HOW CASE DEFINITIONS IMPACT PREVALENCE ESTIMATES OF AUTISM SPECTRUM DISORDERS

Background:   There is the feeling that the prevalence of Autism Spectrum Disorders (ASDs) has been increasing in recent years. However, most US-based studies used different case definitions and a variety of age groups. In 2007, the Centers for Disease Control and Prevention (CDC) published prevalence estimates of 1 in 150 eight-year old children with ASD across 14 different sites all using the same case definition. This case definition was based on review of historical records and required evidence consistent with DSM-IV diagnostic criteria.

Objectives: To determine if applying different case definitions of ASD to the same group of children impacts prevalence estimates.

Methods: Four case definitions of ASD were applied to a sample of 127 children: 1) a parent-reported diagnosis of ASD, 2) first diagnosis given, 3) score on the SCQ, and 4) CDC record review case definition. Each case definition was applied to two age groups: 58 children ages 4-5 years and 69 children ages 6-7 years.

Results: Prevalence estimates differed by case definition. For 4-5 year olds, the prevalence estimates of ASD per 1000 were: 1) 1.40 using a parent-reported diagnosis, 2) .96 using ASD as the first diagnosis, 3) .99 using SCQ cutoff, and 4) 1.33 using CDC record review standards. Similar results were obtained for 6-7 year olds: 1) 1.66 using a parent-reported diagnosis, 2) 1.18 using ASD as the first diagnosis, 3) 1.33 using SCQ cutoff, and 4) 1.47 using CDC record review standards.

Conclusions: As expected, how studies define ASD has an impact on the resulting prevalence estimates. Estimates ranged from .96 to 1.66 when different definitions were applied to the same group of children. Differences within and across age groups will be discussed in addition to implications for clinical and policy practices.