Child development

SACS-R and How It Watches for Autism

Save

The idea behind the SACS-R is repetition. Autism does not announce itself at a single visit, so instead of one questionnaire the same trained observer checks a handful of highly predictive social behaviors three times across a toddler's second year. In the Australian community study that validated it, that design produced an unusually low false-alarm rate for an autism identification tool.

Last updated: July 2026

Talk to a clinician

Gale can help you find a clinician in your state and request a visit.

Find care →

What is the SACS-R?

SACS-R is the Social Attention and Communication Surveillance–Revised: a set of age-specific checklists that a trained clinician — in the program where it was developed, a maternal and child health nurse — completes by observing a toddler at routine visits. Each checklist covers 12 to 15 early social-communication behaviors, and within each one, five are designated key items because they proved most predictive of a later autism diagnosis 1.

The checks fall at 12, 18 and 24 months, and a preschool extension called the SACS-PR adds a checkpoint at 42 months 1. That schedule is not a detail attached to the instrument; it is the instrument. One checklist in isolation is a much thinner thing than three of them spread across eighteen months of a child's development.

It was built and validated in Victoria, Australia, in a prospective community study of roughly 13,500 children, using nurses at ordinary community health visits rather than autism specialists in a research clinic 1. The SACS-R was designed for the person who already sees a toddler regularly — not for a specialist a family has to be referred to first.

How does a child get flagged on the SACS-R?

By showing an atypical presentation on at least three of the five key items on a checklist. That is the whole rule. A child who meets it is classified as at high likelihood for autism and referred for a diagnostic assessment 1. There is no total score, no percentage, and no severity band; the SACS-R yields one of two outcomes rather than a position on a scale.

That design makes the instrument unusually resistant to over-interpretation. Two children's results cannot be ranked against each other. Nobody can watch a number drift down over time as if it were a temperature. The output is a referral decision, and everything that follows depends on the assessment rather than on the checklist that prompted it.

It also means there is no version of this a parent can complete at home. Judging whether a particular behavior is typical or atypical for a 12-month-old is precisely the part that requires training and a great many prior observations, and it is not something a written description can hand over.

Surveillance and screening are not the same thing

Screening is a discrete event: a validated instrument administered at a set age, returning a positive or negative result. Surveillance is the continuous process around it — noticing, asking, and recording development at every contact, so that a concern can surface between the scheduled screens as well as at them 2. The SACS-R sits deliberately on the surveillance side of that line, which is why the word is in its name.

The practical difference shows up in what happens to a child who looks fine at 12 months. Under a pure screening model, the next data point is the next scheduled screen. Under surveillance, the same observer carries a prior impression forward, so a change — a toddler who used to look up when called and now does not — registers as a change rather than as one ambiguous observation.

Surveillance vs screening is a distinction almost no family is ever taught, and it quietly explains a lot of otherwise confusing advice. It is why a clinician can report that a screen came back negative and still want to see the child again in three months, without that being a contradiction.

What behaviors does the SACS-R watch?

Early social communication, specifically. Autism is characterized by differences in two domains — social communication and interaction, and restricted or repetitive behaviors and interests 3 — and the SACS-R concentrates on the first of them, because in the second year of life those are the behaviors that are visible and countable inside a routine visit 1.

That means behavior of a particular kind: whether a toddler coordinates attention with another person, whether they use their eyes and their hands together to direct someone else's attention, whether they answer social bids and initiate their own. Eye contact and autism is the piece of this families hear about most, and also the most misread — the observation that carries weight is not whether a child ever makes eye contact, but whether looking is woven into a back-and-forth exchange.

Because the checklist watches social communication rather than autism directly, other explanations sit underneath the same observations. That is the reason a flag leads to an assessment rather than to a label: an assessment is what can separate autism from a hearing difference, a language delay, or social communication disorder, and a checklist cannot.

How accurate is the SACS-R?

Across the 12- to 24-month checkpoints in the Victorian study, the SACS-R had a positive predictive value of 83% (95% CI, 0.77-0.87), a negative predictive value of 99%, and specificity of around 99%, with diagnostic stability of 88% at 24 months 1. Put plainly: when it flagged a child, it was usually right, and it seldom flagged children it should not have.

A positive predictive value of 83% 1 is high for autism identification in a general community population, and it comes from the three-key-item rule doing narrow, deliberate work. The cost of that precision appears in sensitivity. The paper's estimated sensitivity reaches 96% (95% CI, 0.94-0.98) only once the 42-month preschool checkpoint is included 1, which is another way of saying that some autistic children are not identifiable on these behaviors at 12, 18 or 24 months and are found later.

The same study reported autism prevalence of 2.0% — 1 in 50 — among children aged 11 to 30 months, rising to 3.3%, or 1 in 31, once the window extended to 42 months 1. Those are figures from one Australian cohort inside a specific surveillance program, and prevalence estimates move with how hard and how early a population is looked at.

Is the SACS-R used in the United States?

Not as routine practice. Under AAP guidance, an American child is watched developmentally at each well-child visit, given a general developmental screen three times before their third birthday — at 9, 18 and 30 months — and given an autism-specific screen twice, at 18 months and again at 24 4. That is the same instinct as the SACS-R, delivered through different instruments and a different workforce. An American family is far more likely to meet a parent-report questionnaire than a nurse-observed checklist.

The gap the SACS-R was built to close, though, is not an Australian problem. In CDC surveillance data for 2020, the median age of earliest known autism diagnosis was 49 months 5 — past a child's fourth birthday, and well after the 18- and 24-month visits where autism-specific screening is meant to happen 4. Between when identification becomes possible and when it typically occurs sits a long stretch of a childhood.

Whether repeated nurse-led surveillance would narrow that gap in a different health system is a question about workforce and funding as much as about instruments. What is worth knowing is that the model exists, that it was tested at community scale, and that it performed well where it was studied.

What happens after a SACS-R flag

A referral for a diagnostic assessment — and, in the part families most often miss, usually a parallel referral for support that does not wait on the assessment result. The behaviors the SACS-R watches are largely the same ones early communication-focused therapy targets, so the interval between a flag and an evaluation does not have to be dead time.

Speech-language pathologists work across screening, assessment, and treatment of social communication in autism 6, which is why autism speech therapy is frequently the first service a flagged toddler actually receives. Occupational and developmental services often follow. None of it presumes what the assessment will conclude, and none of it is wasted if the conclusion is something other than autism.

Parents at this stage tend to arrive at one version of the same question: whether what they are seeing is autism, temperament, or nothing at all. Shyness vs autism is the shape it usually takes, and it is not a question a checklist was ever built to answer. A SACS-R referral does not mean a decision has been reached about a child. It means the next appointment is worth keeping.

Common questions

No. The revised instrument, SACS-R, is what the Australian community study validated, together with its preschool extension, the SACS-PR, at 42 months. The earlier SACS is a separate instrument with its own earlier development work. When accuracy figures for the SACS-R are quoted, they belong to the revised version and its schedule of repeated checks.

No, and that is intentional. It is a behavioral checklist rather than a scored scale, so there is no total, no range, and no direction to interpret. The output is a classification: a child showing atypical presentation on at least three key items is identified as at high likelihood for autism and referred. Everything else comes from the assessment.

No. The checklist depends on a trained observer judging whether a toddler's behavior is typical or atypical for their exact age, which is a skill built through many prior observations rather than something a description conveys. What a parent contributes is different and equally useful: specific examples of what the child does, and when it changed.

Because social-communication skills emerge on a schedule, and a behavior that is unremarkable at 12 months can be meaningful at 24. Repeated checks let the same observer compare a child to their own earlier picture rather than only to a norm, which catches both children who were missed initially and children whose development changed direction.

In the validation study, most flagged children went on to receive a diagnosis, but a meaningful minority did not — and those figures came from a specific program with trained nurses in one region. They do not transfer cleanly to a different setting or a different instrument. The referral is a reason for a careful assessment, not its result.

Say so, and be specific about what you are seeing and when it started. Surveillance is built around exactly this: a negative check is one observation, not a closed question, and a parent's persistent concern is one of the reasons to look again. A hearing test is also commonly part of the answer.

Related

Say it back

How would you explain this to someone you love?

Two or three sentences, just as you’d say it. Gale reflects back what you focused on — a mirror, not a quiz.

Talk to a clinician

Gale can help you find a clinician in your state and request a visit.

Find care →

When to raise a developmental concern regardless of a checklist result

  • —Loss of skills a toddler already had — words that disappear, gestures that stop, a child who used to seek a parent out and no longer does — at any age and regardless of earlier checks
  • —A child who consistently does not turn to their name and has never had a formal hearing test, since hearing loss can look like a social-communication difference and is ruled out first
  • —A toddler who does not combine looking, gesturing, and vocalizing to get an adult's attention in any situation, including when they clearly want something
  • —A concern that persists after a reassuring check — repeated surveillance exists precisely because one negative observation does not close the question

This article describes what the SACS-R is and how it is used. It is not itself a screening instrument, cannot be applied to a child, and does not replace an evaluation by a clinician trained to administer it.

References

  1. 1.Barbaro J, Sadka N, Gilbert M, et al. (2022). Diagnostic Accuracy of the Social Attention and Communication Surveillance–Revised With Preschool Tool for Early Autism Detection in Very Young Children. JAMA Network Open. 2022;5(3):e2146415. doi:10.1001/jamanetworkopen.2021.46415 ✓The SACS-R's structure and administration — age-specific checklists of 12 to 15 early social-communication behaviors, five key items each, administered at 12, 18 and 24 months with the SACS-PR extending to 42 months; the at-least-three-key-items rule producing a high-likelihood classification rather than a score; and its accuracy in a prospective community study of roughly 13,500 children in Victoria, Australia (PPV 83%, NPV 99%, specificity ~99%, diagnostic stability 88% at 24 months, estimated sensitivity 96% with the 42-month checkpoint, prevalence 2.0% at 11-30 months and 3.3% at 11-42 months).
  2. 2.Centers for Disease Control and Prevention (2024). Developmental Monitoring and Screening. CDC — Learn the Signs. Act Early.. linkThe distinction between ongoing developmental monitoring or surveillance and formal developmental and autism screening at recommended ages — used here to explain why the SACS-R is described as surveillance rather than as a one-off screen.
  3. 3.Centers for Disease Control and Prevention (2025). Signs and Symptoms of Autism Spectrum Disorder. CDC — Autism Spectrum Disorder (ASD). linkThat autism is characterized by differences in two domains — social communication and interaction, and restricted or repetitive behaviors and interests — used here to place the SACS-R's focus within the first of those domains.
  4. 4.American Academy of Pediatrics (2024). Developmental Surveillance and Screening. American Academy of Pediatrics — Patient Care. link ✓The AAP-recommended U.S. schedule — developmental surveillance at every well-child visit, general developmental screening at 9, 18 and 30 months, and autism-specific screening at 18 and 24 months — used here to contrast the American model with the SACS-R's nurse-led surveillance schedule.
  5. 5.Maenner MJ, Warren Z, Williams AR, et al. (CDC ADDM Network) (2023). Prevalence and Characteristics of Autism Spectrum Disorder Among Children Aged 8 Years — Autism and Developmental Disabilities Monitoring Network, 11 Sites, United States, 2020. MMWR Surveillance Summaries. PMID 36952288 ✓That the median age of earliest known autism diagnosis in CDC surveillance-year 2020 data was 49 months — used here as the gap between when identification is possible and when it typically happens.
  6. 6.American Speech-Language-Hearing Association (2024). Autism (Practice Portal). ASHA Practice Portal — Clinical Topics. link ✓That speech-language pathologists work across screening, assessment, and treatment of social communication in autism — used here to explain why communication-focused therapy is often the first service a flagged toddler receives.

6 sources, numbered by first appearance. General health information, not medical advice. AI-assisted editorial content — citations link their sources. Editorial policy

← Back to the health library