A practical approach to identifying autistic adults within the electronic health record.
Add a quick NLP keyword scan to ICD codes when you pull EHR data on young adults to find 95-99 percent of autistic clients.
01Research in Context
What this study did
The team built a computer tool that reads electronic health records. It looks for autism clues that billing codes miss.
They tested the tool on 18- to young learners already seen at two large hospitals. Gold-standard clinician charts showed who truly had autism.
The tool mixed ICD codes with natural-language notes like "social scripts" or "sensory break." This extra step took minutes once set up.
What they found
The new method caught 95-99 out of every 100 autistic patients. Old code-only searches caught far fewer.
False alarms stayed low: about 2 in 10 flagged youths did not have autism. That rate is good enough for large service-planning lists.
Adding the word hunt boosted sensitivity by 20-30 percent without extra staff time.
How this fits with other research
Barton et al. (2019) found a 5-item toddler screener works well for babies. Waldron et al. (2023) works for adults. Same goal, different ages—no conflict, just a life-span gap.
Sievers et al. (2020) created a personality-test rule for intellectually-able adults. Their tool needs a clinic visit. The EHR tool needs only chart access, so you can use both: one for outreach, one for in-person follow-up.
Osório et al. (2025) showed autistic toddlers walk with jerkier steps. Future EHR versions could add gait-keywords to flag even younger cases during routine check-ups.
Why it matters
If you plan transition services, run the NLP-plus-code search first. You will build a fuller client list in minutes. Then invite those youths for intake, saving weeks of missed cases. Share the list with vocational and college-support teams so they start funding requests earlier.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Ask your IT team to run the free autism keyword list on last year’s 18-25 EHR files and compare the new count to the old ICD-only list.
02At a glance
03Original abstract
The electronic health record (EHR) provides valuable data for understanding physical and mental health conditions in autism. We developed an approach to identify charts of autistic young adults, retrieved from our institution's de-identified EHR database. Clinical notes within two cohorts were identified. Cohort 1 charts had at least one International Classification of Diseases (ICD-CM) autism code. Cohort 2 charts had only autism key terms without ICD-CM codes, and at least four notes per chart. A natural language processing tool parsed medical charts to identify key terms associated with autism diagnoses and mapped them to Unified Medical Language System Concept Unique Identifiers (CUIs). Average scores were calculated for each set of charts based on captured CUIs. Chart review determined whether patients met criteria for autism using a classification rubric. In Cohort 1, of 418 patients, 361 were confirmed to have autism by chart review. Sensitivity was 0.99 and specificity was 0.68 with positive predictive value (PPV) of 0.97. Specificity improved to 0.81 (sensitivity was 0.95; PPV was 0.98) when the number of notes was limited to four or more per chart. In Cohort 2, 48 of 136 patients were confirmed to have autism by chart review. Sensitivity was 0.95, specificity was 0.73, and PPV was 0.70. Our approach, which included using key terms, identified autism charts with high sensitivity, even in the absence of ICD-CM codes. Relying on ICD-CM codes alone may result in inclusion of false positive cases and exclusion of true cases with autism.
Autism research : official journal of the International Society for Autism Research, 2023 · doi:10.18637/jss.v098.i11