Do Domains of Developmental Risk Identified By Parents during ASD Screening and Areas of Risk Identified during ASD Diagnostic Assessment Differ: Findings from a Case Series of Ethnically Diverse, Low-Income Toddlers

Background:  Universal screening for Autism Spectrum Disorder (ASD) is a recommended strategy to address disparities in age of ASD diagnosis. The M-CHAT is the most widely used instrument to screen children primary care settings. Recent studies suggest that M-CHAT performs differently based on family demographic characteristics. However, few studies have explored how such differences correspond to performance on ASD diagnostic assessments.

Objectives: To explore the relationship between domains of developmental risk identified by low-income families during ASD screening and areas of risk identified during ASD diagnostic assessment.

Methods:  Screening (MCHAT-R/F) and diagnostic (ADOS-2) data were collected from 44 children who were referred for an ASD diagnostic evaluation and were participants in randomized control trials studying methods to improve early identification of ASD. All met ASD risk criteria based on the MCHAT-R/F initial screen and follow-up interview. MCHAT-R/F items were categorized based on the four ADOS-2 domains: communication, reciprocal social interaction, play, and restricted and repetitive behaviors. For each domain, we assessed whether areas of risk identified in the MCHAT-R/F and ADOS-2 were concordant or discordant. For discordant results, we assessed patterns of risk identified by parents compared to the child’s performance on the ADOS-2.

Results:  Twenty-four of 44 children meet ASD diagnostic criteria, yielding a PPV of 55%, consistent with published MCHAT-R/F performance characteristics. Twenty children were not diagnosed with ASD. In this low-income (91%), majority minority sample (84%), children who were not diagnosed with ASD were significantly more likely to come from larger, single- parent households. Based on responses to items on the MCHAT-R/F, 80% of parents endorsed risks in the domain of communication; 79% in reciprocal social interaction; 15% in play, and 20% in restricted and repetitive behaviors. Based on performance on the ADOS-2, 75% of children met diagnostic criteria related to communication; 35% of children related to reciprocal social interaction; 21% of children related to play, and 39% related to restricted and repetitive behaviors. The most common discordant pattern was related to reciprocal social interaction, which was identified as an area of risk by 80% of parents and was confirmed during diagnostic assessment in only 35% of children. The most discriminatory item on the MCHAT-R/F for ASD was item one, “If you point at something across the room, does your child look at it?”; 80% of children who failed this item were diagnosed with ASD. The most discordant item was item eighteen, “Does your child understand when you tell him or her to do something?”; 79% of children who failed this item did not receive an ASD diagnosis.

Conclusions:  This study identified two areas that may play a role in classifying low-income, largely minority children who do not meet diagnostic criteria for ASD as having elevated risk.  The first is parental appraisal of children’s social interaction skills and the second is parental expectation of children’s response to parent-initiated communication. Further research that explores parental understanding of developmental norms is needed and could lead to more accurate assessment of ASD risk in this population.