Elsevier

Research in Autism Spectrum Disorders

Volume 4, Issue 4, October–December 2010, Pages 577-593
Research in Autism Spectrum Disorders

Review
A meta-analytic review of the effectiveness of behavioural early intervention programs for children with Autistic Spectrum Disorders

https://doi.org/10.1016/j.rasd.2010.01.014Get rights and content

Abstract

The effectiveness of behavioural intervention programs for children with Autistic Spectrum Disorders was addressed by a meta-analysis, which reviewed 14 studies. The findings suggest that the behavioural programs are effective in improving several developmental aspects in the children, in terms of their treatment gains, and also relative to eclectic-control programs in the same studies. Factors that were found to be correlated with the effectiveness of the behavioural programs were the intensity and the duration of the programs, the parental training, as well as the age and the adaptive behaviour abilities of the children at intake. The review showed that the high intensity of the programs is correlated with better gains in some developmental domains. Moreover, the high adaptive behaviour abilities and the young age of the children are also related with better progress the children have in some specific domains.

Introduction

Autistic Spectrum Disorders (ASD) were historically viewed as lifelong developmental disabilities, characterised by intellectual, communication, social, emotional, and adaptive behavioural deficits (APA, 1994, WHO, 1992). Many intervention approaches have been suggested to help children with ASD (e.g., behavioural therapy, Son-Rise Program, TEACCH, PECS, holding therapy, diets, secretin, sensory integration therapy, music therapy, speech therapy, homeopathy, etc.). Despite the considerable number of intervention programs, the majority of the literature relating to these programs remains at the level of description, and there is little evidence provided for the effectiveness of many of these approaches. One exception to this generalization concerns treatments based on the applied behaviour analysis (ABA) approach, which are the most thoroughly evaluated, and among the best known interventions for ASD.

Research on behavioural treatments for children with ASD started around 40 years ago (see Larsson, 2007). However, the research paper that initiated most recent debates in this area was that reported by Lovaas (1987). The outcomes reported by Lovaas (1987) were remarkable, but at the same time there were significant methodological flaws with the study. Those methodological flaws provoked a strong debate, and a further set of studies concerning the effectiveness of behavioural EIPs, which attempted to address some of the issues arising from the perceived flaws in the original Lovaas (1987) study (e.g., Anderson et al., 1987, Bibby et al., 2002, Birnbrauer and Leach, 1993, Smith et al., 1997). The findings obtained from these studies, on the whole, were promising, and most noted that behavioural EIPs were effective as a treatment approach for ASD, in that they improved the overall functioning of many children with ASD. Nevertheless, these findings have been challenged in terms of the magnitude of gains, and criticized for a further range of methodological weaknesses. For example, some of these studies (e.g., Howard et al., 2005, Sallows and Graupner, 2005) showed significant improvements in the developmental trajectory of children with ASD, while some other studies found less improvement (e.g., Bibby et al., 2002, Smith et al., 1997). Some of these studies found significant differences between behavioural EIPs and control groups (e.g., Howard et al., 2005, Sheinkopf and Siegel, 1998), while some did not (e.g., Farrell, Trigonaki, & Webster, 2005). The majority of the studies contain some methodological weaknesses (see Eikeseth, 2001), such as small sample sizes (e.g., Hoyson et al., 1984, Smith et al., 2000a), no comparison group (e.g., Bibby et al., 2002, Harris and Handleman, 2000), no random assignment of the children in the groups (e.g., Farrell et al., 2005, Smith et al., 1997), no matched groups (e.g., Eldevik et al., 2006, Fenske et al., 1985), and different measures used for assessment between and within children (e.g., Howard et al., 2005, Sheinkopf and Siegel, 1998).

Despite the range of individual studies, a definitive answer regarding the effectiveness of behavioural EIPs in general has not yet been given. Further analysis of this question, beyond the discussion of individual reports, remains a necessity for various reasons. The demonstration of general effectiveness that cannot come from an individual study could have implications in cases, such as that of the Republic of Ireland, where the funding of ABA schools is in danger, because of a perceived lack of evidence regarding the general, as opposed to specific, effectiveness of this approach. Additionally, for many commissioning authorities, the behavioural EIPs have a relatively high operational cost per person, at least initially (although there is a debate relative to the long term costs of not providing treatment; Jacobson and Mulick, 2000, Jacobson et al., 1998, Marcus et al., 2000, Papps and Dyson, 2004). Notwithstanding this financial issue, the use of this kind of EIP for children with ASD has increased (Bibby et al., 2002, Jacobson et al., 1998). Consequently, both parents, and funding bodies, need reliable answers about the effectiveness of those programs and some guidance about the conditions under which such programs might be effective.

One way to study the literature on the field is meta-analysis. Meta-analysis is only one of many ways to summarise, integrate and interpret selected sets of studies. It is a research tool, developed at the end of the 1970s (Lipsey & Wilson, 2001), which translates the results from different studies to a common metric, and statistically explores the relations between the study characteristics, and their findings (Mullen, 1989, O’Mara et al., 2005).

When the present meta-analysis was initiated, there were no meta-analytic reviews published in the field of behavioural early intervention programs for children with ASD. However, recently, three meta-analyses (Eldevik et al., 2009, Reichow and Wolery, 2009, Spreckley and Boyd, 2009) have appeared in peer-reviewed journals, which, however, are characterised by significant differences in their design, inclusion criteria, and included studies and findings. Spreckley and Boyd (2009) carried out a review to the effectiveness of applied behaviour intervention programs for preschool children with ASD, in their cognitive, language, and adaptive behaviour abilities. This meta-analysis involved only four studies, in which ABA programs were compared with standard care. The analysis suggested that there is not adequate evidence for the superiority of ABA programs’ outcomes to those of standard care, but the study had several limitations such as high variability in the included studies, and poor homogeneity (Spreckley & Boyd, 2009).

In a more inclusive meta-analysis with different inclusion criteria, Reichow and Wolery (2009) carried out a systematic review, which used meta-analytic techniques and was aiming mainly to provide a quantitative analysis of the studies on EIBI. That meta-analysis included 13 studies, which varied in their experimental design and quality. It provided a descriptive analysis of the studies, an effect sizes (ESs) analysis (both comparing baseline follow-up and experimental-control group data), while meta-analytic techniques were only used for the within-group changes in IQ and not for the other developmental aspects or the between group differences. Regarding the ESs analysis, there is a concern about the use of the same formula for both the calculation of developmental changes (baseline follow-up) and the differences between experimental and control programs. The findings of this study, contrary to the Spreckley's and Boyd's (2009), indicated that EIBI programs are effective, on average for children with ASD.

Lastly, Eldevik et al. (2009), in a replication and extension of the Reichow and Wolery (2009) meta-analysis, with focus on methodological improvement, employed a new meta-analysis, with stricter inclusion criteria (e.g., more precise definition of EIBA, more uniformity in outcomes, use of the raw data and inclusion of interrater reliability). That meta-analysis included nine studies and focused mainly on effect sizes based on differences between experimental and control groups, for IQ and adaptive behaviour abilities. The results from this analysis also supported the effectiveness of EIBI. Thus, aforementioned three meta-analyses constituted an important step towards the study of the effectiveness of the EIBI. However, many questions remain to be answered, such as the effectiveness of EIBI in improving other developmental aspects of children with ASD, as well as about the factors which are related to the effectiveness.

Section snippets

Purpose of the meta-analysis limitation

The present meta-analysis could be considered an extension to the previous meta-analyses about the effectiveness of the behavioural EIPs for children with ASD, which also tried to address some constraints of them. It included more studies than Spreckley's and Boyd's (2009) study, more developmental dimensions and measures of the effectiveness of the intervention programs than Eldevik's et al. (2009), and used the suggested formula by Lipsey and Wilson (2001) for standardised mean change ESs,

Selection of studies

A major threat to the validity of meta-analytic studies is the selection of studies, which is often characterised as ‘biased’, if it is based only on published studies (Lipsey and Wilson, 2001, Torgerson, 2006). This problem is colloquially known as the ‘file drawer’ problem, and refers to those unpublished studies that might exist, and whose results may fail to support the pattern established by published findings (Mullen, 1989). A very common technique that is used in meta-analysis in order

Results and discussion

The results from the meta-analysis are reported in two sections. The first section has some descriptive statistical elements which regard the variables which were found to be correlated with the effectiveness of the behavioural EIPs, and were included in the meta-analysis. Apart from those variables that will be presented, several other variables were assessed, but either no correlation was found with the progress of the children, or their frequency was small, and that is why they are not be

Conclusions

This meta-analysis represents a synthesis of 14 studies on the effectiveness of behavioural EIPs for children with ASD. The purpose of the analysis was to evaluate the magnitude of the effectiveness of the behavioural EIPs, comparing both pre–post-treatment performances, and behavioural-control outcomes, as well as to study the impact of various different factors on these outcomes. All the analysis carried out while controlling for the effect of the methodological quality of the studies.

The

References (60)

  • Autism Society of America. (2007, July 26). Stress on Families. Retrieved July 26, 2006 from...
  • K. Bassett et al.

    Autism and Lovaas treatment: A systematic review of effectiveness evidence (BCOHTA 00:1T)

    (2000)
  • Beadle-Brown, J., Dorey, H., & Murphy, G. (2004). Early intervention for autism study: Summary of findings (University...
  • B.J. Becker

    Synthesizing standardized mean-change measures

    British Journal of Mathematical and Statistical Psychology

    (1988)
  • J.S. Birnbrauer et al.

    The Murdoch early intervention program after 2 years

    Behaviour Change

    (1993)
  • R.D. Boyd et al.

    Outcome survey of early intensive behavioral intervention for young children with autism in a community setting

    Autism

    (2001)
  • H. Cohen et al.

    Early intensive behavioural treatment: Replication of the UCLA model in a community setting

    Developmental Behavioral Pediatrics

    (2006)
  • J. Cohen

    Statistical power analysis for the behavioural sciences

    (1988)
  • S. Eikeseth

    Recent critiques of the UCLA young autism project

    Behavioral Interventions

    (2001)
  • S. Eldevik et al.

    Effects of low-intensity behavioural treatment for children with autism and mental retardation

    Journal of Autism and Developmental Disorders

    (2006)
  • S. Eldevik et al.

    Meta-analysis of early intensive behavioral intervention for children with autism

    Journal of Clinical Child and, Adolescent Psychology

    (2009)
  • P. Farrell et al.

    An exploratory evaluation of two early intervention programmes for young children with autism

    Educational and Child Psychology

    (2005)
  • G. Green

    Early behavioral intervention for autism: What does research tell us?

  • H. Guilliksen

    The increasing importance of mathematics in psychological research (part 3)

    The Score

    (1986)
  • S.L. Harris et al.

    Age and IQ at intake as predictors of placement for young children with autism: a four- to six-year follow-up

    Journal of Autism and Developmental Disorders

    (2000)
  • S.L. Harris et al.

    Changes in cognitive and language functioning of preschool children with autism

    Journal of Autism and Developmental Disabilities

    (1991)
  • R.P. Hastings et al.

    Stress in UK families conducting intensive home-based behavioural intervention for their young child with autism

    Journal of Autism and Developmental Disorders

    (2001)
  • L.V. Hedges et al.

    Statistical methods for meta-analysis

    (1985)
  • M. Hoyson et al.

    Individualized group instruction of normally developing and autistic-like children: A description and evaluation of the LEAP curriculum model

    Journal of the Division for Early Childhood

    (1984)
  • B. Ingersoll et al.

    Including parent training in the early childhood special education curriculum for children with autism spectrum disorders

    Topics in Early Childhood Special Education

    (2006)
  • Cited by (205)

    View all citing articles on Scopus
    View full text