In this issue of JAMA, the US Preventive Services Task Force (USPSTF) presents final recommendations for screening children aged 18 to 30 months for autism spectrum disorder (ASD).¹ Autism spectrum disorder can be responsible for significant, long-term impairment in social interaction, communication, and functional capacity. Emerging evidence suggests that early intensive behavioral therapy has the potential to improve outcomes.² The prevalence of ASD has increased in recent years and is estimated to be as high as 1 in 68 children,³ suggesting that improving long-term outcomes could have substantial societal impact.

The USPSTF reviews topics that focus only on asymptomatic populations and services provided in (or referable from) primary care and makes its recommendations by following a transparent set of methodologic processes.⁴ The USPSTF concluded that, “the current evidence is insufficient to assess the balance of benefits and harms of screening for ASD in young children for whom no concerns of ASD have been raised by their parents or a clinician”¹ and has given the recommendation an I grade. The I grade stems from the small size and variable quality of intervention studies and from a lack of direct evidence that screening leads to clinical improvement.¹

Although this recommendation may be disappointing to many people, the USPSTF has appropriately applied its methodology to the question of ASD screening and has fulfilled its charge of applying rigorous analysis to the best available evidence. Asking the USPSTF to consider downstream policy or practice ramifications of its decision, or others’ interpretations of it, would be to devalue its critical role as an impartial evaluator of evidence. Professional societies, constituent groups, families, and individual clinicians will need to contextualize the USPSTF’s evaluation within a broader discussion on policy and patient care and assess, based on this broader context, whether universal ASD screening is warranted.

In considering the USPSTF’s recommendation, the issue of “direct evidence” is critical. Although validation studies might prove a screening test to be good at identifying a certain condition, and separate interventional studies might show improvement in those treated for that condition, this piecemeal analytic framework demonstrates only the strength of individual links in a complicated process chain that includes screening, diagnostic evaluation, engagement with treatment, and improved health outcomes. Conversely, higher-level direct evidence addresses the more overarching question of whether universal screening results in clinical improvement—a broader analytic framework that considers the process chain in its entirety. Depending on the condition in question and the strength of the contributing research, evidence supporting individual links in a chain may be sufficient to produce a positive recommendation. In the case of ASD screening, the USPSTF decided that the evidence was not.

In explaining this decision, the USPSTF questions whether the interventional studies included in their review could be extrapolated to a screen-detected population and cites the likelihood that children identified through screening would be younger and less severely affected than the participants of these studies.¹ While this may be an appropriate justification, it is only one dimension of a broader question regarding the level of evidence necessary to demonstrate the value of a clinical process as complicated as screening for, diagnosing, and treating ASD.

Autism spectrum disorder is a heterogeneous condition.^3,5 Its considerable variability in prevalence and age of diagnosis demarcates fault lines of socioeconomic status, race, ethnicity, and geography^3,6—suggesting both disparities in access to services and divergent, regional cultures of practice. The complex behavioral interventions administered to children with ASD vary in both content and delivery, involve a set of clinicians outside the domain of primary care, and produce variable outcomes among the heterogeneous population of children with ASD.⁷ At each step along this screening-to-treatment chain are children (and families) who get screened or do not, get diagnosed with ASD or do not, and respond to therapy or do not. This sorting happens based on nonrandom factors, likely (but not necessarily) having to do with accessibility of services, patient-provider communication, service engagement, cultural relevance of care, and health literacy. Studying such a process chain as individual links ignores these complexities, and it is inadequate to prove, or disprove, the effectiveness of ASD screening.

Some critics of the USPSTF’s decision have argued that accumulating direct evidence that accounts for all the complexities around diagnostic confirmation and engagement with treatment is prohibitively expensive and not feasible.^8,9 While it is important to recognize the challenges involved with accumulating such evidence, the complexity of a clinical problem is no justification for adjusting the USPSTF’s standards for a positive recommendation.

Rather, consistent with the intent of an I statement,¹⁰ the current USPSTF report is an opportunity to motivate better, more relevant research. Here, the evolution of the USPSTF’s report on adult depression provides an instructive example of how that can happen. The USPSTF currently endorses depression screening when systems (previously termed “depression care supports”) are in place to ensure accurate diagnosis, effective treatment, and follow-up.¹¹ In its now archived 2009 statement,¹² the USPSTF had accompanied this endorsement with an explicit recommendation against universal depression screening in the absence of such supports. As with ASD, depression is fraught with issues of diagnostic subjectivity, comorbidities, cultural nuances, and difficulties with engagement. By analyzing the highest level of direct evidence for depression screening, the USPSTF was able to determine a critical set of circumstances under which the practice leads to better outcomes. Today, depression care supports have become so “much more widely available and accepted as part of mental health care,”¹¹ that the USPSTF was able to omit its comment on what to do when these supports are absent. A similar evolution can occur for ASD screening; in fact, the National Institute of Mental Health is currently supporting 5 large-scale studies on early identification of ASD and linkage to services, designed to address overarching questions encompassing universal screening, expedited diagnosis, and engagement with treatment.¹³ Such initiatives are a promising step in discerning the direct relationship of ASD screening to clinical outcomes.

Critics of the USPSTF’s decision have also raised the possibility that the I recommendation may be prone to misinterpretation and thus have unintended consequences.⁹ This concern stems largely from the Affordable Care Act’s mandate to provide first-dollar coverage for all preventive services with an A or B grade. Although the wisdom of this policy has been questioned,¹⁴ the link of a positive USPSTF recommendation to insurance coverage is not meant to deny coverage for preventive services with lesser evidence grades. Rather, an evidence grading is but one of many factors that contribute to whether a service should be practiced and reimbursed. Although the possibility of such misinterpretation is real, the USPSTF should not base its recommendations on this concern. In addition, in both previous communication¹⁰ and in response to public comment concerning its assessment of ASD screening,¹ the USPSTF has been clear in how an I statement recommendation should be interpreted: it is neither a recommendation for or against screening nor a recommendation against insurance coverage. Rather, it is “a call for more research.”¹⁰

Society benefits from the existence of advocacy groups, particularly when those groups represent the needs of vulnerable populations. It also benefits from the existence of professional medical societies that organize and speak on behalf of groups of health care professionals and their patients. However, underlying the policies and guidelines that such organizations create, there should be a bedrock of evidence, vetted by a scientific body that is free (to the extent possible) from outside influence or broader policy considerations. In rendering its determination of insufficient evidence for ASD screening, the USPSTF demonstrated its understanding of the real-world complexities of primary care and its commitment to be a rigorous, transparent arbiter of best available evidence. The ensuing debate around the findings of the USPSTF with respect to screening for ASD has thrown into relief the importance of this circumscribed but critical role. The USPSTF has delivered an important message, which should spur more research and enhance the knowledge base around universal ASD screening. The USPSTF embraced this issue in all its complexity. Physicians, other health professionals, policy makers, insurers, and other stakeholders should do the same.

Corresponding Author: Michael Silverstein, MD, MPH, Department of Pediatrics, Boston Medical Center, 88 E Newton St, Vose 3, Boston, MA 02118 (michael.silverstein@bmc.org).

Conflict of Interest Disclosures: Both authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.