The research team implemented and evaluated a training program built around the STAT, the Screening Tool for Autism in Toddlers and Young Children, an instrument designed for children between 24 and 36 months of age. The STAT is not a lengthy, multi-hour battery; it is a focused, interactive assessment that plays directly to the strengths of clinicians who are skilled at observing young children but who have never been taught the specific behavioral markers that distinguish autism from other developmental conditions. The tool probes play, requesting behaviors, and joint attention—the capacity to share attention with another person, which is one of the earliest and most reliable divergences in autistic development. Developed by Wendy Stone and colleagues and psychometrically validated in the early 2000s, the STAT has since accumulated an impressive evidence base, including recent work showing strong agreement between STAT-based classifications and the gold-standard Autism Diagnostic Observation Schedule, Second Edition, in both specialist and primary care settings.

What makes the Wisconsin study distinctive is not merely the training itself but the scaffolding wrapped around it. After completing STAT training, participating primary care providers joined a year-long autism learning community that met monthly for case-based discussions and knowledge exchange. This model echoes the ECHO Autism framework pioneered by Micah Mazurek and Kristin Sohl, in which community clinicians present real cases to a hub of autism experts and receive iterative mentorship rather than a one-and-done workshop. The logic is sound: diagnostic skill, like surgical skill, decays without practice and feedback. A single training day may teach the mechanics of an instrument, but a sustained community of practice is what converts mechanics into clinical judgment.

The study enrolled English-speaking primary care providers in Wisconsin and assessed them before and after training on a battery of measures: knowledge of autism, screening and diagnostic practices, attitudes toward diagnosing autism in primary care, and self-reported comfort with autism identification and management. At baseline, the picture was sobering. Among the 23 providers with baseline data, the researchers found limited use of autism-specific screening and diagnostic tools—a finding consistent with a 2023 scoping review by Johnson and colleagues documenting that autism diagnosis in primary care remains rare, unevenly regulated, and often constrained by state scope-of-practice rules. In other words, the very clinicians positioned to catch the earliest signs of autism were, for the most part, not equipped with instruments designed for the task.

The pre-to-post comparisons tell a more hopeful story, though with an important statistical caveat. Among the 11 providers with matched pre- and post-training data, paired-samples t-tests revealed significant improvements across multiple domains of provider comfort: identifying autism risk, distinguishing autism from other developmental conditions, making formal diagnoses, and—perhaps most delicately—providing diagnostic feedback to families. That last domain deserves emphasis. Delivering an autism diagnosis is not a technical act but a communicative one, and parents consistently report in qualitative studies that the manner of disclosure shapes their subsequent engagement with services. Training that builds confidence in this conversation, not just in the assessment, addresses a dimension of care that pure psychometric instruction often neglects.

The gains were not uniform. Improvements in care coordination domains—referral practices and linkage to mental health services—were more limited. This asymmetry is clinically meaningful. A diagnosis made confidently but followed by a fumbled referral is only half a solution, because the entire rationale for early diagnosis is early access to intervention. Research reviewed by Zwaigenbaum and colleagues has established that intervention before age three can alter developmental trajectories, and the American Academy of Pediatrics’ 2020 clinical report urges identification and evaluation as early as possible. The Wisconsin findings suggest that diagnostic training must be paired with equally deliberate training in the post-diagnostic pathway, or the bottleneck simply moves downstream.

Why does the bottleneck exist in the first place? The arithmetic is unforgiving. The Centers for Disease Control and Prevention’s Autism and Developmental Disabilities Monitoring Network estimated in its most recent surveillance cycle that roughly one in 31 children aged four and eight is identified with autism spectrum disorder. Even at earlier prevalence estimates, the supply of specialist diagnostic capacity has never approached the demand. Wait times at tertiary autism clinics routinely stretch past a year, and children in rural areas, in under-resourced communities, and from families facing insurance or transportation barriers wait longest. Meanwhile, the median age of diagnosis has remained stubbornly higher than the age at which reliable signs first appear, squandering a window of heightened neuroplasticity that early intervention research suggests is precious.

The counterargument to primary care diagnosis has always been accuracy: autism is heterogeneous, its early presentation overlaps with language delay, ADHD, anxiety, and trauma responses, and a false positive or false negative carries real consequences. But the emerging literature is steadily eroding the assumption that only specialists can get it right. A 2023 study in Pediatrics by Rachel McNally Keehn and colleagues evaluated a statewide system of autism evaluation in Indiana and found that primary care clinicians, when trained and supported, achieved diagnostic accuracy comparable to expert teams. A 2023 study in JAMA Network Open by Melanie Penner’s group in Toronto found strong concordance between diagnoses made by pediatricians and those made by a multidisciplinary specialist team. And the ECHO Autism STAT program has demonstrated high agreement between primary care practitioners and autism experts across a dispersed network. The concept of “frank” autism presentations, articulated by Ashley de Marchena and Julia Miller, is central here: children with clear, unambiguous signs of autism—those the authors of the new study call children with obvious signs—represent the cohort for whom primary care diagnosis is most defensible, while complex or ambiguous cases can still be triaged to specialists.

Within this landscape, the Wisconsin study contributes a granular, process-level account of how such a shift can be built. The authors are careful about scope: this was a small cohort, the matched sample for the primary outcome was 11 providers, and the design evaluated implementation and self-reported confidence rather than diagnostic accuracy against a gold standard. Self-reported comfort is a necessary but not sufficient condition for competent practice; the decisive test is whether trained providers’ diagnoses agree with expert adjudication, a question the Indiana and Toronto studies addressed directly and that this study positions but does not itself answer. The researchers also note that scope-of-practice regulations vary by state, with licensed professional counselors’ diagnostic authority, for example, differing across jurisdictions—a legal patchwork that any scale-up effort must navigate. Funding came from the Advancing a Healthier Wisconsin Endowment and a Clinical and Translational Science Institute award, signaling institutional investment in translating the model beyond the study.

Still, the trajectory of the evidence is hard to ignore. Training primary care providers to diagnose autism in toddlers with clear signs, supported by structured tools like the STAT and sustained by learning communities like the one tested here, is associated with meaningful gains in exactly the confidence domains that gatekeep early diagnosis. If future work confirms that this confidence translates into accurate, timely diagnoses at the point of care, the implications ripple outward: shorter waits for the children who most need specialist attention, earlier intervention for those with frank presentations, and a rebalancing of a specialist workforce that can finally focus on the complex cases that genuinely require it. The bottleneck in autism diagnosis was never a law of nature. It was a design choice—and, as this study suggests, one that primary care, properly trained and supported, may finally be ready to undo.

Subject of Research: Evaluation of STAT-based autism diagnostic training for primary care providers in Wisconsin

Article Title: Battling the Bottleneck: Evaluating Autism Diagnostic Training in Primary Care

Article References: Battling the Bottleneck: Evaluating Autism Diagnostic Training in Primary Care. (n.d.). https://doi.org/10.1007/s10803-026-07565-0

Image Credits: AI Generated

DOI: 10.1007/s10803-026-07565-0

Keywords: autism diagnosis, primary care, STAT, screening tool, early intervention, diagnostic bottleneck, provider training, learning community, toddlers, developmental-behavioral pediatrics, ECHO Autism, diagnostic confidence