Focus on implementation: Parent-mediated early intervention for young children with autism spectrum disorders (ASD)


  • Jacqueline Roberts,

    Corresponding author
    1. Autism Centre of Excellence, Griffith University, Mount Gravatt, Queensland, Australia
    • Jacqueline Roberts, Autism Centre of Excellence, Griffith University, Mount Gravatt, Queensland, Australia.

      Cheryl Dissanayake, Olga Tennison Autism Research Centre, La Trobe University, Melbourne, Victoria, Australia.

    Search for more papers by this author
  • Cheryl Dissanayake

    Corresponding author
    1. Olga Tennison Autism Research Centre, La Trobe University, Melbourne, Victoria, Australia
    • Jacqueline Roberts, Autism Centre of Excellence, Griffith University, Mount Gravatt, Queensland, Australia.

      Cheryl Dissanayake, Olga Tennison Autism Research Centre, La Trobe University, Melbourne, Victoria, Australia.

    Search for more papers by this author


The commentaries in this section refer to the review, Oono IP, Honey EJ, McConachie H. Parent-mediated early intervention for young children with autism spectrum disorders (ASD). Cochrane Database of Systematic Reviews 2013, Issue 4. Art. No.: CD009774. DOI: 10.1002/14651858.CD009774.pub2.

Editor's note: Parents of children with autism often have a complicated life. In some cases, because of their child's behavior, they cannot take them to a restaurant or to a movie, let alone on a vacation. Finding child-care may be impossible. Some children with autism appear to be unhappy much of the time. Most parents would fly to the moon if that would make their child better, but it is our responsibility as clinicians to only assign parents tasks that are likely to benefit their child. In their commentaries, Professors Jacqueline Roberts and Cheryl Dissanayake provide their opinions on the fine art of balancing what we ask parents of autistic children to do.

Commentary by Jacqueline Roberts

A steady rise in the number of children diagnosed with autism at increasingly younger ages is likely to be related to an increased focus on, and interest in, parent-mediated interventions for children with autism. Fundamentally parent-mediated intervention makes sense for very young children. Parents are the primary source of input and learning, and parent–child interaction is critical for early social communication development. Early child–caregiver interaction is the primary means by which children develop all aspects of interpersonal communication and learn language. In the case of most children with autism, the parents are baffled by their child's early lack of interest in interaction, lack of engagement and joint attention. They know their child is different well before the age children usually start to talk. When their child's language development is delayed and disordered their concerns intensify.

Parent-mediated intervention potentially provides parents with the knowledge and skills to persist with engaging their children with autism at a young age and facilitate the development of social communication skills including language.

Advantages of parent-mediated intervention

Ideally, all interventions should involve a consistent collaborative approach so that the same strategies are being implemented in natural settings at home and in the community. Strategies will be most effective and meaningful for the child when incorporated into daily routines, and parents are best placed to do this.

Children learn best if engaged and motivated, and parents are in the best position to take advantage of opportunities across the child's day. Parents also know their children intimately and understand what is motivating and reinforcing for their children.

In addition, learning that occurs in natural contexts does not need to be generalized and is likely to be intrinsically reinforcing and hence maintained.

For many parents there will be benefits in having more control and the feeling that they are proactive.


Conversely some parents may find the expectation that they need to provide intervention for their child stressful, which puts additional pressure on parents at a time when they are already stressed.

Parents may be on the autism spectrum themselves and may find social engagement challenging as a result.

Parents may not have the resources to undertake the necessary training and allocate the time and energy required to implement parent-mediated intervention. They may feel guilty as a result.


The findings of the review suggest that the evidence for parent-mediated interventions is not strong, effect sizes are small and findings are inconsistent across studies. This is likely to be due in part to the heterogeneity of children with autism, heterogeneity of families and different types of intervention implemented in the studies reviewed, ranging from approaches based on traditional applied behaviour analysis (1) to approaches focused on engagement and relationship development (2).

Parents vary in the style of intervention they prefer. Some like the highly structured routines of ABA, while others prefer the relationship development focus of transactional developmental approaches. Therefore we need to match process and content to the parent as well as to the child.

Encouragingly, an increase in parent–child engagement was the best reported finding of the research into parent-mediated interventions. However, there was no significant evidence that the interventions reduced parent stress.

Research challenges

It is difficult to evaluate quality and quantity of parent-mediated intervention. For the intervention delivered to the child, checking of fidelity is difficult, and measuring the intensity of interventions that are built into daily routines is challenging.

There is also the challenge of measuring outcomes. Standardized measures may be blunt instruments insensitive to change, particularly changes in parent–child synchrony.

It is difficult to tease out the effect of maturation, to understand how much children would have improved anyway. If they are only just keeping up with their typical peers developmentally or even if they are falling behind it is possible, but difficult to prove, that they would be even further behind without the parent-mediated intervention.

Research directions

We need research focused on individual tailoring of intervention to individual children and their families and research designed to consolidate knowledge of prognostic indicators for different subgroups on the autism spectrum. The development of tools and processes to reliably describe and measure social communication development should be a priority.

Commentary by Cheryl Dissanayake

Parents are critical in early childhood development, and partnering with them in the delivery of intervention needs to be a core component of any early intervention programme for children with autism spectrum disorders. Few would argue otherwise, especially as interventions are increasingly begun earlier in development given the trend towards reduced ages of diagnoses. Indeed, parents are likely to be one of the ‘essential active ingredients’ in a successful intervention programme for children with autism spectrum disorders. In addition to providing consistency for their young child in the home and across different settings, empowering parents to engage with their child in a meaningful way has the added benefit of increasing their sense of efficacy.

However, the question of parent-delivered interventions, without the support of direct delivery to the child by an expert therapist(s), is another question. In its review of parent-mediated interventions, the Cochrane Review has found increases in parental synchrony (while interacting with their child) as one of the main outcomes, which may well be associated with the reported increase in child's language skills and reduced autism spectrum disorders' severity post-intervention. Few other gains in child outcomes (e.g. IQ) are found, particularly when these outcomes are assessed independently of the parents. Indeed, it is of interest that the smallest effect sizes are evident when child outcomes are assessed independently. Parents, on the other hand, report benefits for their child, with large effect sizes, although they do not report reductions in their own levels of stress. Self-reporting is, of course, hampered by possible bias. Nonetheless, there is evidence here that parent-delivered interventions do confer benefits for the child.

One wonders, however, if the amount of benefit is underestimated by our reliance on blunt tools, the majority of which are standardized, used to measure outcomes in cognition and behaviour. It is of interest that one of the strongest findings in the review is based on a fine-grained observational measure of parent–child synchrony. The authors too raise this point regarding accurate and sensitive measurement of outcomes that is not hampered by reporting bias or by instruments that are not sensitive to change.

Another issue is ascertainment of any dilution of effect. The parents are trained by the ‘expert’ therapist, and then administer the intervention to their child. It remains an empirical question whether the parental report of changes is stronger than changes in the children themselves (as measured independently) precisely because of the distance between the ‘expert’ and the target of this expertise: the child. Intermingled here is the issue of fidelity in the delivery of the intervention, another issue not able to be addressed in the review. However, each of these issues deserves attention in their own right. Studies are needed to disentangle these variables of ‘distance’ and ‘fidelity’ so that we may build a strong evidence base about how best to deliver efficacious interventions for young children with autism spectrum disorders.