Alex is 6 years old who was evaluated at Emerge for mood, social interaction, sensory, and behavioral challenges. He scored in the Superior range of intellectual ability on the WISC-IV and was reported to be bright, clever, and kind to others in class. Alex does not like it when his first grade classmates break rules and he wants to play the same pretend game “puppies” on the playground day after day. Alex has huge meltdowns at home over small things like disagreements with his sister. He is a bed wetter, seemingly unaware when he has to “go” and thus he often does not make it to the bathroom on time (Table 9.1).

Alex demonstrated an unusual style of communication and social interaction throughout the ADOS-2 and during other parts of testing. His speech was occasionally formal and advanced, and his utterances were sometimes characterized by an odd use of words or phrases. Though he occasionally offered information to the examiner during conversations, what he disclosed appeared to be blatant fabrications or events that were unlikely to be true. It seemed that Alex was unsure of how to respond to the examiner, and he dealt with this uncertainty by making up stories. For instance, he told stories about his dog catching ten squirrels at once, being on Fox News, and having a pet tiger. When asked follow-up questions about his pet tiger, he said the tiger is currently one inch long and is about to “turn two inches.” He seemed to lack any awareness of how the information he shared would be received by the examiner (i.e., that she would know he was being untruthful).

Alex occasionally asked the examiner about her ideas or experiences; however, he dropped several of the examiner’s conversational bids for him to ask questions and simply ignored her comments. Alex exhibited good eye contact and he directed some facial expressions appropriately at the examiner such as raising his eyebrows and smiling. Alex demonstrated extremely limited insight into typical social situations and relationships. For instance, he could not identify any of his friends’ names, and he shrugged when asked, “What makes someone a friend?” or “How is a friend different from someone whom you just go to school with?” Alex’s difficulty answering these questions is not commensurate with his strong verbal abilities, as evidenced by his performance on the WISC-IV.

Alex’s play was typified by unusual sensory interests and a highly specific focus on certain aspects of the play characters. For instance, when asked to play with several different objects, Alex first picked up a piece of cotton, held it in his hand, and said, “It’s a fluffy rock.” He also laid down the human figures, covered up their faces with plates, and said, “I don’t want to see their bald eyes.” Alex was diagnosed with ASD; and through the cognitive behavioral treatment that followed, Alex began to work on social perspective taking skills. He is becoming less bossy with his peers and his tantrums are reduced in frequency because his emotional awareness and coping skills have improved .

A review of the research indicates that the Autism Diagnostic Observation Schedule, Second Edition (ADOS : Lord & Schopler, 1989) and the Autism Diagnostic Interview-Revised (ADI-R : Lord, Rutter, & Le Couteur, 1994) are the tools generally recommended for autism evaluations and research (Tanguay, 2000). The ADOS-2 is an observation scale whereby a clinician assesses a child’s play and communicative behaviors through a naturalistic observation, using a defined set of activities and assessment criteria (approximately 14 activities, 26 items, and 3 domains). The ADI-R is a semi-structured interview, administered to parents (over 70 items, summarized into an algorithm that includes three domains/dimensions). Some studies name the ADOS as the “gold standard” (Klin et al., 2007) for diagnostic assessment of autism spectrum disorders. However, others refer to the ADI-R as the gold standard (Matson, Nebel-Shcwalm, & Matson, 2007). Matson and colleagues claim, “The ADI-R is considered to be the gold standard for assessment scales used in the diagnosis of autism by many at this time. It has a broader age range of norms, more published psychometric data, and is a best fit with DSM and ICD criteria relative to other scales…” (Matson et al., 2007, p. 43).

Finally, other studies suggest the use of both instruments as the gold standard. A comprehensive literature review conducted in 2013 by Falkmer, Anderson, Falkmer, and Horlin found that “…the true ‘gold standard’ classification and diagnosis of autism is still considered to be multi-disciplinary team clinical assessment , including use of the ADOS and ADI-R, as well as other assessments with consensus clinical judgment” (Falkmer, Anderson, Falkmer, & Horlin, 2013, published online no page number). Research on the usefulness of the ADOS and ADI-R in populations with co-occurring intellectual disabilities shows that these measures are valuable diagnostic tools (Sappok et al., 2013). Robertson and colleagues report that the ADOS and ADI-R measure slightly different aspects of manifestations of the disorder and the combination is a conservative approach to diagnose Autism Spectrum Disorders (Robertson, Tanguay, L’Ecuyer, Sims, & Waltrip, 1999). Another study focused on using these measures to assess preschool children and found good agreement between the two instruments (Dover & Le Couteur 2007). Research on the use of both instruments found approximately 75 % agreement between the ADI-R and ADOS (in 2006) (Dover & Le Couteur 2007). The authors of this text agree with the research claiming that the ADOS-2 and the ADI-R are both considered gold standard instruments, but argue that the purpose and usefulness of the instruments can be differentiated.

In looking at the utility of both the ADOS and ADI-R, both considered Gold Standard instruments, it is useful to consider sensitivity and specificity. For clarification, specificity measures the proportion of negatives (no diagnosis) correctly identified as such and sensitivity measures the proportion of actual positives (those with the diagnosis) correctly identified as such. Sappok and Diefenbacher’s findings indicate that the ADOS is a very sensitive measure with 100 % sensitivity but is lower in differentiating between diagnoses (45 % specificity) (Sappok et al., 2013). The study also utilized the ADI-R which had 88 % sensitivity and 80 % specificity. In the intellectually disabled population, it was more feasible to do the ADOS and less so to do the ADI-R, based on the number of caregivers available to be interviewed.

Studies have examined the relative practicality of utilizing both measures during an autism assessment. Both Filipek et al. (1999) and De Bildt et al. (2004) noted that the time to train on and administer both measures is extensive. In these authors’ experience, practicality is often discovered as the reason why some clinicians do not utilize both instruments consistently during diagnostic evaluations. Le Couteur et al. (2007) notes that for clinical and research settings the Autism Spectrum has broadened and a clear assessment framework is needed. As previously mentioned, the ADOS-2 offers observational data, whereas the ADI-R offers diagnostic interview information. Both interviews and observations have advantages and limitations . One advantage of using both is to “gather clinically relevant information in a systematic and comparable fashion.” These authors advance that the ADOS-2 is mandatory to include; however, most ADI-R data can be obtained through a comprehensive diagnostic interview meeting with parents. In more involved cases with diagnostic uncertainty, it is valuable to record behavior with both the ADOS-2 and ADI-R (Le Couteur et al., 2007).

According to the ADOS-2 manual, the first version of the ADOS was introduced in 1989 and intended for administration with children ages 5–12 with expressive language at least at a 3-year-old level. This measure was proposed to complement the Autism Diagnostic Interview, known as the ADI (Le Coutuer et al., 1989). The combination of the two measures was offered as having adequate reliability and validity for research diagnosis. For administration of the original ADOS, specific training was required. According to the manual, after the ADOS was released, it became evident that there was a need for a version of the measure appropriate for younger children and those with less language. The PL-ADOS was developed with this population in mind. At that time, the ADOS authors were participating in a longitudinal study that referred children at age two and further necessitated the development of this measure. The PL-ADOS, which came out in 1995, was more flexible and included more use of play materials (DiLavore, Lord, & Rutter, 1995). The ADI-R was released including more questions about social interaction and communication for younger children (Lord et al., 1994) making it a better companion measure.

The PL-ADOS was useful for assessing younger children with very limited speech but researchers found it still excluded two subsets of individuals. It did not identify young children with autism who also had some expressive language skills. It also failed to provide a method to assess and identify older adolescents and adults with autism. Test developers set to resolve these issues and the ADOS-G was developed. The ADOS-G is what we now know of as the ADOS. This version of the ADOS included Module 1 based on the PL-ADOS and Module 3 based on the original ADOS (1989). It included a “new” Module 2 developed for “phrase speakers” in the preschool age range and a Module 4 for assessing adolescents and adults. This module was less play based and included more questions about daily living, personal responsibility, and job skills. With the development of the ADOS-2, the original algorithms for Modules 1–3 were revised; a comparison score was created for Modules 1–3, and the ADOS protocols were improved to clarify administration guidelines. A fifth Module, the Toddler Module, was created with associated algorithms designed to assess children with limited expressive language who are aged from 12 to 30 months (Lord et al., 2012). For detailed psychometric data on reliability and validity of the measure see Appendix.

The ADOS-2 was originally released in May of 2012. It contains five assessment modules to evaluate varying developmental levels at different chronological ages. The modules assess skills in communication, reciprocal social interaction, and restricted and repetitive behaviors or interests. The ADOS-2 is a semi-structured measure including open-ended and creative tasks. The goal of the revision was to make the algorithm fit with continued research and provide increased specificity in lower functioning populations, as well as provide a uniform basis of comparison and maintain the strong predictive validity of the measure (Lord et al., 2012). Research findings indicated a need for a unitary social-communication factor (Constantino et al., 2004; Lord, Rutter, DiLavore, & Risi, 1999, 2000; Robertson, Tanguay, L’Ecuyer, Sims, & Waltrip, 1999). Modules 1–3 of the ADOS-2 include “social affect” or SA which is a composite of social and communication skills, and “restricted repetitive behavior” or RRB. The Module 4 algorithm was not revised because of limited available samples of older adolescents and adults, an area needing further study (Lord et al., 2012). Research on the ADOS Module 1 found it to be over-inclusive of those with a mental age of 15 months or below (Gotham et al., 2007) and thus the Toddler Module was developed for those 12–30 months old with play activities, codes, and an algorithm that is better aligned with this population.

In order to administer the ADOS-2, a clinician must be appropriately trained in assessment (i.e., have licensure and credentialing in psychology or medicine), have adequate practice with administration, and read the manuals thoroughly. Additionally, he or she must attend a workshop with an independent trainer or complete the video training program. It is crucial that a clinician has extensive experience in giving the measure and scoring the algorithm and this can take varying amounts of time depending on the clinician. In the authors’ clinical practice, administrators of the ADOS-2 should have completed the training and have co-scored the measure watching a reliable clinician and consistently achieving at least 70 % agreement on the algorithm overall. Additionally, it is required that trainees achieve the same diagnostic conclusion, with 100 % agreement, from the algorithm. This standard is consistent with what Dr. Lord’s team recommended for clinical practice at the ADOS training workshop (WPS, Ann Arbor, Michigan, June 2011). Those trained in the ADOS do not need an additional training for ADOS-2, only a thorough review of the manual, and completion of the ADOS-2 training video program for the Toddler Module. For research reliability, an individual must attend a research reliability workshop and achieve a greater interrater reliability than what is required for clinical work (WPS ADOS training workshop Ann Arbor Michigan, 2011).

An important benefit of ADOS-2 administration is that, in addition to providing diagnostic information, it is excellent for guiding treatment planning. The ADOS-2 provides valuable information on pretend play, joint attention, flexibility, conversation and communication with others, restricted interests, unusual sensory interests, emotional awareness, and emotion regulation. Those with ASDs often have much more difficulty and sometimes present with resistance to ADOS-2 activities versus concrete cognitive or language testing. It is interesting to note whether a child preferred the ADOS-2 or other assessment measures. When introducing pretend play with toys, acting or demonstrating tasks, and conversations, often those who do not have ASDs are more excited or engaged with these measures than they are with cognitive measures.

In contrast, individuals with ASD may refuse to play with toys, citing their preference for video games, sports, books, etc. They often make excuses to avoid “acting” tasks; or they may say that they cannot do it or are too embarrassed. At other times, individuals with ASD might engage with the toys or books but have play skills that are not commensurate with cognitive and language abilities.

They may make up a story in a picture book but with no plot, character development or emotional response. In the authors’ experience, and consistent with past research, their responses on the storybook task alone can offer rich diagnostic data. Children with ASD often struggle to tell stories, fail to identify characters, struggle to describe the sequence of events, and cannot provide the main idea of the story (Suh et al., 2014; Willard, 2013) Some studies have found that children with autism actually perform similarly to neurotypical children when telling stories from picture books, but struggle with more challenging tasks requiring narrative recall (Losh & Gordon, 2014). Children with optimal outcomes may actually show improved narrative production over time, sometimes similar to neurotypical children. Thus, examiners should carefully consider the child’s narrative quality during the storybook task of the ADOS-2, particularly over repeated administrations.

Discussion of friendships and other relationships may be less sophisticated than expected or a child might often respond “I don’t know” or “I can’t remember.” Readministering the ADOS-2 periodically can provide nice data on progress, as goals for treatment may be improving pretend play skills, conversation or joint attention; the ADOS-2 can provide a measure of skill development over time. The authors recommend reevaluations every 1–2 years to provide progress data and direct treatment.