Evidence Based Treatments for Child and Adolescent Mental Health Disorders #1 – Early Intervention In Autism

In a previous “Special Issue Spotter”, I linked to the January edition of the Journal of Clinical Child and Adolescent Psychology which published a 10-year review on evidence-based psychosocial treatments for children and adolescents. The edition is made up of a series of 10 review articles in which experts sift through the last 10 years of research in specific areas to identify best practice programs/ interventions. Over the next few months, I’d like to visit each of these 10 articles and draw out some of the key findings. In this post, I tackle the article by Sally J. Rogers and Laurie A. Vismara entitled “Evidence-Based Comprehensive Treatments for Early Autism”.

It seems the last 10 years have been much better for the early autism literature than the previous. As the authors note, when the original review was conducted in 1998, there were no randomized controlled trials of early autism interventions. In the current paper, 5 trials are included. In total, Rogers & Vismara review a total of 22 early autism interventions studies ranging in quality, scope, method, target group, intervention type and not surprisingly outcome. Whilst there is considerable variation between the studies, it is possible to summarize some of the characteristics of the studies.

Study Characteristics

Age of children: the age of the children involved in the studies ranges from 18 months to 72 months (6 years) with the majority falling in the 24 to 40 month age bracket (2 and 3 year old kids).

Diagnosis: Not surprisingly, the majority of the children meet criteria for Autism, Autism Spectrum Disorder or Pervasive Developmental Disorder Not Otherwise Specified.

Intervention Type: The range of interventions is probably the hardest to summarize, given a) the number of points of difference and b) that review articles are unable to provide extensive detail on the specific interventions. 16 of the 22 studies had manualized treatments, with only 2 studies looking specifically at “eclectic” treatments. Thus most interventions were based on a specific set of procedures. All interventions had behavioural components, focused specifically on shaping behaviours typically deficient in children with Autism (think language, social reciprocity, play, joint attention, self-help, recreational & academic skills). The length of interventions ranged from 12 weeks to 3 years, with intensity ranging from a couple hours per week to 40 hours per week direct intervention. Interventions were delivered by a mix of trained parents, trained workers, specialist early intervention consultants and trainee early intervention workers and delivered across a range of settings (childcare, home, clinic). Some interventions focused on the children themselves (e.g., direct training) whilst others others focused on upskilling the various agents in the child’s life (e.g., parent, childcare worker).

What treament(s) stand out?

Getting recognition as a “well established” treatment in the research literature is a complex task. Without going into extensive detail (Chambless & Hollon, 1998; Nathan & Gorman, 2002 are good starting points), a treatment must be well manualized and defined, so others can replicate the therapy in a different setting; the studies that have evaluated the treatment need to be of high quality; and the outcomes of the studies must show the treatment to be superior to no treatment (or best available treatment if going up against already well established treatments). In the early intervention for Autism field, only one treatment currently meets these criteria (Lovaas). That being said, many of the treatment programs for early intervention in Autism are based on similar theory (i.e., Applied Behaviour Analysis – ABA) and it may simply be a matter of time before they have been sufficiently well evaluated and we have a larger range of established treatment programs. In terms of treatment characteristics, Rogers & Vismara’s review does suggest that programmatic treatments with specific methods built on core theory are better than “eclectic” treatments drawing upon multiple factors. There is also the suggestion (as I read it) that clinic-based, focused daily interventions over long time periods (> 1 year) lead to greater effects thans shorter interventions, although shorter trials do show effects and it may point to a common finding in psychotherapeutic interventions that early response to treatment is a predictor of overall response.

Lovaas Method: The Lovaas method (described on their website), is an time-intensive (5-7 days per week, 2-8 hours per day for up to 3 years), individualized behaviour program for children aged 2-8 that focuses on shaping specific behaviours, particularly around language, social interaction, ability to learn as well as the reduction of inappropriate or destructive behaviours. Delivered by trained professionals, typically in a clinic-based environment, the treatment is intensive, gradual (shaping behaviours through successive approximations) and involves parents in addition to the instructor. It is a type of Discrete Trial Training (DTT) program in which the instructor leads the way in shaping specific behaviours in the child.  Results from the original Lovaas study (1987) showed a 47% “recovery” rate, namely these children attained a normal IQ and tested within the normal range on adaptive and social skills. Subsequent replication studies (Smith, Groen, and Wynn, 2000; Sallows and Graupner, 2005) have demonstrated positive outcomes also. Whilst both studies demonstrated significant improvements in IQ as a result of the intervention, with Sallows & Graupner showing a similar “recovery rate” to the original Lovaas study, there are multiple disclaimers. First, Smith, Groen & Wynn’s subjects did not show the same degree of IQ change, did not improve on measures of adaptive or social functioning and the gains were limited to those subjects with a PDDNOS diagnosis (i.e., not full autism). Sallows & Graupner’s subjects were provided with a augmented version of Lovaas program (involving components of Pivotal Response Training) and a significant portion (12 of 23) showed no improvement even after 4 years of therapy.

Pivotal Response Training: Whilst not getting a “well established” treatment guernsey, Pivotal Response Training (PRT) does get a special mention as a “probably efficacious” treatment. Information from the PRT website (http://psy3.ucsd.edu/~autism/prttraining.html) describes PRT as a naturalistic treatment approach, built upon the same applied behaviour analysis (ABA) theory as Lovaas but employing a different method of learning.  Unlike DTT where the instructor shapes skills according to a specific curriculum, PRT reinforces naturally occuring behaviours demonstrating “motivation” and “responsivity to cues”, particularly those around language, social interaction and play.

What improvements can parents expect for their children from these types of treatments?

Current evidence suggests children involved in these types of programs may demonstrate accelerated intelligence & developmental gains (i.e., age appropriate performance), improvements in use of language/communication and adaptive behaviour (i.e., behaviours that make up our daily life – toileting, dressing, feeding etc) and significant reductions in problematic behaviours (e.g., aggression). In best case scenarios, current evidence does suggest the possibility of “recovery” whereby the child is able to return to standard school placements and demonstrate adaptive behaviour at the level of a “normal” child. This should be interpreted with extreme caution though, as Rogers & Vismara point out, the “best” studies in the area have not demonstrated this “recovery” effect and hope for recovery comes from earlier, less rigorous trials and the recovery rate appears, at this stage, to be capped at 50% (i.e., 1 in 2 children may show this level of improvement). Aside from the direct effects on the child, it should also be noted that parents and childcare workers involved in many of these studies reported considerably improved knowledge about autism and how to respond to behaviours more appropriately.

Caveats on these improvements – what you should also know

When a body of research gets sufficiently large enough, researchers can start looking at what are called “predictors of treatment outcome”. These are characteristics of the treatment or the clients that predict how well the clients will respond to the treatment. Whilst still a relatively young research field, Rogers & Vismara’s article does suggest a number of potential predictors that professionals and parents should be aware of. Specifically, children who are likely to do better in treatment are those:

• Who have a less severe initial presentation – for example, better response to treatment is often found in those children not meeting full autism criteria (pervasive developmental disorder not otherwise specified PDDNOS)

• Whose autism is identified earlier – the earlier the intervention the better

• Whose initial IQ (intelligence) is higher

• Who receive longer treatments – although interestingly this does not yet seem to be related to actual therapy time, but the period of time over which therapy is conducted

• Whose treatment also includes specific efforts to integrate it into the family’s life

• Who show less dysmorphology (unusual physical features)

• Who are more socially initially (i.e., show more spontaneous social initiation) – although it has also been found that those children with greater social avoidance initially show the greatest improvement over time, so this one is a little contradictory.

Primarily, it appears as though the more functional the child is pre-treatment, the better they respond to treatment overall. From reading the review, it appears as a sad fact that many of the children with severe initial presentations (low IQ, communication and social interaction particularly) do not respond as favourably to the treatments currently available.

Evaluating Programs (advice for professionals and parents)

Rogers & Vismara provide an excellent set of suggestions for parents and professionals looking to access programs for children. Specifically, they suggest that professionals and parents should look for programs that:

a) use positive behaviour management & functional behaviour analysis to address unwanted or challenging behaviours

b) provide regular daily opportunities for spontanous communication

c) capitalize on naturalistic teaching methods involving child’s preferences to increase motivation, and generalization

d) seek to provide intervention across multiple settings, specifically home, special classrooms & inclusive settings

e) provide opportunities for peer interactions

f) push to generalize gains

g) involve parents and family members

h) regularly monitor progress

i) co-ordinate care across different settings

j) address other factors, not just the progress of the child (e.g., family health)

Final Thought

Rogers & Vismara’s article is fundamentally a snapshot of the current state of affairs with regards to early intervention treatments in Autism. As such it does not enter into the debate/discussion that I have seen on the web around attitudes towards autism and whether we have become too focused on identifying the problems associated with autism and not recognizing the unique world view that people with autism and autism spectrum disorders bring to our society. In doing this blog post, I have purposefully come from the angle of simply trying to extract from Rogers and Vismara’s article, some of the key findings relating to treatment, so that those searching for this kind of information are aware that these reviews are being written. I am happy to take comments from anyone though who wants to address the issue of treating autism as I am particularly open minded on the issue myself.

Sally Rogers, Laurie Vismara (2008). Evidence-Based Comprehensive Treatments for Early Autism Journal of Clinical Child & Adolescent Psychology, 37 (1), 8-38 DOI: 10.1080/15374410701817808

Special Issue Spotter: Evidence-based psychosocial treatments for children and adolescents: ten year review

Just noticed that the Journal of Clinical Child and Adolescent Psychology’s January 2008 issue was a special edition. In this edition, they provide a ten-year update on evidence based psychosocial treatments for a range of child and adolescent presentations including: anxiety, autism, disruptive behaviour, eating disorder, problems in ethnic minorities, OCD, substance abuse and trauma. A great opportunity to brush up on what is best practice in these areas! You will need subscription to the journal or access through institution to get hold of the articles.

Use of weighted vests with children with autistic spectrum disorders and other disabilities

Stephenson & Carter (2008), in their Journal of Autism and Developmental Disorders article, explore the current literature on the use of weighted vests in treatment for children with autistic spectrum disorders and other disabilities. Despite what seems like widespread use of this type of sensory integration therapy (SI), the literature so far suggests little to no effect.

Use of SI
Stephenson & Carter (based on previous literature) report that around 1/3 of children with autism are receiving some type of SI, with around 12% currently using weighted vests. One quarter of children with autism had used weighted vests at some point in time. Surveys with Occupational Therapists (OTs) have shown high acceptance of the procedure, with over 50% using the vests for people with autism/
PDD, ADHD, developmental delay, cerebral palsy and Down syndrome, mostly pre-school and young elementary school students.

SI is based on the proposition that many of the behaviours (e.g., steretypy, inattention, hyperactivity) observed in children with the conditions described above, relate to a failure to adequately process sensory information. SI provides corrective sensory conditions (e.g., brushing and rubbing of the body, deep pressure and compression of joints, scooter boards) which alter underlying sensory processing. Match the SI to the deficit and improvements are noted in those behaviours triggered by the sensory deficit. Weighted vests specifically are hypothesized to provide deep pressure stimulation which has a calming and organizing effect.

Weighted Vest Literature
Stephenson & Carter found only 7 studies on weighted vests (5 peer-reviewed papers, 1 non peer reviewed, 1 poster presentation) totaling 20 students. Nine had autism, 5 had pervasive developmental disorders (PDD), 1 had an autism spectrum disorder, 1 had developmental delay with autistic like behaviours, 2 had ADHD and 2 had ADHD and speech/language impairments. The ages of the children ranged from 2 years, 7 months to 11 years.

Four of the studies concluded that weighted vests were ineffective, 2 claimed positive effects and 1 claimed mixed results. Stephenson & Carter point to a number of methodological issues, particularly in the two “positive” studies that cast doubt on the validity of the findings. The issues are pertinent to those conducting small n designs.

1) Descriptions of participants were poor, including details of their presentations and how diagnoses were made.

2) The use of standardized measures of cognitive or developmental performance was apparent in only 2 of the studies, suggesting (along with diagnostic procedure ambiguity) a lack of attention to assessment.

3) Whilst studies utilized observers to rate child’s behaviour with and without the vest, there was minimal inter-rater reliability data for intervention periods (i.e., with vest), most observers were not blinded to the child’s vest status, and observation periods were generally short.

4) Interpretation of changes observed in one of the “positive” studies was based on a statistical test which was inappropriate for determining clinically significant change in small n designs. Most changes observed were clinically very small.

5) Studies varied considerably in the actual nature of the intervention, particularly the weight of vest and the time spent wearing the vest. As such, none of the studies helped identify what was an appropriate level of intervention for the various presenting conditions.

Take home message
This article highlights not only the current paucity of methodologically sound literature on SI, but also identifies a number of pitfalls for those conducting small n designs. Small n designs can be extremely useful in providing detailed descriptions of interventions. Good quality texts on small n designs are available and the methodologies still sound, despite being in an era of multi-centre randomized controlled trials (out of the reach of many of us small scale researchers!).

I hope to see more SI articles and reviews land in child and adolescent journals. My clinical experience has been that I have learned much from OTs and their methods. I hope also that some OTs read this review and perhaps make some comments on the status of the literature from their perspective.

Stephenson, J., Carter, M. (2008). The Use of Weighted Vests with Children with Autism Spectrum Disorders and Other Disabilities. Journal of Autism and Developmental Disorders DOI: 10.1007/s10803-008-0605-3