John P. Tuman, Sheniz Moonie, and Danielle Roth-Johnson. “The Administrative Prevalence of Autism Spectrum Disorders in Nevada School Districts: A Pooled Time Series Analysis.” Nevada Journal of Public Health (formerly Journal of the Nevada Public Healt
Objective: To examine the administrative prevalence of Autism Spectrum Disorders (ASD) in all seventeen school districts in Nevada during the period of 1996 to 2004. Methods: Normalized administrative prevalence rates (per 1,000 children ages 6-17) for ASD, Mental Retardation (MR), Learning Disability (LD), and Speech and Language Impairment (SLI) were calculated. Covariates for board certified pediatricians per 1,000 students, real Federal special education funding per student, and other measures of school resources were employed. Models were estimated with pooled Ordinary Least Squares (OLS) regression with panel corrected standard errors. A separate analysis compared pooled OLS results to results from Latent Growth Curve models (LCGM). Results: The average administrative prevalence of ASD in Nevada school districts increased from .56 per 1,000 in 1996 to 2.37 per 1,000 in 2004. The upward trajectory of ASD prevalence during the time series was not associated with declines in MR, LD or SLI prevalence, suggesting that diagnostic substitution does not explain the upward trend in the administrative prevalence of ASD. Federal funds distributed partly for detection of disabilities was associated with ASD prevalence (p<0.01) (results were not due to endogeneity). The concentration of pediatricians in each school district, changes in the regulatory definition of ASD, and real salaries for school personnel were shown to have no effect on prevalence. The results of the pooled OLS models were robust when compared to the Latent Growth Curve models.