Case definition in epidemiologic studies of AD/HD

Purpose To propose a five-step multigating, multimodal procedure for research case definition and identification of attention-deficit/hyperactivity disorder (AD/HD) in population-based epidemiologic studies. Methods Subjects included a birth cohort of all children born in Rochester, MN, during 1976 to 1982 who remained in Rochester after age 5. Systematic screening for potential cases of AD/HD required four initial steps that included information from school and medical records, a computerized diagnostic index, and private psychiatry records—sources available for all children. In step 5, research diagnostic criteria were applied to all potential cases. The model for defining cases consisted of applying combinations of three categories of information (DSM-IV criteria, questionnaire results, and clinical diagnosis). Validity of the model was tested by comparing different measures (e.g., medication treatment, substance abuse, school outcomes, comorbidities) between subgroups of AD/HD cases who met DSM-IV criteria and those who did not. Chi-square and Wilcoxon sum tests were used to compare categorical and continuous scaled variables between these two groups. Results Among 5718 subjects, 1344 potential cases were identified, 379 of whom met research criteria for AD/HD. No significant difference in gender, treatment, school outcome and specialty of professionals making clinical diagnoses was found between cases who did (N = 228) and did not (N = 151) meet DSM-IV criteria. Compared with cases that met DSM-IV criteria, cases who did not meet DSM-IV criteria were more likely to be inattentive (33.8% vs. 17.1%; P < 0.001), older when criteria were met (mean age 12.8 vs. 10.5; P < 0.01), and have fewer psychiatric comorbidities (17.9% vs. 32.0%; P < 0.001). Conclusion These data underscore the importance of using multiple sources and combinations of documented information for case definition. If DSM-IV criteria alone had been used in this study, 151 mild cases of AD/HD would not have been identified. Studies done in this thorough manner will further the understanding of frequency, risk factors, natural history, treatment, and outcome of children within the spectrum of AD/HD.