Context: The Centers for Disease Control and Prevention estimates that 1 in 68 children has an autism spectrum disorder (ASD) and the majority of children are not diagnosed until after 4 years of age. Current approaches rely on developmental surveillance, general developmental screening, and/or parental concerns. Systematic screening has been advocated for identifying ASD at earlier ages.
Objective: We systematically reviewed the evidence about benefits and harms of routine screening for ASD in primary care settings.
Methods: We explicitly focused on studies of screening instruments for use in young (≤36 months of age), unselected populations (e.g., universal screening approaches).
Results: We identified 17 unique screening studies reported in 22 papers. The most commonly studied tool was the Modified Checklist for Autism in Toddlers (M-CHAT), including the most recently available variant (M-CHAT-Revised with Followup [M-CHAT-R/F]), which has a positive predictive value of 48 percent in diverse populations of children ages 16 to 30 months. Forty-two studies of good and fair quality addressed interventions for young children. Among these, 17 involved direct provision of intervention to children. Fifteen of these 17 studies assessed cognitive outcomes, and outcomes were significantly more improved in the treatment versus comparison arm in 10 studies. Sixteen of these 17 studies assessed language outcomes, and outcomes were significantly improved in the treatment versus comparison group in 10 studies. Thirteen studies involved parent training. Five of these 13 studies addressed cognitive outcomes, and outcomes were significantly improved in the treatment versus comparison group in one study. Twelve of the 13 studies addressed language outcomes, and outcomes were significantly improved in the treatment versus comparison group in three studies. Thus, 20 studies overall measured cognitive outcomes and 11 reported greater benefit for the intervention group compared to the control group, and language outcomes were significantly improved in treatment versus comparison arms in 13 of 28 studies assessing language. Twelve studies focused on play and interaction and typically measured joint attention as the outcome. Nine out of 10 studies evaluating joint attention outcomes reported greater benefit in the treatment arm compared to the control arm. None of the studies focused on screen-detected children.
Conclusions: Both the M-CHAT and the M-CHAT-R/F, when including the followup interview procedure, have a positive predictive value of around 50 percent in community practices for children between 16 and 30 months of age. Screening tools are widely available. Multiple treatments are available to young children with ASD. Early intensive interventions demonstrate statistically significant improvements in cognitive and language outcomes in children compared to eclectic treatments obtained in the community or other comparison groups, although the studies are generally small, and, within the studies, some children benefit while others do not. We found no studies that directly compared long-term outcomes of screened versus nonscreened children. More research is needed to determine the benefits and harms of screening the general population.