26418
The Influence of Diagnostic Criteria on Autism Spectrum Disorder Classification: Findings from the Metropolitan Atlanta Developmental Disabilities Surveillance Program, 2012
The Centers for Disease Control and Prevention (CDC) has consistently applied a rigorous and reliable surveillance method to estimate the prevalence of autism spectrum disorder (ASD). This method employs a record-based coding scheme based on criteria outlined in the Diagnostic and Statistical Manual of Mental Disorders – Fourth Edition – Text Revision (DSM-IV-TR). The American Psychiatric Association made considerable changes to ASD diagnostic criteria in 2013 (i.e., DSM-5) that influence CDC ASD surveillance methods.
Objectives:
The objectives of this study were to: (1) evaluate agreement between the DSM-IV-TR and DSM-5 surveillance definitions for ASD, (2) quantify the percentage of children who met only the DSM-IV-TR surveillance definition and the percentage of children who met only the DSM-5 surveillance definition, and (3) evaluate differences in characteristics of these latter two groups of children.
Methods:
A child was eligible for ASD surveillance if he/she: 1) was eight years old during the 2012 surveillance year, 2) had a legal guardian who resided in metropolitan Atlanta, GA, and 3) was served for a developmental condition as evidenced by a discharge diagnosis, billing code, reason for referral, or education eligibility noted in evaluation records. Surveillance staff reviewed health and education records of eligible children for social deficits characteristic of ASD (e.g., limited interest in other children). All records that contained a social deficit were abstracted to collect accounts of developmental history, ASD symptoms, developmental tests, and co-occurring conditions diagnosed by the community professional who evaluated the child. Clinicians with advanced degrees and specialized training and experience in ASD then applied two standardized coding schemes to the abstracted data: one based on DSM-IV-TR criteria that has previously been used to estimate ASD prevalence, and another based on DSM-5 criteria that was developed for this project and future surveillance efforts.
Results:
Results found substantial agreement between the DSM-IV-TR and DSM-5 surveillance classifications of ASD (kappa=0.80). There were no differences in child race/ethnicity, child sex, or intellectual disability between children who met only the DSM-IV-TR surveillance definition and those who met only the DSM-5 surveillance definition. Children who met the DSM-IV-TR but not the DSM-5 surveillance definition (4% of the sample) were more likely to have developmental concerns and evaluations in the first three years. Children who met the DSM-5 but not the DSM-IV-TR surveillance definition (6% of the sample) were more likely to have been receiving autism-related services or previously diagnosed with ASD.
Conclusions:
The ASD surveillance definition based on DSM-5 criteria is largely comparable to that based on DSM-IV-TR criteria, and both identify children with similar demographic and intellectual characteristics. DSM-IV-TR surveillance criteria may detect few children with long-standing social-communication concerns that do not meet the number and pattern of deficits specified in DSM-5. DSM-5 surveillance criteria may detect few children who have an ASD diagnosis and are not captured with DSM-IV-TR surveillance. CDC record-review surveillance is uniquely equipped to continue to evaluate the influence of diagnostic criteria on ASD classification and prevalence estimates in large and diverse samples of children.