Concern May Not Equal Action: Outcome Data in a Universal ASD Screening Sample

Background: Recently, the US Preventative Services Task Force (USPSTF) issued a call for more research on effects of universal screening for autism spectrum disorder (ASD).  This call came in a draft statement indicating insufficient evidence exists for recommending autism screening for all children.  Instead, the USPSTF supported screening for ASD in the presence of a concern.   The American Academy of Pediatrics immediately issued a re-affirmation that universal screening of toddlers at 18 and 24 months was best practice in addition to screening any time there is a concern.  Although studies have looked at universal screening, and others have looked at outcomes from early intervention, no studies have been conducted that link children identified through universal screening to their outcomes.  

Objectives: This study provides a detailed examination of actions and outcomes within a small group of screen-identified children from a large community sample, including re-evaluations after third birthdays. 

Methods: Participants (n=42, aged 15-33 months) were typically developing controls and children who screened positive following universal autism screening within a large community pediatric practice (n=796). Three children in the sample were referred for ASD evaluation prior to universal screening. Diagnostic evaluations including ADOS at time of screening (n=42) and after 36 months of age (n=22) were performed to determine group status (autism [n=14], language delay [n=14], or typical [n=14] after initial evaluation). Licensed psychologists with toddler and autism expertise, unaware of group status, analyzed 10-minute video samples of participants’ original autism evaluations. Raters were asked for autism referral impressions based solely on observations to determine concern about atypical behaviors.  Data regarding medical histories, concerns, and risk factors at the time of screening, and eventual actions following screening and evaluation were gathered, as were assessments of cognitive and adaptive abilities at initial evaluation and at follow up during the fourth year of life. 

Results: Although concerns of parents or others were prevalent in the autism group (12/14 children), the presence of concerns did not translate to action for many parents  (only 7/14 in the autism group had discussions with physicians prior to screening).  Prior to universal screening, 3 of the children were referred to early intervention based on concerns for ASD, 2 were referred for speech or feeding.  Following universal screening, 9 children were enrolled in services, 1 did not enroll, 1 was eventually not diagnosed with ASD, and 3 Spanish-speaking families were lost to follow up after identification, so outcomes are unknown.  In the language delay group, 5 families enrolled in speech services following universal screening, 2 did not, 1 enrolled prior to screening, 6 were lost to follow up.  Children in language and autism groups had improved or stable cognitive scores with few exceptions, and mixed adaptive score changes.

Conclusions: Screening in the presence of concerns is supported, but many parents within the sample did not take action on concerns prior to universal screening.  The universal screening process may have accelerated identification of children with ASD, resulting in earlier intervention enrollment and stable or improved cognitive and adaptive outcomes.