The Systematic Observation of Red Flags for Autism in the First Year of Life for Infants at Risk

Background: Studies focused on high-risk (HR) siblings of children with ASD demonstrate that core ASD symptoms can be detected as early as 12 months (Bryson, et. al, 2008; Watson et. al, 2013; Zwaigenbaum et. al, 2005), highlighting the opportunity for earlier screening. The Systematic Observation for Red Flags (SORF/Cite) is an observational screening tool based on DSM-5 diagnostic criteria designed to quantify presence of ASD symptoms. The SORF has been validated to differentiate 16-24 month-olds with ASD from those with developmental delay and typical development (Dow et al., 2017). Given the pressing need for earlier detection, the current study examined utility of the SORF among 12-month-old HR infants.

Objectives: In a prospective, longitudinal study of infant siblings, we examined: 1) sensitivity and specificity of the SORF at 12-months for predicting ASD at 24-months, and 2) associations between 12-month SORF performance and 24-month phenotypic outcome

Methods: Participants included 122 infants seen for communication, developmental and social-behavioral assessment at 12 and 24-months of age. Participants were either HR (N=45,) or low-risk infants (LR; N=77). The Communication and Symbolic Behavior Scales-DP Behavior Sample (Wetherby & Prizant, 2002) and the Mullen Scales of Early Learning (Mullen, 1995) were administered at 12-months. The SORF was coded via video review of the CSBS and provides a measure of ASD symptoms (represented as a score and total number of red flags) across two domains: Social Communication and Social Interaction (SC) and Restricted Repetitive Behaviors (RRB). At 24-months, participants received a diagnostic evaluation that included the CSBS, Mullen, ADOS-2, and clinician best estimate diagnosis.

Results: See results in Table 1 and 2. 12-month-olds later diagnosed with ASD scored significantly higher than TD infants on the SORF SC domain, RRB domain, and overall composite. Infants with ASD also had significantly more SC and total red flags (RFs), with the difference in RRB RFs approaching significance. Highest sensitivity and specificity was observed for the SORF Composite with a cutoff of 18, which correctly identified 24 of 31 ASD infants, yielding a sensitivity of .77 and specificity of .76. Optimal cutoff for SORF Total RFs was 7, with 20 of 31 infants having 7 or more RFs at 12 months, yielding a sensitivity of.65 and specificity of.75. The SORF composite score and SC domain score at 12-months were both highly correlated with ADOS-2 CSS and verbal and nonverbal skills at 24-months.

Conclusions: Our results show that 12-month-olds later diagnosed with ASD exhibited significantly greater SORF total and SC domain scores compared to TD-infants. Additionally, SORF Composite cutoff of 18 resulted in higher sensitivity and specificity than SORF red flags cutoff of 7. SORF RRB scores were less associated with ASD when considered in isolation, in line with other research showing RRBs unfold over time and are quite subtle in the first year (Zwaigenbaum et. al, 2015, Bryson et. al 2008). The high sensitivity and specificity of the SORF at 12-months highlights the need for 12-month screening studies that include larger samples of both high-risk and community-based infants.