Assessment & Research

Detecting autism spectrum disorder from early intervention charts: methodology and preliminary findings.

Towle et al. (2009) · Journal of autism and developmental disorders 2009
★ The Verdict

Early-intervention charts already hold trustworthy autism counts—no new screening budget required.

✓ Read this if BCBAs writing grants or state reports who need quick, cheap prevalence data.
✗ Skip if Clinicians looking for bedside diagnostic tools.

01Research in Context

01

What this study did

Towle et al. (2009) pulled 304 early-intervention charts and checked if the notes could give a clear count of autism cases.

They did not run new tests. They simply read what was already written and compared the totals to published norms.

02

What they found

The chart totals matched the autism rates you see in journal articles.

No extra screening forms were needed; the everyday paperwork was enough.

03

How this fits with other research

Prigge et al. (2013) later added university clinic charts and found four toddlers the usual CDC method missed.

That builds on Towle et al. (2009): same idea, just more data sources.

Lotfizadeh et al. (2020) tried to do the same thing with billing codes instead of notes. It failed; too many kids were labeled wrong.

Together the three papers show rich EI notes work, insurance codes do not.

04

Why it matters

If you need a fast count of how many kids with autism your region serves, ask for the EI files first. You can give your director real numbers without running a single new test.

Free CEUs

Want CEUs on This Topic?

The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.

Join Free →
→ Action — try this Monday

Email your EI coordinator and request last year’s intake list; flag any child with ‘ASD’ or ‘autism’ in the service column.

02At a glance

Intervention
not applicable
Design
other
Sample size
304
Population
autism spectrum disorder, developmental delay
Finding
not reported

03Original abstract

Federal laws mandating a "single point of entry" for early intervention (EI) create a potential database for surveillance of early childhood disabilities. This study evaluated EI records for estimating rates of autism spectrum disorder (ASD) using a chart abstraction protocol, with good interrater agreement (k = .86). Sampling 304 EI records yielded a point prevalence of (per 1,000) 8.5 (95% CI: 4.8-10.9) and a cumulative incidence of 7.4 (95% CI: 5.5-12.4). These rates are similar to recent published estimates. Additionally, the male-to-female ratio for autism, and rates of other developmental disorders were found to be consistent with current literature. These results suggest that local systems EI records may provide an excellent resource for ASD surveillance and research.

Journal of autism and developmental disorders, 2009 · doi:10.1007/s10803-008-0643-x