velalagan, Weng Tong Wu, James Rufus John, Ping-I Lin, and Valsamma Eapen; Supervision: Ping-I Lin and Valsamma Eapen.

Afsharnejad B, Falkmer M, Black MH, Alach T, Lenhard F, Fridell A, Coco C, Milne K, Bölte S and Girdler S (2022) KONTAKT® social skills group training for Australian adolescents with autism spectrum disorder: A randomized controlled trial. European Child and Adolescent Psychiatry 31(11): 1695-1713.
American Psychiatric Association (2013). Diagnostic and statistical manual of mental disorders. 5th ed. Washington, DC.
Australian Bureau of Statistics (2019). Disability, ageing and carers, Australia: Summary of findings. Available at: https://www.abs.gov.au/statistics/health/disability/disability-ageing-and-carers-australia-summary-findings/latest-release (accessed 31 August 2023).
Babb S, Raulston TJ, McNaughton D, Lee JY and Weintraub R (2021) The Effects of Social Skill Interventions for Adolescents With Autism: A Meta-Analysis. Remedial and Special Education 42(5):343–357.
Baron-Cohen S, Wheelwright S, Skinner R, Martin J and Clubley E (2001) The autism-spectrum quotient (AQ): Evidence from Asperger syndrome/high-functioning autism, males and females, scientists and mathematicians. Journal of Autism and Developmental Disorders 31(1): 5-17.
Bellini S, Peters JK, Benner L and Hopf A (2007) A Meta-Analysis of School-Based Social Skills Interventions for Children with Autism Spectrum Disorders. Remedial and Special Education 28(3):153–162
Carroll C, Patterson M, Wood S, Booth A, Rick J and Balain S (2007). A conceptual framework for implementation fidelity. Implementation Science 2(1):40.
Chou YC (2020) Navigation of Social Engagement (NOSE) Project: Using a Self-Directed Problem Solving Model to Enhance Social Problem-Solving and Self-Determination in Youth with Autism Spectrum Disorders. Education and Training in Autism and Developmental Disabilities 55(1):101-114.
Constantino J and Gruber C (2012). Social Responsiveness Scale (SRS-2). Western Psychological Services, Torrance, CA.
Cunningham JA, Kypri K and McCambridge J (2013) Exploratory randomized controlled trial evaluating the impact of a waiting list control design. BMC Medical Research Methodology 13(1):150
Egger M, Smith GD, Schneider M and Minder C (1997) Bias in meta-analysis detected by a simple, graphical test. British Medical Journal 315(7109):629-634.
Frankel FD and Myatt RJ (2013). Children’s friendship training. Routledge.
Gantman A, Kapp SK, Orenski K and Laugeson EA (2012). Social skills training for young adults with high-functioning autism spectrum disorders: A randomized controlled pilot study. Journal of Autism and Developmental Disorders 42(6):1094–1103
Gates J, Kang E and Lerner M (2017). Efficacy of group social skills interventions for youth with autism spectrum disorder: A systematic review and meta-analysis. Clinical Psychology Review (52):164–81.
Gresham F and Elliott S (1990). Social skills rating system manual. American Guidance Service.
Gresham F and Elliott S (2008). Social skills improvement system: Rating scales. Pearson Assessments, Bloomington, MN
Guyatt G, Oxman A, Vist G, Kunz R, Falck-Ytter Y, Alonso-Coello P and Schünemann, H (2008) GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. British Medical Journal 336(7650):924–926.
Hansen SN, Schendel DE and Parner ET (2015) Explaining the Increase in the Prevalence of Autism Spectrum Disorders: The Proportion Attributable to Changes in Reporting Practices. JAMA Pediatrics 169: 56-62.
Higgins J and Thompson S (2002). Quantifying heterogeneity in a meta-analysis. Statistics in Medicine 21:1539–1558.
Hirvikoski T, Jonsson U, Halldner L, Lundequist A, Schipper E, Nordin V and Bolte S (2015) A Systematic Review of Social Communication and Interaction Interventions for Patients with Autism Spectrum Disorder. Scandinavian Journal of Child and Adolescent Psychiatry and Psychology 3:147–168.
Hume K, Odom SL, Steinbrenner JR, Smith DaWalt L, Hall LJ, Kraemer B, Tomaszewski B, Brum C, Szidon K and Bolt DM (2022). Efficacy of a School-Based Comprehensive Intervention Program for Adolescents With Autism. Exceptional Children 88(2):223–240
IBM Corp. (2007) IBM SPSS Statistics for Windows (17.0) [Computer software].
Idris S, van Pelt BJ, Jagersma G, Duvekot J, Maras A, van der Ende J, van Haren N and Greaves-Lord K (2022) A randomized controlled trial to examine the effectiveness of the Dutch version of the Program for the Education and Enrichment of Relational Skills (PEERS®). BMC Psychiatry 22(1):293.
January AM, Casey RJ and Paulson D (2011) A meta-analysis of classroom-wide interventions to build social skills: Do they work? School Psychology Review 40(2):242–256.
Kiresuk T, Smith A and Cardillo J (1994). Goal attainment scaling: Applications, theory, and measurement. Lawrence Erlbaum Associates, Inc. NY
Knight VF, Huber HB, Kuntz EM, Carter EW and Juarez AP (2019) Instructional Practices, Priorities, and Preparedness for Educating Students with Autism and Intellectual Disability. Focus on Autism and Other Developmental Disabilities 34: 3-14.
Ko JA, Miller AR and Vernon TW (2019) Social conversation skill improvements associated with the Social Tools And Rules for Teens program for adolescents with autism spectrum disorder: Results of a randomized controlled trial. Autism: The International Journal of Research and Practice 23(5):1224–35.
Laugeson, E. (2017) PEERS® for Young Adults: Social Skills Training for Adults with Autism Spectrum Disorder and Other Social Challenges. 1st ed. Routledge, NY
Laugeson EA, Ellingsen R, Sanderson J, Tucci L and Bates S (2014) The ABC’s of teaching social skills to adolescents with autism spectrum disorder in the classroom: The UCLA PEERS (®) Program. Journal of Autism and Developmental Disorders 44(9):2244–56.
Laugeson E and Frankel F (2010) Test of Adolescent Social Skills Knowledge, Mental Status Checklist. The PEERS Treatment Manual. Routledge, NY
Laugeson EA, Frankel F, Mogil C, Dillon AR (2009). Parent-Assisted Social Skills Training to Improve Friendships in Teens with Autism Spectrum Disorders. Journal of Autism and Developmental Disorders 39(4):596–606.
Laugeson EA, Gantman A, Kapp SK, Orenski K and Ellingsen R (2015) A Randomized Controlled Trial to Improve Social Skills in Young Adults with Autism Spectrum Disorder: The UCLA PEERS(®) Program. Journal of Autism and Developmental Disorders 45(12):3978–3989.
Lerner MD, Calhoun CD, Mikami AY and De Los Reyes A (2012) Understanding Parent–Child Social Informant Discrepancy in Youth with High Functioning Autism Spectrum Disorders.Journal of Autism and Developmental Disorders 42(12):2680–2692.
Lipscomb S, Haimson J, Liu AY, Burghardt J, Johnson DR and Thurlow M (2017) Preparing for life after high school: The characteristics and experiences of youth in special education. Findings from the National Longitudinal Transition Study 2012. Volume 1: Comparisons with other youth. Report for the Department of Education, Institute of Education Sciences, National Center for Education Evaluation and Regional Assistance. Report no. NCEE 2017-4016, March. Washington, DC.
Loomes R, Hull L and Mandy WPL (2017) What Is the Male-to-Female Ratio in Autism Spectrum Disorder? A Systematic Review and Meta-Analysis. Journal of the American Academy of Child and Adolescent Psychiatry 56(6):466-474.
Lord C, Risi S, Lambrecht L, Cook EH, Leventhal BL, DiLavore PC, Pickles A and Rutter M (2000) The Autism Diagnostic Observation Schedule—Generic: A standard measure of social and communication deficits associated with the spectrum of autism. Journal of Autism and Developmental Disorders 30:205–223.
Lounds Taylor J, Dove D, Veenstra-VanderWeele J, Sathe NA, McPheeters ML, Jerome RN and Warren Z (2012) Interventions for Adolescents and Young Adults With Autism Spectrum Disorders. Comparative Effectiveness Review No. 65. AHRQ Publication No. 12-EHC063-EF, Agency for Healthcare Research and Quality. August. Rockville, MD.
Matthews NL, Laflin J, Orr BC, Warriner K, DeCarlo M and Smith CJ (2020). Brief Report: Effectiveness of an Accelerated Version of the PEERS® Social Skills Intervention for Adolescents. Journal of Autism and Developmental Disorders 50(6):2201–2207.
Matthews NL, Orr BC, Warriner K, DeCarlo M, Sorensen M, Laflin J and Smith CJ (2018). Exploring the Effectiveness of a Peer-Mediated Model of the PEERS Curriculum: A Pilot Randomized Control Trial. Journal of Autism and Developmental Disorders 48(7):2458–2475.
Moher D, Shamseer L, Clarke M, Ghersi D, Liberati A, Petticrew M and Stewart L (2015). Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Systematic reviews 4(1).
Nielsen TC, Nassar N, Boulton KA, Guastella AJ and Lain SJ (2013) Estimating the Prevalence of Autism Spectrum Disorder in New South Wales, Australia: A Data Linkage Study of Three Routinely Collected Datasets. Journal of Autism and Developmental Disorders 54:1558–1566.
Ouzzani M, Hammady H, Fedorowicz Z and Elmagarmid A (2016). Rayyan—A web and mobile app for systematic reviews. Systematic Reviews, 5(210).
Peters M, Godfrey C, McInerney P, Soares C, Khalil H and Parker D (2015). The Joanna Briggs Institute reviewers’ manual 2015: Methodology for JBI scoping reviews.
Rabin SJ, Israel-Yaacov S, Laugeson EA, Mor-Snir I and Golan O (2018). A randomized controlled trial evaluating the Hebrew adaptation of the PEERS(®) intervention: Behavioral and questionnaire-based outcomes. Autism Research : Official Journal of the International Society for Autism Research 11(8):1187–1200.
Rabin SJ, Laugeson EA, Mor-Snir I and Golan O (2020). An Israeli RCT of PEERS®: Intervention Effectiveness and the Predictive Value of Parental Sensitivity. Journal of Clinical Child & Adolescent Psychology 50(6):933–949.
Ratto AB, Turner-Brown L, Rupp BM, Mesibov GB and Penn DL (2011). Development of the Contextual Assessment of Social Skills (CASS): A Role Play Measure of Social Skill for Individuals with High-Functioning Autism. Journal of Autism and Developmental Disorders 41(9):1277–1286.
Reichow B, Steiner AM and Volkmar F (2012) Social skills groups for people aged 6 to 21 with autism spectrum disorders (ASD). Cochrane Database of Systematic Reviews (7):CD008511.
Seltzer MM, Krauss MW, Shattuck PT, Orsmond G, Swe A and Lord C (2003) The symptoms of autism spectrum disorders in adolescence and adulthood. Journal of Autism and Developmental Disorders 33(6):565–581.
Shattuck PT, Seltzer MM, Greenberg JS, Orsmond GI, Bolt D, Kring S, Lounds J and Lord C (2007) Change in Autism Symptoms and Maladaptive Behaviors in Adolescents and Adults with an Autism Spectrum Disorder. Journal of Autism and Developmental Disorders 37(9):1735-1747.
The Cochrane Collaboration (2020) Review Manager Web (RevMan Web). 5.4. Available at: Available at: https://training.cochrane.org/online-learning/core-software-cochrane-reviews/revman/revman-non-cochrane-reviews
Vernon TW, Miller AR, Ko JA, Barrett AC and McGarry ES (2018) A Randomized Controlled Trial of the Social Tools And Rules for Teens (START) Program: An Immersive Socialization Intervention for Adolescents with Autism Spectrum Disorder. Journal of Autism and Developmental Disorders 48(3):892–904.
Viechtbauer W (2010 Conducting Meta-Analyses in R with the metafor Package. Journal of Statistical Software 36(3):1–48.
Vuattoux D, Castiglia D and Chabane N (2021) Considerations and Perspectives on Interventions to Improve Social Cognition in Adolescents With ASD Without ID: Involving Parents in Social Skills Groups. Frontiers in Psychiatry 28;12.
Wang P and Spillane A (2009) Evidence-Based Social Skills Interventions for Children with Autism: A Meta-analysis. Education and Training in Autism and Developmental Disabilities 44:318–342.
Wolstencroft J, Robinson L, Srinivasan R, Kerry E, Mandy W and Skuse D (2018) A Systematic Review of Group Social Skills Interventions, and Meta-analysis of Outcomes, for Children with High Functioning ASD. Journal of Autism and Developmental Disorders 48(7):2293–2307.
Yoo HJ, Bahn G, Cho IH, Kim EK, Kim JH, Min JW, Lee WH, Seo JS, Jun SS, Bong G, Cho S, Shin MS, Kim BN, Kim JW, Park S and Laugeson EA (2014) A randomized controlled trial of the Korean version of the PEERS(®) parent-assisted social skills training program for teens with ASD. Autism Research: Official Journal of the International Society for Autism Research 7(1):145–161.

Table 1 is available in the Supplementary Files section.

The authors declare no competing interests.

Download PDF

Version 1

posted

You are reading this latest preprint version

Effectiveness of Social Skills Interventions in young people aged 12 to 25 years with Autism Spectrum Disorder: Systematic Review and Meta-analysis

Status:

Version 1

Abstract

Figures

Background

Methods

Search strategy

Eligibility criteria

Study selection

Data extraction

Quality assessment and Risk of Bias

Outcome measures

Data analysis

Results

Search results

(Insert Fig. 1)

Study characteristics

(Insert Table 1)

Quality assessment and Risk of Bias

(Insert Table 2)

(Insert Table 3)

Effects of intervention

(Insert Fig. 2)

(Insert Fig. 3)

(Insert Fig. 4)

(Insert Fig. 5)

Follow-up measure

Publication bias

Implementation fidelity

Discussion

Summary of main results

Quality of evidence

Clinical relevance

Strengths and Limitations

Future directions

Conclusion

Declarations

References

Tables

Additional Declarations

Supplementary Files

Status:

Version 1