Autism screening

Screening is one step in identification

Families may encounter related processes:

Permission to administer or score a screener does not itself confer authority to diagnose autism.

The CDC clinical screening page says screening tools do not diagnose autism and a positive result should be followed by thorough assessment; CDC lists tools but endorses none. The current AAP screening guidance says increased likelihood identified through screening or surveillance warrants diagnostic referral, while intervention should address identified developmental delays without waiting for an autism diagnosis.

Routine pediatric screening has recommended ages

The current American Academy of Pediatrics screening page recommends autism-specific screening for all children at the 18- and 24-month health-supervision visits, together with regular developmental surveillance. CDC's February 2026 monitoring and screening page also reports AAP recommendations for general developmental screening at 9, 18, and 30 months and autism screening at 18 and 24 months.

These ages are routine pediatric checkpoints, not a deadline for identification. CDC says a missed recommended screen should be completed at the next visit, and screening should also occur when a parent or clinician has a concern. Increased likelihood found through surveillance may justify additional screening or direct diagnostic referral. Older children, adolescents, and adults need age-appropriate evaluation pathways, not a toddler tool used beyond its validated range.

CDC also notes that the 2016 USPSTF found insufficient evidence for or against universal screening of children age 3 or younger who have no signs or concerns; that finding was not a recommendation against screening.

What happens during screening

A caregiver may complete a questionnaire about development and autism-related characteristics. A trained professional may observe or guide brief activities. Some tools require a follow-up interview after certain initial results. Follow the current manual for age, language, administration, scoring, repeat rules, and referrals.

Use an authorized language version whose instructions and evidence fit the person and setting. A qualified interpreter may support communication but should not improvise items or scoring. Document the tool and version, respondent, language, interpreter role, unanswered items, access supports, and deviations. Preserve AAC access and consider hearing, vision, motor, sensory, health, culture, literacy, and opportunity before interpreting a result. Include the person's account when possible; family report does not replace it.

The AAP clinical report on autism identification, evaluation, and management says more accurate, culturally sensitive approaches are needed and ongoing surveillance through school age is important. Some children are identified later when social differences become clearer as school demands increase. This supports attention to changing context, not reuse of a toddler tool outside its validated age range.

Read the result under the tool's rules

Reports may use terms such as negative, lower likelihood, elevated likelihood, positive, inconclusive, or follow-up needed. Ask for the exact term, score date, tool version, age and language range, completed follow-up steps, limitations, and recommended action.

A concerning result means more evaluation is warranted. It does not mean the screener has diagnosed autism. A result below a cutoff reduces concern only within the tool's tested conditions. It cannot erase family or professional observations, developmental change, skill loss, or a person's request for evaluation.

Sensitivity is how often a tool identifies people who meet the validation study's reference condition; specificity is how often it classifies those who do not. A false positive is an elevated result when the reference evaluation does not identify autism; a false negative is a lower-likelihood result when that evaluation does. Predictive values vary with population and setting. These group statistics never give certainty for one person, and persistent concerns remain actionable after a lower-likelihood result.

A screen can lead to several next steps

NIMH's autism overview describes screening and diagnostic evaluation as separate stages. A comprehensive evaluation may include medical and neurological examination, developmental history, observation, cognitive and communication assessment, adaptive functioning, family history, and hearing or other tests as indicated. The team and components depend on age, question, jurisdiction, and professional scope.

Families can ask:

• What exactly did the screen find, and what could it miss?
• Who can complete the diagnostic evaluation for this age?
• Should hearing, vision, speech-language, occupational, medical, educational, or genetic questions be assessed?
• Which supports can begin while an evaluation is pending?
• How do we request state early-intervention or school-system evaluation?
• What records and observations should we bring?
• How will language, culture, AAC, disability access, and the person's own account be included?

The AAP advises starting intervention for identified developmental delays while an autism evaluation is pending. Screening by itself does not establish that ABA, another therapy, or any fixed intensity is appropriate. Services should follow actual needs, preferences, evidence, eligibility, and qualified assessment.

Respond promptly to concerns and skill loss

The February 2026 CDC family guidance tells families to discuss concerns with a clinician, request developmental screening, and contact the state early-intervention program for a child under three or the local public school for a child three or older. Eligibility and process vary by jurisdiction. A family can pursue medical and educational routes at the same time.

Report loss of previously used skills, new seizures, pain, sudden behavior change, hearing or vision concern, or other medical symptoms promptly. Use emergency services for an immediate threat. Routine screening should never delay urgent medical or safety action.

A fictional screening example

Maya is a fictional twenty-two-month-old whose family uses Spanish and English. Her parent chooses the tool's authorized Spanish version; a qualified interpreter supports the visit without changing items or scoring. The parent reports strong interest in music, several effective gestures, limited response to some sounds, and a recent change in words used. The initial result falls in the tool's follow-up range. This categorical result has no opportunity denominator and is not Maya's personal probability of being autistic.

The clinician completes the manual's required follow-up and records the initial and follow-up classifications, tool and version, respondent, language, interpreter role, access supports, unanswered items, and deviations. The clinician explains that neither stage diagnoses autism or explains the change in words or response to sounds. Referrals address developmental diagnostic evaluation and hearing, while the family contacts early intervention without waiting for a diagnosis.

The record keeps the screen, hearing question, developmental evaluation, and service eligibility as separate decisions. Maya's interests, gestures, family priorities, and existing supports remain part of every next step.