Developing UK ASD Research Capacity: Regional and UK ASD Research Databases Include 2500 Children Representative of the UK ASD Population

Background: ASD research often requires large numbers of participants, or participants with uncommon characteristics. To improve recruitment to studies, and enable families to take part in research, we developed two large research databases: the Database of Children with ASD in the North East (Daslⁿe), which has a population sampling frame, and a national database, the Autism Spectrum Database-UK (ASD-UK).

Objectives: To demonstrate the extent to which the databases include children who are broadly representative of the overall childhood ASD population by presenting:

1. Whether children of parents who consent to join ASD-UK are similar to those who do not want to participate.
2. The similarities and differences between the children included in Daslⁿe (shown representative of the regional population, McConachie et al., 2009 Archives of Disease in Childhood) and the children recruited in the first 30 months to ASD-UK.

Methods: Children with ASD are identified in North East England through local ASD assessment teams, and lists held by education. For ASD-UK, parents are approached through more than 60 Child Health teams. All parents give informed consent, and complete information packs about themselves and their child (or children) with ASD, and siblings. Children’s diagnosis data are validated through information obtained from clinicians. Daslⁿe and ASD-UK have permission to collect a small amount of data about non-consenting families.

Results: After 10 years, Daslⁿe includes 1400 children – around 55% of children aged 2-18 diagnosed with ASD from the local population. After 30 months’ recruitment, 1003 children have been included on ASD-UK.

Similarities between children/parents who consented to join ASD-UK, and those who did not: Families who consented to join ASD-UK were very similar to those who reported they did not want to participate (n=184) on the following characteristics: DSM-IV ASD diagnosis, gender (4:1 Male:Female) and Townsend social deprivation score derived from postcode (consent versus refusal difference 0.75 on a 21 point scale).

Similarities between children recruited to Daslⁿe and ASD-UK: There were similar proportions of males and females recruited to each database, and also similar proportions of children with a parent-reported learning disability. Considering DSM-IV ASD diagnosis, children recruited to ASD-UK were more likely to have an ASD diagnosis (62% vs. 50%), and slightly less likely to have an Autism or Asperger syndrome diagnosis. Daslⁿe children were slightly more likely to attend mainstream school (55% vs. 50%).

Conclusions: Researchers wishing to recruit from ASD-UK and the population-based Daslⁿe can do so knowing that the two databases are both as representative as possible of the ASD child population.