There is controversy about the need to screen every toddler for autism.
The American Academy of Pediatrics argues that a pediatrician should screen every child at both 18 and 24 months of age. Conversely, the United States Preventive Services Task Force released a draft recommendation that children without any presenting symptoms should not be screened.
Pediatrician Aaron E. Carroll has laid out a rationale for why every child should be screened. Key points include the high prevalence of autism (estimated now to be 1 in 45), the availability of an easy to administer and cost-effective screening device (typically a 23-item questionnaire), and the demonstrated effectiveness of early treatment that is dependent on early detection. None of this is controversial.
Given these considerations, why then would there be opposition to universal screening? The contention is that screening is unnecessary if a toddler is showing no signs of autism, and should be reserved only for cases in which there is some presenting symptomatology. Such is the basis for the task force’s lack of endorsement for universal screening.
This position is misguided. The fundamental purpose of screening is to provide an empirically grounded method, provided by a professional, to uncover undetected symptoms that are often subtle or ambiguous and hence can easily be missed. For example, even though we promote awareness of potential signs of skin cancer and encourage people to screen themselves, we endorse professional screening because an individual is not in the best position (physically and professionally) to detect or interpret potential signals of malignancy, we have reliable methods for screening, and early screens can lead to early diagnosis and more effective intervention. This is different, in spirit, than performing an evaluation when an individual seeks out an evaluation because they have done, in essence, their own screening.
Applying this perspective to autism helps clarify the utility of universal screening. Parents should not have the burden of being diagnosticians who conduct their own screens of their children. Of course, we want them to have as much awareness of the signs of autism as possible, be vigilant about observing them, and seek out clinical evaluation if they become concerned. That said, the signs of autism manifest in many ways, especially as the symptoms may be expressed across a spectrum of severity. They may emerge slowly within the stream of daily development as it unfolds over the first two years of life. It is hard to know when the absence of a developmental milestone is a potential red flag or simply normal variation in timing. This is why the field has devoted considerable scientific effort to developing sensitive behavioral screens for autism, to be administered at key developmental periods (18 months and 24 months) when symptoms can be detected with reliability and evidence-based guidelines can be followed to determine if a full evaluation is warranted. Such an evaluation is then used to initiate interventions if necessary.
We continue to see evidence that early intervention works, especially newer programs. For example, there is a growing body of data supporting the effectiveness of the Early Start Denver Model (ESDM), a novel intervention for autism that is designed for implementation during toddlerhood. One study documented that there are both immediate and longer-term gains — two years later — in terms of reducing symptoms, as compared to typical intervention. And we know from many studies that typical intervention leads to substantial gains compared to no intervention, and the earlier we administer the intervention, the better the prognosis.
The point of rational, cost- and time-effective, and empirically validated universal screening for autism is to ensure that we don’t miss any potential cases at those developmental moments when intervention may have the most powerful effects. Why would we want to do less than that?