Autism is currently one of the most visible and widely discussed human conditions. Its increased prevalence has brought it to the attention of society in the United States, with world-wide recognition (Lord et al.
2020). Much discussion surrounds the conceptualization of autism as a disability or as a set of unique skills that can be viewed as strengths (Urbanowicz et al.
2019). Although there is truth in both, there is also much verification that the life course for many individuals with autism, from infancy and into adulthood, is challenging for them and their families (Shattuck et al.
2018). In efforts to have a positive impact on this life trajectory, personnel in early intervention, schools, clinics, and other human service programs search for practices that could be most effective when working with children and youth with autism (Lai et al.
2020). The increased prevalence of autism (Maenner et al.
2020) has also intensified the demand for effective educational and therapeutic services, and intervention science is providing mounting evidence about practices that positively impact outcomes. The purpose of this study was to identify a set of focused intervention practices that have clear scientific evidence of positive effects with autistic
1 children and youth (i.e., evidence-based practices). For this paper, we define scientific evidence as reports in a peer-reviewed journal of an experimental study of acceptable methodological quality that addresses the efficacy of a focused intervention practice.
The imperative for establishing and continuing to update information about evidence-based practices (EBPs) is because the knowledge about the demographics, key ability features, and intervention science related to autism continues to advance. At the same time, there is also a purveyance of interventions that have little or no evidence of effectiveness and yet are described as “cutting edge” (Siri and Lyons
2014). For example, Paynter et al. (
2020) recently noted that the Autism Research website operated by the National Autism Society in the U.K. (
http://www.researchautism.net/autism-interventions/alphabetic-list-interventions) catalogued over 1000 interventions, and many of which lacked evidence. Lack of evidence does not mean that interventions are ineffective, as studies may not yet have been conducted. However, current human services policies in the U.S. requires that intervention practices have research evidence of their effectiveness (e.g., Individuals with Disabilities Education Act, Medicaid waiver provisions, insurance coverage regulations).
The contemporary focus on EBPs for children and youth with autism can be tracked back to Cochrane’s (
1972) proposition that health and medical services be based on empirical, scientific evidence of its efficacy. The movement to base health practices on scientific evidence gained further traction through Sackett’s and colleagues’ (
1996) advocacy of evidence-based medicine. An important contribution of the evidence-based medicine movement, which Cochrane also suggested, was that such identification and verification of evidence-based practice
is just the first step. The selection and application of such scientifically based practices depends on the skills and wisdom of the health care worker in selecting appropriate practices for the individual and applying them with fidelity. This multi-step process of blending information about scientifically identified, efficacious practices with practitioners’ knowledge and skill has been adopted in the evidence-based movements in education (Davies
1999; Odom et al.
2005), psychology (American Psychological Association
2006) and other human services (American Speech and Hearing Association
2004).
Evidence-Based Intervention Approaches
Focused intervention practices and comprehensive program models (i.e., previously identified as comprehensive
treatment models and modified because of the potentially ableist language implied) are two broad classes of interventions that appear in the research literature (Smith
2013). Focused intervention practices are designed to address a single skill or goal of a student with autism (Odom et al.
2010a,
b). These practices are operationally defined and address specific learner outcomes. Teachers, clinicians, or other practitioners select and use the practices (e.g., prompting, reinforcement, time delay) in interventions or instruction that addresses a learners individual learning goal. Focused intervention practices could be considered the building blocks of educational programs for children and youth with autism.
In contrast, comprehensive program models consist of a set of practices designed to achieve a broad learning or developmental impact on the core deficits of autism (Odom et al.
2010a,
b). Comprehensive programs are organized around a conceptual framework, procedurally manualized, focus on a breadth of outcomes, and are implemented with a substantial number of hours per week across one or more years (Odom et al.
2014). Examples of such programs are the early intensive behavior intervention program based on the UCLA Young Autism Project (Smith et al.
2000), the LEAP preschool model (Strain and Bovey
2011), and the Early Start Denver Model (Dawson et al.
2012). Teachers or other professionals adopting comprehensive program models must commit to being trained and implementing the entire model that might replace a current program or in addition a current program. Because of the differences between the two classes of interventions, and to need to specify a clearly articulated and practical focus for the review, comprehensive program models were not included in this review.
Previous Literature Reviews of EBPs for Children and Youth with Autism
Before the mid-2000s, the identification of EBPs for children and youth with autism was accomplished through narrative reviews by an individual or set of authors or organizations (e.g., Simpson
2005), which were useful, but did not follow a stringent review process. Many traditional systematic review processes, such as the Cochrane Collaborative (
https://www.cochrane.org/) or Project AIM (Sandbank et al.
2020), have only included studies that employed a randomized experimental group design (i.e., randomized control trial or RCT) and have excluded single case experimental design (SCD) studies. By excluding SCD studies, such reviews omit a vital experimental research methodology recognized as a valid scientific approach (What Works Clearinghouse
2020).
It is important to note that SCDs are often excluded in more traditional systematic reviews and meta-analyses, in part because such designs are thought to not be sufficiently scientific or rigorous. For example, on the previously noted Autism Research website (
http://www.researchautism.net/autism-interventions/other-aspects-autism-interventions/process-for-evaluating-studies/our-ratings-system), the National Autism Society describes SCD studies as “Grade C” methodology (i.e., contrasted with Grade A and B studies which are group designs). Other reviews have excluded SCD studies entirely (Sandbank et al.
2020). While an individual SCD study provides limited evidence of efficacy, multiple replication studies by different research groups build the strength of evidence. Systematic reviews that minimize or exclude SCD studies ignore the largest body of scientific findings about focused intervention practices. When excluding SCD research, researchers are ignoring the admonition by Sackett et al. (
1996): “… if no randomised trial has been carried out for our patient's predicament, we must follow the trail to the next best external evidence and work from there (p. 72).”
Many systematic reviews of interventions for autistic children and youth now appear in the research literature. Such reviews are useful in their focus on individual practices like functional communication training (Gregori et al.
2020), intervention for autistic children/youth of a certain age (Sandbank et al.
2020), or interventions occurring in certain locations such as schools (Martinez et al.
2016), and allow for more in depth review of contextual factors impacting the intervention or outcomes. They do not, however, provide a comprehensive critical summary of evidence across practices, ages, and outcomes. To date, only the National Professional Development Center on ASD (NPDC) and National Standards Project (NSP) have conducted comprehensive reviews of focused intervention practices for children and youth with autism and included both group and SCD studies.
The NSP published their comprehensive review in two phases. In Phase 1, their search process accessed articles from the early years of experimental intervention research for autistic children and youth from 1957 through September 2007 (National Autism Center,
2009). Peer-reviewed journal articles were included if the intervention/treatments were implemented in school, home, community, vocational, clinic settings and included children with autism who did not have significant co-occurring conditions. In Phase 2, the NSP investigators followed the same process as occurred in Phase 1 (National Autism Center,
2015) adding articles published from 2007 to 2012. Their analyses generated 14 practices for children and youth with autism that met their criteria for evidence-based.
The NPDC investigators also conducted two iterations of reviews of the intervention literature. The first review included articles published over the 10-year period from 1997 to 2007 (Odom et al.
2010a,
b) and used the research design quality indicator criteria established by the CEC-Division for Research (Gersten et al.
2005; Horner et al.
2005) to evaluate articles for inclusion or exclusion from the review. In the second review, the NPDC team (Wong et al.
2014,
2015) used a more comprehensive search strategy, extended the coverage of the literature to include 22 years of studies (1990–2011), revised their methodological review criteria and process to include current criteria established by What Works Clearinghouse. Using a standard review protocol, they trained a national panel of 159 reviewers to evaluate the methodological quality of the journal articles. From the articles meeting quality criteria, the NPDC team identified 27 EBPs in the second review.
The purpose of this study, now being conducted by the National Clearinghouse for Autism Evidence and Practice (NCAEP), was to update the Wong et al. (
2015) review, incorporating autism intervention literature from 2012 to the end of 2017. The questions addressed by this review were: What focused intervention practices are evidence-based? What outcomes areas did evidence-based focused intervention practices address? What are the characteristics of the research designs, participants, and intervention implementation?
Method
As noted, the previous review (Wong et al.
2015) included articles from 1990 to 2011. In the current review, the methods from the previous review were followed as closely as possible to access and incorporate acceptable articles published from 2012 to 2017. The two groups of articles were then combined for the re-analysis that resulted in the identification of evidence-based practices. The systematic review included five phases: search, screening, quality appraisal, data extraction and synthesis, following the PICO conceptual framework originated by Richardson et al. (
1995) and followed by the Cochrane Collaboration. The methods are described in the subsequent sections and more detail about each step of the process can be found in the freely available report on the NCAEP website (
https://ncaep.fpg.unc.edu/).
Search Process
The NCAEP research team and a research librarian from the University of North Carolina at Chapel Hill developed and refined the literature search plan. The databases utilized in the search were: Academic Search Premier, Cumulative Index to Nursing and Allied Health Literature (CINAHL), Excerpta Medica Database (EMBASE), Educational Resource Information Center (ERIC), PsycINFO, Social Work Abstracts, PubMed, Thomson Reuters (ISI) Web of Science, and Sociological Abstracts. Search terms were intentionally broad to be as inclusive as possible and included terms related to diagnosis (autism OR autistic OR Asperger OR ASD OR ASC [autism spectrum condition] OR pervasive developmental disorder OR PDD/PDD-NOS) and practice (intervention OR treatment OR practice OR strategy OR therapy OR program OR procedure OR approach OR methods OR education OR curriculum). The only filters used were peer-reviewed, language (English) and publication date (2012–2017). Deduplication was used to eliminate duplicate articles from the initial search prior to screening.
Screening Process
Prior to the title/abstract and full-text phases of the screening process, NCAEP team members participated in trainings to review the inclusion and exclusion criteria (see subsequent sections) and screening procedures. For the title/abstract screening, the NCAEP team reviewed the title and abstract of the article and indicated if the article should be excluded or further reviewed in a full-text screening. Following the title/abstract screening, the team gathered the full-text version of all articles that were not excluded. During the full-text screening, team members indicated if an article should be included or excluded, and for excluded articles indicated a reason for exclusion. Both steps were completed internally by single reviewers.
Inclusion/Exclusion Criteria for Studies in the Review
Articles included in this review were published in peer-reviewed, English language journals between 1990 and 2017. The study inclusion criteria are described in the subsequent sections and summarized in Table
1.
Table 1
Inclusion and exclusion criteria
Literature | Article published (or online prepublication) in peer-reviewed journal | Grey literature such as dissertations, conference presentations or proceedings |
Language | Article published in English | Article published in non-English journal |
Intervention | Intervention was focused intervention practice Intervention was behavioral, developmental, academic and/or vocational | Intervention was comprehensive treatment program Intervention was medical or psychopharmacological |
Outcomes | Outcomes were behavioral, developmental, academic, mental health or vocational for autistic children and youth | Outcomes were physical health, neuroimaging, or EEG Only outcomes for family or caregivers reported |
Study Design | Article examined efficacy of intervention with group or single case design | Article primarily descriptive or correlational Article tested moderation of effects on previously published or nonsignificant main effects |
Population/ Participants | Some participants identified as autistic Some participants between birth and 22 years of age | Outcomes for participants with autism/in specified age range were not presented separately |
Population/Participants
A study had to have participants whose ages were between birth and 22 years of age and were identified as having autism spectrum disorder (ASD), autism, Asperger syndrome, pervasive developmental disorder, pervasive developmental disorder-not otherwise specified, or high-functioning autism. Studies varied on the description of the autism diagnosis- some had ADOS or ADIR data, others reported clinical diagnoses by psychiatrists or physicians, while others used diagnoses provided by public school evaluations. Participants with autism who also had co-occurring conditions (e.g., intellectual disability, genetic syndrome, mental health conditions) were included in this review. Studies with participants identified as “at risk for autism” were not included in the review.
Interventions
The focused intervention practices examined in a study had to be behavioral, developmental, and/or educational in nature. Studies in which the independent variables were only medications or nutritional supplements/special diets (e.g., melatonin, gluten-casein free, vitamins) were excluded from the review. In addition, only interventions that could be practically implemented in typical educational, clinical, home, or community settings were included. As such, intervention practices requiring highly specialized materials, equipment, or locations unlikely to be available in most educational, clinic, community, or home settings were excluded (e.g., dolphin therapy, hippotherapy, hyperbaric chambers). Interventions requiring the supervision of trained medical personnel were excluded (e.g., chelation, neurofeedback, or acupuncture/acupressure).
Outcomes
Studies had to generate behavioral, developmental, academic, vocational, or mental health outcomes (i.e., dependent variables). Outcome data could be discrete behaviors (e.g., social initiations, stereotypies) assessed observationally, ratings of behavior or student performance (e.g., parent/teacher questionnaires), standardized assessments (e.g., nonverbal IQ tests, developmental assessments), and/or informal assessment of student academic performances (e.g., percentage of correct answers for instructional task). Studies only reporting physical health outcomes were excluded. Studies that targeted only caregiver and/or staff outcomes or only examined how those outcomes mediated student outcomes were excluded.
Study Designs
Studies included in the review had to employ a group design or SCD to test the efficacy of focused intervention practices. Adequate group designs included randomized control trials, sequential multiple assignment randomized trials, quasi- experimental designs, or regression discontinuity designs that compared an experimental/intervention group receiving the intervention to a control or comparison group that did not receive the intervention or received another intervention (Shadish et al.
2002). SCDs had to demonstrate the functional relationships between the intervention (or independent variable) and the autistic child/youth outcomes (Kazdin
2011). Acceptable SCDs for this review were withdrawal of treatment (ABAB), concurrent multiple baseline, multiple probe, alternating treatment, and changing criterion designs (Horner and Odom
2014), as well as SCDs that included hybrid designs (e.g., an ABAB set of phases in an multiple baseline design). Studies that were solely descriptive, examined only predictors, reviewed existing literature, or were meta-analyses were excluded. In addition, non-concurrent multiple baseline studies and component analyses without a baseline condition were also excluded.
Quality Appraisals
Protocols for reviewing group design and SCD studies used to determine methodological acceptability were developed in the Wong et al. (
2015) review and appear in the supplemental materials for this article. The protocols drew from the methodological quality indicators developed by Gersten and colleagues (
2005) for group design and Horner et al. (
2005) for SCD, as well as the review guidelines established by the WWC. In addition, the last item on each protocol asked reviewers to make a judgement about whether the study reported positive effects for the intervention. Protocols went through two iterations of pilot testing within the research group and then were reviewed by two national leaders in research methodology and intervention research, with expertise in SCD and group design, respectively for finalization. Only minor updates were made in the current review (e.g., include SMART design as a design option).
National Board of Reviewers
To assist in quality appraisals, external reviewers were recruited through professional organizations (e.g., Association for Behavior Analysis International, Council for Exceptional Children), professional contacts, social media, the NCAEP website, and solicitation to reviewers from the previous review. The criteria for qualifying for the subsequent training was that the individual had to have a graduate degree or be a current Ph.D. student, had to have finished coursework in experimental group design and/or single case design research, and had to have had coursework related to and/or experience working with individuals with autism. The reviewers self-identified their methodological expertise and interests as group, SCD, or both. Reviewers completed an online training process described fully in the project report (
http://go.unc.edu/Hk72T). Following training, they coded a “master-file” article (i.e., an article in which correct review answers had been established by our team) that employed the respective design. For the review of the master-file study, reviewers had to meet an 80% inter-rater agreement criterion for study elements. If potential reviewers did not meet the criteria, they were allowed to review the training materials and complete the task a second time (i.e., with a different master-file article for the article review).
Two hundred and twenty-one reviewers completed the training and met inter-rater agreement criteria with the master code files; 55% completed requirements for single case design articles (n = 122), 10% completed requirements for group design articles (n = 21), and 35% completed requirements for both design types (n = 78). Most reviewers received their degrees in the area of special education or applied behavior analysis and were faculty, graduate students, or practitioners. Reviewers were acknowledged in the report and BCBA/BCaBA reviewers received continuing education credit if requested.
Ten articles were randomly assigned to each reviewer, with the exception that a check was conducted after assignment to make sure that reviewers had not been assigned an article for which they were an author. They first completed a set of screening questions about the articles (e.g., type of study design) followed by the methodological review items for SCD or group design. If an article met all individual methodological items on the protocol, reviewers noted whether the study had positive effects for autistic participants on at least one outcome variable and listed the variables with positive effects. Last, reviewers described the key features of the study (e.g., participant characteristics) and the intervention procedures. Each article was independently reviewed by two external reviewers. Once both reviews for a given article were complete, the NCAEP team identified any disagreements between the reviewers related to methodological quality and effects. If needed, an NCAEP team member was assigned to complete a third review and make a final determination about quality and/or effects. A third review was required for 42% of the articles, including 27% that were reviewed for quality and effects and 15% that were reviewed for effects only.
Inter-rater Agreement
The NCAEP team collected inter-rater agreement on quality appraisals for 1,085 articles. The formula for inter-rater agreement was rating agreements divided by agreements plus disagreements multiplied by 100%. Agreement was calculated for (a) methodological quality review items on the review protocol, (b) summative evaluation of whether a study met or did not meet quality criteria, and (c) evaluation of whether or not studies that met quality criteria had positive effects for autistic participants on at least one outcome variable. Mean inter-rater agreement on the individual items for quality review was 85% (range = 55–97%) for group design articles and 93% (range = 87–97%) for SCD articles, generating a total mean item agreement of 90%. Mean inter-rater agreement for summary decisions about article inclusion was 65% for group design articles and 80% for SCD articles, generating a total agreement of 73%. This agreement percentage was lower because if a single reviewer rated even a single item as negative, it led to exclusion of the article from the study. As noted, when a disagreement between reviewers occurred, a third internal review was conducted by an NCAEP staff member, which led to a consensus decision about inclusion or exclusion of the article. Of the included articles, there was 86% agreement for group design articles and 74% agreement for SCD articles on the presence of positive effects, yielding a total agreement of 80% (i.e., between the respective reviewers who agreed on the previous inclusion decision). When disagreement occurred on the positive effects report, internal NCAEP staff conducted another review just focusing on the decision about positive effects to yield a consensus judgment.
The NCAEP Team followed a three-step data extraction process. First, team members compared external reviewers reports of participant characteristics (e.g., diagnosis, age) and outcomes (i.e., dependent variables), and made final determinations for this information. Second, team members thoroughly reviewed each article to identify primary interventions. In this identification step, the reviewer could: assign an article to one or more of the 27 practice categories identified in the previous review (Wong et al.
2014); assign the article to a practice category that had been identified as having some evidence in the previous review; and/or identify a new possible category of practices represented in the article. Third, different team members reviewed each article assigned to a given practice in the previous step to confirm that it fit within the practice description. During this step of data extraction, team members also identified manualized interventions that fit within a practice. Manualized interventions shared procedural features that were similar to other interventions in the category but had unique features that distinguished them as a salient model and had an identifiable title. For example, Social Stories™ is a trademarked intervention by Carol Gray and Garand (
1993), that fits within Social Narratives practice description, but is also distinct as a particular type of Social Narrative. Also, during the data extraction phase of the review, our team identified additional articles that were removed for not meeting eligibility requirements and/or quality standards, which may have been missed in the original quality review. These decisions were confirmed by a second team member. Then, the data were compiled for analysis and synthesis.
Synthesis and Identification of EBPs
When all included articles were assembled into categories, the team made a final determination about whether a practice met the level of evidence necessary to be classified as an EBP using criteria for evidence established by the previous NPDC team. The NPDC’s criteria were drawn from the work of Nathan and Gorman (
2007), Rogers and Vismara (
2008), Horner and colleagues (2005), and Gersten and colleagues (2005), as well as the earlier work by the APA Division 12 (Chambless and Hollon
1998). Its rationale is based on the necessity of having a sufficient number of empirical demonstrations of efficacy through high quality, peer-reviewed journal articles and replications of those demonstrations by independent research groups.
Different criteria were established for group and single case design evidence. To be identified as evidence-based, a category of practice had to contain (a) two high quality group design studies conducted by two different research groups, (b) five high quality single case design studies conducted by three different research groups and involving a total of 20 participants across studies, or (c) a combination of one high quality group design study and three high quality single case design studies with the combination being conducted by two independent research groups. Independence of research groups was defined as the research being located in different settings and the key constituent members of the authorship of published articles being different from other research groups.
Discussion
The current report updates and extends the work on evidence-based, focused intervention practices begun with an initial review of the literature from 1997 to 2007 (Odom et al.
2010a,
b) and extended through a second report that covered the literature from 1990 to 2011 (Wong et al.
2015); extending this systematic review through 2017 added 567 articles to the review. As the intervention literature has provided more empirical information and as practices have evolved, some of the classifications required reconceptualization and revision of previous definitions. In an active research area, knowledge does not stand still, and in fact identification of EBPs should be dynamic, reflecting the growth of knowledge across time (Biglan and Ogden
2019).
The five new practices in this review are Augmentative and Alternative Communication (AAC), Behavior Momentum Intervention, Direct Instruction, Music-Mediated Intervention, and Sensory Integration. It is important to note that Sensory Integration refers explicitly to the classical sensory integration model developed by Jean Ayres (
2005) and not to a variety of interventions that address sensory issues but have been found to be unsupported (Case-Smith et al.
2015; Watling and Hauer
2015). Also, several EBP categories from previous reviews were reclassified into other EBP categories (i.e., Scripting into Visual Supports, Peer-Mediated Intervention and Instruction and Structured Play Groups into Peer-Based Intervention and Instruction).
The growth of the published empirical literature led to a new classification of interventions noted above as MIMC. These interventions were developed by individual research groups and had sufficient evidence to meet the criteria as an EBP. However, the procedural features overlapped directly with established EBP categories. As such, MIMC could be considered “EBPs within broader EBP categories.” Ten interventions met the criteria for MIMC and have sufficient evidence to meet the EBP criteria. In addition, these interventions have clearly established manualized procedures or software, often including training protocols, which may better facilitate their uptake and implementation (Kasari and Smith
2013). Two interventions, PECS® (Frost and Bondy
2002) and PRT (Koegel and Koegel
2006; Stahmer et al.
2011), previously identified as EBP categories were reclassified under AAC and Naturalistic Intervention, respectively. The MIMC classification does not convey in any way that the interventions are less strong or efficacious than previously indicated, because they both have an extensive set of research that supports their efficacy. Also, some MIMCs have features that cross EBP categories. For example, Project ImPACT (Ingersoll and Dvortcsak
2019) was classified within the Parent-Implemented EBP but it also shares some characteristics with Naturalistic Intervention.
Because this review included articles for two time periods, it was possible to examine trends across time. In our analysis, SCD remained the primary methodology employed, constituting 85% of the acceptable studies. Multiple baseline and multiple probe designs were more frequently used in the more current set of articles as compared to the earlier review, perhaps because they do not require a treatment to be withdrawn in order to demonstrate an experimental effect of the intervention (i.e., as is necessary in ABAB designs). Although group designs were only 17% of the studies included overall, there was a dramatic increase in the number of RCT studies included between the review periods, which may reflect greater access to the number of participants with autism needed to establish power for RCT analyses and/or priorities of funding agencies.
Information on the intervention setting, implementer, and group size was available from the 2012–2017 review period. Commenting on the earlier research literature, Parsons and Kasari (
2013) lamented the fact that most intervention research was not occurring in the educational settings where many children and youth with autism spend a great part of their life. In the current review, nearly 50% of the research was conducted in education settings, the largest of any setting reported. While an important step in the right direction, the majority of the research was still being conducted in individual sessions by research staff members. Certainly, directions for the future would be to examine more often the efficacy of interventions when implemented in “authentic” educational settings by practitioners such as teachers, speech language pathologists, psychologists, and other service providers.
Outcomes for intervention participants shifted somewhat from the 1990–2011 to the 2012–2017 review period. As noted, communication, social, and behavior outcomes occurred most frequently across both review periods, as would be expected given that these are the challenges that define autism. There were notable increases in studies that targeted academic skills, vocational skills, and mental health, although the number of studies addressing these outcomes remains low relative to the number of studies. These outcomes are important, especially for adolescents and young adults with autism, and their low frequency document needed areas for research in the future.
For this review, outcomes were coded by domain and individual dependent variables were not coded. For focused intervention practices, researchers tend to address individual skills or behaviors. Also, for SCD studies, researchers employ dependent measures that can be used repeatedly across time to monitor changes in student performance (Kazdin
2011). A safe assumption is that for SCD studies, the outcome assessments employed observational or rating scale data. For group design studies, these methodologies may also be used, but researchers also would have more frequently employed standardized, norm-referenced measures. Describing specific assessment methodology of EBP studies would certainly be a feature of future research.
A number of demographic findings have implications for future research as well. As in the earlier review, the majority of the 2012–2017 studies were conducted with preschool and elementary-aged children with autism, indicating an important need for increased research with infants/toddlers, adolescents, and young adults with autism. Also, although it fits the overall demographic of autism, most studies were conducted with male participants, and information about differential effects for female participants with autism continues to be under-examined. Finally, there are gaps related to the reporting, inclusion, and analysis of participants with co-occurring conditions, which should be addressed in future studies.
In the 2012–2017 review period, information was collected about whether researchers reported the race/ethnicity/nationality of autistic participants, which did not happen in the previous review. However, West and colleagues (
2016) recoded the earlier set of articles to retrieve those data, which can serve as a point of comparison. West et al. found that only 17.9% of the articles from 1990 to 2011 reported race/ethnicity, while in the 2012–2017 review period, 30% of the reviewed articles reported these data. In both sets of studies, Black and Hispanic/Latino were the most frequent nonwhite racial/ethnic categories reported. The number of participants from nonwhite racial and ethnic groups in the subset of studies that reported these data was strikingly low compared to what would be expected based on community demographics. For example, only 8% of research participants were Hispanic/Latino, while in the United States alone, 26% of the school age population identifies as Hispanic (U.S. Department of Education
2017). Also, differential treatment outcomes related to race/ethnicity/nationality were not examined, which is similar to findings by Pierce et al. (
2014) in an analysis of studies published between 2000 and 2010. Last, socioeconomic class (SES) of participants is rarely described for autistic participants in research studies, so the possibility of determining how SES affects treatment outcomes is largely not possible.
Limitations
Several limitations exist for this review. As noted, the review included only studies published from 1990 to 2017. Two limitations exist regarding this timeframe. First, studies that occurred before 1990 were not included, although one might expect early (i.e., pre-1990) studies of important and effective practices to have been replicated in publications over subsequent years. Second, because of the time required to conduct a review of a very large database and involve a national self-selected set of reviewers, there was a lag between the end date for a literature review (i.e., 2017) and the date which the review is published. Certainly, studies have been published in the interim that could have impacted the EBP classifications.
Regarding the methodology of the review, this was clearly a systematic review of the literature and not a meta-analysis. So, the magnitude of effect size was not examined. Also, the review only contained peer-reviewed journal articles and not grey literature, as sometimes occurs in meta-analyses (McAuley et al.
2000). In addition, studies with null findings were not included. In fact, experimental studies are rarely set up to prospectively test for null findings, although there are methodological procedures for addressing such a question (Greene et al.
2007). Research studies with a hypothesis of treatment condition differences that instead “prove the null hypothesis” run the greater risk of Type II error. Last, although methodological quality indicators were drawn from recognized authoritative sources in the field, it is possible that a more detailed analysis of methodology than was practical for this review could have influenced the studies included in this review.
Implications and Conclusions
The current review, as noted previously, is an important link in the research to practice process. When translated into useable and accessible information for practitioners (Sam et al.
2020a,
b) and supported through professional development and implementation science (Sam et al.
2020a,
b), these practices can be essential components of effective programs for autistic children and youth. That is, practitioners may match EBPs to specific learning goals for autistic children and youth (Cox and Sam n.d.) analogous to the way medical practitioners match specific treatments to the health needs and characteristics of their patients in personalized medicine. By assembling multiple EBPs to address specific learning needs, practitioners can build a technical eclectic program for children and youth with autism (Lazarus and Beutler
1993). In such an approach, practitioners establish programs with strong program quality as a foundation, develop individualized and clearly articulated goals for children/youth, select and implement practices that may have different theoretical bases but also have demonstrated efficacy (Odom et al.
2012). In a recent study, Sam et al. (
2020a,
b) employed such a program in 59 elementary schools and found significantly positive effects for program quality, teachers use of EBP with fidelity, and autistic children goal attainment.
In conclusion, the current study provided an updated review of the empirical evidence supporting focused intervention practices. In this review, nearly 1,000 studies were identified, with over half being published between 2012 and 2017. The updated review led to a revised set of 28 primary EBPs and also 10 interventions that classified as MIMCs. Examination of trends from the earlier review and current update suggest that SCD studies continue to be the modal form of research although RCT are being used more often. This analysis suggests that important directions for future research include intervention effects related to race, ethnicity, and gender as well as increased research for both infants/toddlers and adolescents/young adults with autism.
Acknowledgments
Support for this project was provided by the Institute of Education Sciences, U.S. Department of Education through Grant R324B160038 (Postdoctoral Training Program on Special Education Research) awarded to University of North Carolina at Chapel Hill and the U.S. National Institutes of Health, Grant T32HD040127. The opinions expressed represent those of the authors and do not represent the U.S. Department of Education or the U.S. National Institutes of Health. Funding for this work was also provided by the Ireland Foundation, Mr. John E. Rucker, and the Frank Porter Graham Child Development Institute. The authors wish to acknowledge the support of the following individuals at the University of North Carolina and the Frank Porter Graham Child Development Institute who provided assistance, feedback, and technical support during the course of the project: Amelia Gibson, Kathleen Thomas, Ann Sam, Victoria Waters, Jeff Alpi, Andrea Ross, Stephanie Ridley, Luke Hayek, Lindsay Rentschler, Tara Regan, Coral Morrow, Crisma Emmanuel, Benjamin Carter, Juliet Alegria, Mary Tran, and Ashley Freuler.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.