Abstract | Summary | Original Article

Efficacy of behavioural interventions for reducing problem behaviour in persons with autism: A quantitative synthesis of single-subject research.

Campbell, JM

Research in Developmental Disabilities 24: 120-138. 2003.

Question: How effective are different behavioural interventions for children presenting with challenging behaviour?

Background: Children with autism vary widely in their symptoms. When children present with challenging and aggressive behaviours, the triggers are individual and the behaviours specific to the child; thus, studies of interventions to change the problem behaviours are often done with individual children, rather than with large groups (which is considered more appropriate for other outcomes).

Data Sources: The author searched PsycLit, ERIC and MEDLINE for the years 1966 to 1998. He also hand searched a large number of pediatric, psychiatric, and behavioural therapy journals.

Study Selection: Papers were selected if the articles were specifically about autistic disorder (and not "autistic-like" behaviour, for example), if the articles reported on single-case studies, if pre and post data were presented and if there were individual scores taken several times during the course of the intervention. Studies had to report on reduction of self-injurious behaviour, stereotypies, aggression and damage to property.

Data Extraction: No details provided.

Main Results: The literature search produced 117 articles that met the criteria for inclusion in the review. The primary finding was that behavioural treatments reduce problem behaviours in people with autism, no matter what the problem and no matter what behavioural intervention is used. Treatments that use repeated experiments to identify situations that either trigger or help to maintain problem behaviours work best to stop the behaviour. Treatment based on experimental functional analysis worked best to completely eliminate the problem behaviours.

Conclusions: This study found no difference in the effectiveness of behavioural interventions, no matter the kind of behaviour (e.g., self-injurious behaviour, stereotypical behaviour, aggression, or property damage). Effectiveness was not influenced by age, I.Q., or verbal ability. It appears that the response to problem behaviour, whether punishment such as time out, or reward does not differ in its effectiveness if it is preceded by experimental functional analysis.