Should all children be screened for autism spectrum disorders? No: screening is not ready for prime time.

The Centers for Disease Control and Prevention has declared that autism spectrum disorders (ASDs) are “an urgent public health concern,” with recent estimates that one in 110 children has an ASD.1 Before routine screening was recommended, the median age of diagnosis was 5.7 years,2 despite 70 to 95 percent of children having identifiable symptoms by three years of age.1 The primary care physician is challenged to identify children at risk of ASDs before they enter school, with referral for early diagnosis and effective intervention when developmental delays are detected. Screening for ASDs in primary care was advanced by the American Academy of Pediatrics (AAP) through its policy statement on developmental surveillance and screening, and its clinical report on the identification and evaluation of children with ASDs.3,4 The guidelines aim to identify infants and toddlers at risk of developmental delay or disability, leading to further diagnosis and treatment in federally mandated early intervention programs. The AAP recommends developmental surveillance at every preventive care visit, with additional use of standardized general screening tests at the nine-, 18-, and 30-month visits, and ASD-specific screening tests at the 18and 24-month visits, because general screening tests may not identify toddlers with ASDs.3,5 Continued surveillance for ASD symptoms through school age is also recommended because higher functioning children may not be identified as symptomatic until social demands become more complex. The recommendations for ASD screening were based on promising results using the Modified Checklist for Autism in Toddlers (M-CHAT), a parent questionnaire with a reported sensitivity of 87 percent and positive predictive value of 80 percent.6 Subsequent evaluation of the M-CHAT has revealed weakness, with a positive predictive value as low as 11 percent (or a false-positive rate of 89 percent) when used as a questionnaire alone; however, the positive predictive value increases to 65 percent (false-positive rate of 35 percent) with a structured interview follow-up. For example, if the initial evaluation is positive, then a second visit with the physician or office staff would be scheduled to answer additional questions and possibly refer the child for further testing and intervention.7 Screening tools that have since been developed may be more specific while also identifying children at risk of other developmental disorders. One example, the Infant-Toddler Checklist, screens for ASDs and language delay in children nine to 24 months of age, with a positive predictive value for language disorders higher than 70 percent and sensitivity for ASDs higher than 90 percent.8 With its use in screening more than 10,000 children who were one year of age in 137 pediatric practices, the positive predictive value remained high at 75 percent.9 Community nurses achieved sensitivities of approximately 70 to 85 percent by combining ongoing developmental surveillance for ASD symptoms and observation Should All Children Be Screened for Autism Spectrum Disorders?