Surveillance of Autism Spectrum Disorders (ASD) in Select Children who Live in Minneapolis: Do Children of Somali Descent Have a Higher ASD Prevalence?
Michael Reiff, PhD, University of Montana
Project Summary:
The purpose of this project is to implement the Center for Disease Control and
Prevention's (CDC) Autism and Developmental Disabilities Monitoring (ADDM) Network protocol to
estimate ASD prevalence among a defined population of children in Minneapolis and assess whether the children of Somali descent have a higher ASD prevalence than children of other racial and ethnic groups.

Goal 1: Define a population of Minneapolis children in which to ascertain ASD prevalence and
obtain access to relevant records. Objective 1: The population tentatively identified for this surveillance
effort was based on 2005-2009 American Community Survey 5-year population estimates and represents approximately 19,752 children in Minneapolis ages 5-9 years.2 MN LEND will confirm and refine as appropriate the population tentatively identified for this surveillance effort. Objective 2: MN LEND (Dr. Reichle) will collaborate with the Minneapolis Public Schools, various charter schools, and Drs. Reiff and Esler with medical/clinical providers in the Twin Cities area who diagnose autism to access educational and clinical records.

Goal 2: Implement established ASD public health surveillance methodology to gather valid and
reliable data. Objective 1: The MN LEND will employ ADDM case identification methodology
including: (1) Case ascertainment: source files from educational and medical sources will be screened to identify children who meet residency requirements and are suspected of having ASD. (2) Clinician
review: composite summary records for all abstracted cases will be reviewed and scored by an ASD
clinician reviewer meeting ADDM certification criteria. Using the ADDM coding guide, clinical
reviewers will determine if a child's symptoms and behaviors are consistent with the ADDM ASD
surveillance case definition. Objective 2: MN LEND will collaborate with the University of Wisconsin
(UW) ADDM site to ensure that data are collected in a manner consistent with ADDM methodology.

Goal 3: Determine whether Somali children have a higher estimated ASD prevalence than children
from other racial and ethnic groups. Objective 1: The MN LEND will access census data for the defined
population, overall and stratified by race/ethnicity and Somali descent, to estimate the denominator for
ASD prevalence. Objective 2: ASD population prevalence will be estimated, overall and stratified by
race/ethnicity/Somali descent and compared using appropriate statistical methods. Objective 3: A
sensitivity analysis will be performed to examine the stability of ASD prevalence estimates to account for
non-response of school districts and uncertainty regarding Minneapolis census population estimates.

Goal 4: Verify final case status in a sample of children identified from ASD surveillance activities
through direct clinical evaluation. Objective 1: Verify case status of all identified Somali children who
meet the ASD surveillance case definition. Objective 2: Verify case status of a random sample of non-
Somali children who meet the ASD surveillance case definition.

To achieve the project's goals in the timeframe given, and to ensure comparability of data with the
ADDM Network, we would request technical assistance from CDC to guide in: (a) training of abstractors
and clinician reviewers on ADDM methodology3. To facilitate rapid data collection and eliminate the
need for software development, additional technical assistance may be requested to access and utilize the CDC's Alliance for Research in Child Health Epidemiology (ARCHE) database.