Diagnosis of autism

The diagnosis of autism is based on a person's reported and directly observed behavior. There are no known biomarkers for autism spectrum conditions that allow for a conclusive diagnosis.

In most cases, diagnostic criteria codified in the World Health Organization's International Classification of Diseases (ICD) or the American Psychiatric Association's Diagnostic and Statistical Manual of Mental Disorders (DSM) are used. These reference manuals are regularly updated based on advances in research, systematic evaluation of clinical experience, and healthcare considerations. Currently, the DSM-5 published in 2013 and the ICD-10 that came into effect in 1994 are used, with the latter in the process of being replaced by the ICD-11 that came into effect in 2022 and is now implemented by healthcare systems across the world. Which autism spectrum diagnoses can be made and which criteria are used depends on the local healthcare system's regulations.

According to the DSM-5-TR (2022), in order to receive a diagnosis of autism spectrum disorder, one must present with "persistent deficits in social communication and social interaction" and "restricted, repetitive patterns of behavior, interests, or activities." These behaviors must begin in early childhood and affect one's ability to perform everyday tasks. Furthermore, the symptoms must not be fully explainable by intellectual developmental disorder or global developmental delay.

Challenges
There are several factors that make autism spectrum disorder difficult to diagnose. First off, there are no standardized imaging, molecular or genetic tests that can be used to diagnose ASD. Additionally, there is a lot of variety in how ASD affects individuals. The behavioral manifestations of ASD depend on one's developmental stage, age of presentation, current support, and individual variability. Lastly, there are multiple conditions that may present similarly to autism spectrum disorder, including intellectual disability, hearing impairment, a specific language impairment such as Landau–Kleffner syndrome, ADHD, anxiety disorder, and psychotic disorders. Furthermore, the presence of autism can make it harder to diagnose coexisting psychiatric disorders such as depression. Diagnosing will be much harder in adults, since most people with ASD who reach adulthood undiagnosed, learn diverse (and often intense) masking techniques which make external diagnosis almost impossible.

DSM-5-TR criteria
The DSM-5-TR criteria lists 5 criteria (with examples) which include 2 groups of criteria (the first two):
 * Persistent impairments in social communication and interaction, characterized by difficulties in social-emotional exchange, nonverbal communication, and forming or understanding relationships.
 * Restricted, repetitive patterns of behavior, interests, or activities, as manifested by at least two of the following: repetitive actions or speech, strict adherence to routines, intense fixations, and unusual sensory responses.
 * Symptoms must be evident early in development, though they may only become noticeable when social demands exceed abilities or may be masked by learned coping strategies later in life.
 * Symptoms cause significant impairment in social, occupational, or other important areas of current functioning.
 * The symptoms are not better explained by intellectual developmental disorder or global developmental delay.

Diagnostic process
Ideally the diagnosis of ASD should be given by a team of clinicians (e.g. pediatricians, child psychiatrists, child neurologists) based on information provided from the affected individual, caregivers, other medical professionals and from direct observation. Evaluation of a child or adult for autism spectrum disorder typically starts with a pediatrician or primary care physician taking a developmental history and performing a physical exam. If warranted, the physician may refer the individual to an ASD specialist who will observe and assess cognitive, communication, family, and other factors using standardized tools, and taking into account any associated medical conditions. A pediatric neuropsychologist is often asked to assess behavior and cognitive skills, both to aid diagnosis and to help recommend educational interventions. Further workup may be performed after someone is diagnosed with ASD. This may include a clinical genetics evaluation particularly when other symptoms already suggest a genetic cause. Although up to 40% of ASD cases may be linked to genetic causes, it is not currently recommended to perform complete genetic testing on every individual who is diagnosed with ASD. Consensus guidelines for genetic testing in patients with ASD in the US and UK are limited to high-resolution chromosome and fragile X testing. Metabolic and neuroimaging tests are also not routinely performed for diagnosis of ASD.

The age at which ASD is diagnosed varies. Sometimes ASD can be diagnosed as early as 18 months, however, diagnosis of ASD before the age of two years may not be reliable. Diagnosis becomes increasingly stable over the first three years of life. For example, a one-year-old who meets diagnostic criteria for ASD is less likely than a three-year-old to continue to do so a few years later. Additionally, age of diagnosis may depend on the severity of ASD, with more severe forms of ASD more likely to be diagnosed at an earlier age. Issues with access to healthcare such as cost of appointments or delays in making appointments often lead to delays in the diagnosis of ASD. In the UK the National Autism Plan for Children recommends at most 30 weeks from first concern to completed diagnosis and assessment, though few cases are handled that quickly in practice. Lack of access to appropriate medical care, broadening diagnostic criteria and increased awareness surrounding ASD in recent years has resulted in an increased number of individuals receiving a diagnosis of ASD as adults. Diagnosis of ASD in adults poses unique challenges because it still relies on an accurate developmental history and because autistic adults sometimes learn coping strategies, known as "masking" or "camouflaging", which may make it more difficult to obtain a diagnosis.

The presentation and diagnosis of autism spectrum disorder may vary based on sex and gender identity. Most studies that have investigated the impact of gender on presentation and diagnosis of autism spectrum disorder have not differentiated between the impact of sex versus gender. There is some evidence that autistic women and girls tend to show less repetitive behavior and may engage in more camouflaging than autistic males. Camouflaging may include making oneself perform normative facial expressions and eye contact. Differences in behavioral presentation and gender-stereotypes may make it more challenging to diagnose autism spectrum disorder in a timely manner in females. A notable percentage of autistic females may be misdiagnosed, diagnosed after a considerable delay, or not diagnosed at all.

Considering the unique challenges in diagnosing ASD using behavioral and observational assessment, specific US practice parameters for its assessment were published by the American Academy of Neurology in the year 2000, the American Academy of Child and Adolescent Psychiatry in 1999, and a consensus panel with representation from various professional societies in 1999. The practice parameters outlined by these societies include an initial screening of children by general practitioners (i.e., "Level 1 screening") and for children who fail the initial screening, a comprehensive diagnostic assessment by experienced clinicians (i.e. "Level 2 evaluation"). Furthermore, it has been suggested that assessments of children with suspected ASD be evaluated within a developmental framework, include multiple informants (e.g., parents and teachers) from diverse contexts (e.g., home and school), and employ a multidisciplinary team of professionals (e.g., clinical psychologists, neuropsychologists, and psychiatrists).

, psychologists wait until a child showed initial evidence of ASD tendencies, then administer various psychological assessment tools to assess for ASD. Among these measurements, the Autism Diagnostic Interview-Revised (ADI-R) and the Autism Diagnostic Observation Schedule (ADOS) are considered the "gold standards" for assessing autistic children. The ADI-R is a semi-structured parent interview that probes for symptoms of autism by evaluating a child's current behavior and developmental history. The ADOS is a semi-structured interactive evaluation of ASD symptoms that is used to measure social and communication abilities by eliciting several opportunities for spontaneous behaviors (e.g., eye contact) in standardized context. Various other questionnaires (e.g., The Childhood Autism Rating Scale, Autism Treatment Evaluation Checklist) and tests of cognitive functioning (e.g., The Peabody Picture Vocabulary Test) are typically included in an ASD assessment battery. The diagnostic interview for social and communication disorders (DISCO) may also be used.

Screening
About half of parents of children with ASD notice their child's atypical behaviors by age 18 months, and about four-fifths notice by age 24 months. If a child does not meet any of the following milestones, it "is an absolute indication to proceed with further evaluations. Delay in referral for such testing may delay early diagnosis and treatment and affect the [child's] long-term outcome."


 * No response to name (or gazing with direct eye contact) by 6 months.
 * No babbling by 12 months.
 * No gesturing (pointing, waving, etc.) by 12 months.
 * No single words by 16 months.
 * No two-word (spontaneous, not just echolalic) phrases by 24 months.
 * Loss of any language or social skills, at any age.

The Japanese practice is to screen all children for ASD at 18 and 24 months, using autism-specific formal screening tests. In contrast, in the UK, children whose families or doctors recognize possible signs of autism are screened. It is not known which approach is more effective. The UK National Screening Committee does not recommend universal ASD screening in young children. Their main concerns includes higher chances of misdiagnosis at younger ages and lack of evidence of effectiveness of early interventions. There is no consensus between professional and expert bodies in the US on screening for autism in children younger than 3 years.

Screening tools include the Modified Checklist for Autism in Toddlers (M-CHAT), the Early Screening of Autistic Traits Questionnaire, and the First Year Inventory; initial data on M-CHAT and its predecessor, the Checklist for Autism in Toddlers (CHAT), on children aged 18–30 months suggests that it is best used in a clinical setting and that it has low sensitivity (many false-negatives) but good specificity (few false-positives). It may be more accurate to precede these tests with a broadband screener that does not distinguish ASD from other developmental disorders. Screening tools designed for one culture's norms for behaviors like eye contact may be inappropriate for a different culture. Although genetic screening for autism is generally still impractical, it can be considered in some cases, such as children with neurological symptoms and dysmorphic features.

Misdiagnosis
There is a significant level of misdiagnosis of autism in neurodevelopmentally typical children; 18–37% of children diagnosed with ASD eventually lose their diagnosis. This high rate of lost diagnosis cannot be accounted for by successful ASD treatment alone. The most common reason parents reported as the cause of lost ASD diagnosis was new information about the child (73.5%), such as a replacement diagnosis. Other reasons included a diagnosis given so the child could receive ASD treatment (24.2%), ASD treatment success or maturation (21%), and parents disagreeing with the initial diagnosis (1.9%).

Many of the children who were later found not to meet ASD diagnosis criteria then received diagnosis for another developmental disorder. Most common was ADHD, but other diagnoses included sensory disorders, anxiety, personality disorder, or learning disability. Neurodevelopment and psychiatric disorders that are commonly misdiagnosed as ASD include specific language impairment, social communication disorder, anxiety disorder, reactive attachment disorder, cognitive impairment, visual impairment, hearing loss and normal behavioral variation. Some behavioral variations that resemble autistic traits are repetitive behaviors, sensitivity to change in daily routines, focused interests, and toe-walking. These are considered normal behavioral variations when they do not cause impaired function. Boys are more likely to exhibit repetitive behaviors especially when excited, tired, bored, or stressed. Some ways of distinguishing typical behavioral variations from autistic behaviors are the ability of the child to suppress these behaviors and the absence of these behaviors during sleep.