- About the Center
- Evidence-Based Practices
- Early Identification of ASD Module
- Manual of Procedures (MOP)
- Additional Resources
- News and Events
- Working With States
- Early Identification Processes
Module Menu >>
Module: Early Identification of Autism Spectrum Disorders
EARLY IDENTIFICATION PROCESSES
Surveillance, Screening and Diagnosis for ASD
As Shattuck et al. (2009) recently suggested, there is considerable need for further research, innovation, and improvement in early identification, given the gap between the age when ASD characteristics are first noticed and when a diagnosis is made. There are several levels of recommended practices that lead to the early identification of an ASD. Surveillance is the ongoing process of identifying children at risk. Surveillance methods are often simply questions that primary care providers ask all parents at well-child visits about developmental milestones. Screening is a brief assessment that uses standardized instruments to identify children at significant risk of having a disorder. The results of screening indicate the presence of risk characteristics for delay or disability, but do not provide a diagnosis. Diagnostic assessment involves a detailed assessment process using standardized tools, observations, and clinical history to define the child’s difficulties and determine the presence of a clinical diagnosis. A diagnostic assessment is the final step in early identification of a specific disorder.
The American Academy of Pediatrics endorsed new practice parameters for ASD screening and evaluation to occur with primary care pediatric practices as summarized in Identification and Evaluation of Children with Autism Spectrum Disorders (Johnson et al., 2007). There are several new early identification guidelines laid out in this document, which represents a significant new milestone for the autism field. Among the most important recommendations are the following:
These autism specific refinements follow previously established parameters for the universal screening of all children for potential developmental delays at 9, 18 and 24 (or 30) months during well-child visits. To view this surveillance and screening algorithm, go to the following website: http://www.medicalhomeinfo.org/health/autism.html
Surveillance, Screening and Diagnosis for ASD
The primary purpose for gathering Information obtained through screening and surveillance is to help practitioners determine whether or not a referral for diagnostic assessment for ASD is warranted. Surveillance begins with an understanding of child development and may be conducted by any professional who has contact with a child. However, surveillance is routinely conducted during well-child visits by the primary care physician or other office professional.
The following behaviors have been suggested by the American Academy of Pediatrics (2006) and others as red flags for ASD in the first two years of life:
The ASD Video Glossary is an excellent, no cost, web-based tool that is designed to demonstrate the subtle differences between typical and delayed development in young children and spot the early red flags for ASD. There are over one hundred video clips that feature children who are typically developing and children diagnosed with an ASD.
Screening for Autism Spectrum Disorders
Universal versus targeted screening. There are two levels of screening – population-based/universal screening and targeted screening. Population-based or universal screening is designed to evaluate all children and to identify those at risk for developmental differences. Tools that may be used for this type of screening include the Ages and Stages Questionnaire (Squires et al., 2002), the Infant-Toddler Checklist (Wetherby, & Prizant, 2002), the PEDS (Glascoe,1998), and the Denver Developmental Screening Test – II (Frankenburg, 1992). Many pediatricians conduct routine universal developmental screening during well-child visits to detect delays in various domains of development (e.g., language, motor skills). These population-based screening procedures are general tools used to identify a wide range of developmental problems (as well as children who do not end up having any difficulties). These general screeners are useful because they are (or should be) administered universally, to all children, and will identify those who need more targeted, specific screening for autism spectrum disorders.
In contrast, targeted screening focuses on a specific group of children (such as high risk children or those who may be showing differences in their development and behavior). Targeted screening uses instruments that examine specific behaviors often associated with ASD in young children, such as social interest, pointing, and eye contact. These instruments are summarized in Table 1 in the Resource section of this module.
Recommended Ages for Screening
The ideal age to screen infants and toddlers for autism is still debated. With the current state of our knowledge about how and when signs of autism emerge, screening should occur repeatedly during the first three years of life, since onset can happen as late as 36 months. Screening for ASD much before 18 months is currently not feasible. There are no validated tools for children much younger than 18 months. Signs of autism may be quite subtle before the first birthday and easier to identify retrospectively than to “see” as they are happening. As many as a third to half of children who develop autism did not show any signs until symptom onset or regression in the middle of the second year of life. Therefore, screening prior to 18 months will miss many cases. As the AAP recommended in their new practice parameters (Johnson et al., 2007), universal screening at both 18 and 24 months should cast the broadest net for identification, and repeated screenings at 30 and 36 months should identify most children at risk for ASD.
Selection of a screening tool must be made carefully and with consideration of its psychometric properties. Psychometric properties include the tool’s reliability and validity, as well as its sensitivity and specificity. The reliability of an instrument reflects the consistency of the measure in obtaining similar results each time it is used under the same condition with the same person. The validity of an instrument informs the strength of the conclusions drawn from the findings, offering a check on the accuracy of a finding. The sensitivity of a tool refers to its ability to correctly identify those children who need additional evaluation; specificity refers to its ability to accurately identify those who do not need further evaluation. Most screening measures have better sensitivity than specificity, as their purpose is casting a wide net to identify anyone who might be at risk. Thus, it is acceptable to have some “false positive” screens or over-identification. These children will be identified as unaffected or at low risk after further screening or evaluation. It is important to convey this possibility to parents, who may be alarmed when a child who they had no concerns about is identified as potentially at risk. Potential benefits of early identification (e.g., early treatment) can be explained to parents, while also assuring them that a more complete evaluation will clarify whether the screening results are correct in identifying risk.
A summary of commonly used screening measures with acceptable psychometric properties is provided in Table 1.
Conducting a Diagnostic Assessment
The purpose of the diagnostic evaluation is to confirm whether or not a child with any of the early signs and risk factors for an autism spectrum disorder and/or those with concerning results from a screening tool would meet the criteria for an autism diagnosis. Diagnostic evaluations should include three central components: gold standard diagnostic measures, parental history, and direct observation; all are necessary for a comprehensive picture of the child’s functioning. The first step of the core assessment process is to review with parents the child’s early developmental history and current concerns. The critical aspects of this history-taking are reviews of communication, social, health, motor, and behavioral development; additionally, brief screening of potential medical issues should be conducted at this stage in order to determine the need for more detailed evaluation (possibly including referral to specialists). A review of available records rounds out the history-taking aspect of the evaluation. Combined with this review is direct observation of and interaction with the child. Specific instruments that elicit signs of autism, such as those in Table 2, are recommended to be used in addition to informal observation. Finally, assessment of cognitive, language, and adaptive skills is an important part of a diagnostic evaluation of infants and toddlers with ASD.
In contrast to screening measures, diagnostic instruments are expected to have both high sensitivity and specificity. These properties of the instrument, particularly sensitivity, are difficult to know for newly developed instruments because long-term follow up to determine how many cases were missed by the initial evaluation is necessary to determine sensitivity. A summary of commonly used diagnostic measures with good psychometric properties is provided in Table 2 of the Resources section of this module.
Diagnostic assessments are typically completed by developmental pediatricians, psychologists, psychiatrists, or neuropsychologists with the support of other multidisciplinary team members, such as occupational therapists and speech and language therapists. Many diagnostic assessments occur in specialty clinics for which there may be long waiting lists. With the increases in the prevalence of ASD, there is a need for more accessible community-based diagnostic resources, especially to address potential disparities in access to diagnostic services due to race, ethnicity, or socio –economic status. Many school districts and birth–to-three agencies have professionals who are fully equipped to diagnose autism as part of the assessment services provided by public agencies.
No one tool or procedure “diagnoses” autism. Information from autism specific instruments and other developmental assessments provides evidence to inform clinical judgments as to whether or not a child meets the DSM-IV criteria for ASD. Diagnosis is a clinical process, and it is always made by a skilled clinician, who diagnoses autism based on their expert integration of all the information they have obtained about the child, not on any one score or behavior.
Educational Eligibility and Clinical Assessment
Beginning at age three when a child is referred for special education, an educational classification of autism may be used to determine eligibility. The educational definition is compatible with DSM-IV criteria for ASD and does not require a child have a clinical diagnosis of autism. Sometimes, having a clinical diagnosis of ASD from a medical professional is not enough to qualify children for special educational services. It is quite typical for a team of qualified school personnel to conduct educational evaluations to determine a child’s educational disability in these developmental areas: cognition, speech-language, adaptive behavior, and social-emotional and behavior. In addition, vision and hearing is screened and the child’s developmental and health history is gathered. Observations are made in school, home and community settings. This process would provide the information needed for the Individualized Family Service Plan (IFSP) or Individualized Education Plan (IEP) team to deem a child eligible for early intervention or special education services.