Evidence-based case conference

Diagnosing autism spectrum disorder (ASD) among adults

Clinical case

Patient: 18-year-old man

Present illness: The patient was brought to your department of psychiatry by his worried mother. While the patient sat still on the chair, directing an unmoving suspicious gaze towards you, his mother explained in a hurried, high-pitched voice, as follows: “Adam (the patient's name) has always been an odd boy. We've recently heard a lot about Asperger's syndrome and autism in the newspapers and on TV. I'm convinced that he suffers from them and told our family doctor, but he wasn't sure and sent us to you. Adam was very difficult to raise from the very beginning. Maybe he was a bit late in starting to walk or to talk but we did not think it was unusual, in comparison with his sister. When he went to the kindergarten and to primary school, he never had any friends. He always played alone and was interested in cartoon monsters. All he did was to collect monster cards. Things just had to be the way he wanted and he had tantrums when something unexpected happened. Oh yes, he also killed a lot of frogs with firecrackers—isn't this weird? Isn't he a weird boy? He was never good at school, probably always the worst or the second worst in class, and kids started bullying him. He started skipping class, but with understanding teachers he has so far managed to finish primary school and junior high school, and he is now a senior in high school. But he probably will now have to drop out: he has not gone to school for two months now. I am so worried. What will he do for his life?”

Present status: All this while Adam sat still, without changing his facial expression except for occasional grimaces. When asked what he himself thought was his problems, he murmured, “I don't know. But you know, I have recently taken a picture of E2. It is a newly-built bullet train for the Northeastern Railway, and runs at 250 km/h. Their previous E1 ran at 220 but the new one has a record speed of 345. But I also like slow trains, like models K1H2 and K1H202.” His mother intervened and explained that the train is now his pet subject and he spends hours trying to take pictures of trains passing by. When further asked what his school life was like, he explained that he did not like his peers because they always made fun of him and that he would not like to talk about them. He suddenly pointed to a train shown in a picture in your office and mentioned “Ah, that is the Venice Simplon-Orient-Express pulled by Ae 6/6 of the Swiss National Railway. Do you like it?”

You decided to order an intelligence test first. While entering the order into your computer, you now remembered to ask his mother, “By the way, has Adam been particularly sensitive to noises or smells or touch? Or did he have a high pain threshold?” “Yes, yes, how do you know that? He did not and does not like to be touched at all. It was very difficult to raise him.”

Formulate your clinical question

You make the next appointment for 6 weeks later, after the results of the intelligence test come back. You have to decide whether to send the patient for further work-up by a child psychiatrist, but you also remember that diagnostic criteria for autism and related disorders have recently changed in Diagnostic and Statistical Manual Fifth Edition (DSM-5), and you also would like to know what additional information you may need and be able to gather before sending every adult patient suspected of autism spectrum disorder (ASD) to your child psychiatrist colleague.

The clinical question to identify a screening tool for autism spectrum disorder could be formulated as follows:

  • Patients: Adults suspected of autism spectrum disorder

  • Screening test: Any standardised observer-rated scale or self-rating scale

  • Comparison: Gold standard diagnosis of autism spectrum disorder according to DSM-5

  • Outcomes: Stratum-specific likelihood ratios (or sensitivity and specificity)

Literature search

As always, you first search for a MeSH term corresponding broadly with autism spectrum disorder by simply entering ‘autism’ in the PubMed MeSH database (http://www.ncbi.nlm.nih.gov/mesh). It reveals that is does have a MeSH term ‘Autism Spectrum Disorder’, but looking at its tree you realise that ‘Child Development Disorders, Pervasive’, is a higher (broader) term that encompasses ‘Asperger Syndrome’, ‘Autism Spectrum Disorder’ and ‘Autistic Disorder’. (ASD in DSM-5 is a broader concept that roughly corresponds with pervasive developmental disorder in DSM-IV or International Classification of Diseases 10th Edition (ICD-10), and subsumes autistic disorder and Asperger's syndrome in DSM-IV.) So you choose ‘Child Development Disorders, Pervasive’ [Mesh], and use it in PubMed Clinical Queries (http://www.ncbi.nlm.nih.gov/pubmed/clinical) while specifying Category: diagnosis and Scope: narrow. This retrieves 387 references but this list could be further narrowed down to 97 when you apply the age limit to Adult: 19+ years. Browsing down the list, you immediately notice one study focusing on adults and using DSM-5: Grodberg D, Weinger PM, Halpern D, et al. The autism mental status exam: sensitivity and specificity using DSM-5 criteria for autism spectrum disorder in verbally fluent adults. J Autism Dev Disord 2014;44 :609–14.

The study examined the Autism Mental Status Exam (AMSE) against DSM-5 diagnoses among high-risk verbally fluent adults. AMSE aims to provide a standardised diagnostic screening approach that is feasible in all clinical settings. It has eight items assessing eye contact, interest in others, pointing skills, language, pragmatics of language, repetitive behaviours and stereotypy, unusual or encompassing preoccupations and unusual sensitivities. Each item is rated as 0 or 1 or 2, based on the clinical observation and/or self-reports and observer-reports. The developers of AMSE have prepared a web site where they provide four vignette cases with videos and scores, and two post-tests, for training purposes.

Critical appraisal of the literature on diagnostic test

ARE THE RESULTS AT RISK OF BIAS?

Did participating patients present a diagnostic dilemma?

YES. This study enrolled consecutive self-referred verbally fluent adult patients between the ages of 18 and 45 years who visited an autism research centre and consented to a comprehensive assessment protocol.

Did investigators compare the test to an appropriate, independent reference standard?

YES. The reference diagnosis was the best estimate clinical diagnosis according to DSM-5, based on the full Autism Diagnostic Observation Schedule (ADOS)1 (plus the Autism Diagnostic Interview-Revised (ADI-R)2 in one-third of the sample) and the psychiatric history. ADOS and ADI-R are considered to be the gold standard assessment tools for autism spectrum disorder.

Were those interpreting the test and reference standard blind to the other results?

YES. The AMSE was scored by a psychiatrist who first saw the patient. Then the ADOS and ADI-R were administered by an independent, blinded psychologist, and, finally, the best estimate clinical diagnosis was made by an independent psychiatrist based on all the information except for the AMSE score.

Did investigators perform the same reference standard to all patients regardless of the results of the test under investigation?

YES. Two-thirds of the patients did not receive ADI-R; however, ADOS is usually considered to provide enough information.

WHAT ARE THE RESULTS?

What likelihood ratios (LR) were associated with the range of possible test results?

Of the 50 enrolled patients, 23 were judged to have ASD according to DSM-5. The study reports sensitivity and specificity for each of the scores ranging from 1 through 9. When this table is converted into actual numbers according to the proposed cut-off,⇓

The positive likelihood ratio can then be calculated as 12.3 (95% CI 3.8 to 40.5), and the negative likelihood ratio as 0.09 (0.03 to 0.30), using the spreadsheet available on our web site (http://ebmh.med.kyoto-u.ac.jp/toolbox.html).

Since, in general, a likelihood ratio greater than 10 substantively increases the post-test probability of the diagnosis when the test is positive (ie, rules in the diagnosis), and that smaller than 0.1 substantively decreases the post-test probability when it is negative (ie, rules out the diagnosis), the performance of the AMSE appears to be very good.3

HOW CAN I APPLY THE RESULTS TO PATIENT CARE?

Will the reproducibility of the test result and its interpretation be satisfactory in my clinical setting?

This is probably the biggest issue for any clinician-rating scales. Although the tool has been developed to be feasible in all clinical settings, the developers of the AMSE explicitly warn that the scale be used only by clinicians with diagnostic expertise in ASD. They are currently examining the effectiveness of a training curriculum across different mental health professionals.

However, to the great advantage of the broader range of clinicians, the developers provide a web site explaining AMSE with case vignettes and videos (http://autismmentalstatusexam.com/). When I went through the materials, it took me two hours to train myself and to reach perfect agreement for the two post-tests.

Are the study results applicable to the patients in my practice?

As you are a general adult psychiatrist, your pre-test probability for ASD may be lower than that in specialised clinics such as the one in this study (46%).

Assuming the pre-test probability of 33% for Adam, then, because he would score 6 on AMSE (eye contact—0, interest in others—0, pointing skills—0, language—0, pragmatics of language—2, repetitive behaviours and stereotypy—1, unusual or encompassing preoccupations—2 and unusual sensitivities—1), his post-test probability is now elevated to 86%.

Will the test results change my management strategy?

YES.

Will patients be better off as a result of the test?

YES, VERY HOPEFULLY. The definitive diagnosis by a child psychiatrist, be it positive or negative, should decrease the mother's anxiety and, if positive, should lead to more specialised care fit for the patient.

•

What will you do with your patient?

The probability for Adam to have autistic spectrum disorder is quite high, and you decide to refer the patient to your colleague child psychiatrist for diagnostic work up and for further treatment.⇓

•
Archives of Evidence -Based Case Conferences published so far, according to clinical questions

  • Competing interests: TAF has received lecture fees from Eli Lilly, Janssen, Meiji, MSD, Otsuka, Pfizer and Tanabe-Mitsubishi, and consultancy fees from Sekisui Chemicals and Takeda Science Foundation. He has received royalties from Igaku-Shoin and Nihon Bunka Kagaku-sha publishers. He has received grants or research support from the Japanese Ministry of Education, Science and Technology, the Japanese Ministry of Health, Labour and Welfare, the Japan Society for the Promotion of Science, the Japan Foundation for Neuroscience and Mental Health, Mochida and Tanabe-Mitsubishi. He is diplomate of the Academy of Cognitive Therapy.

  • Provenance and peer review: Not commissioned; externally peer reviewed.

  1. close Rutter M, LeCouteur A, Lord C, et al. Autism diagnostic interview-revised. Los Angeles, Western Psychological Services 2003;
    Google Scholar
  2. close Lord C, Rutter M, DiLavore P, et al. Autism diagnostic observation schedule manual. Los Angeles, Western Psychological Services 1999;
    Google Scholar
  3. close Furukawa TA, Strauss S, Bucher HC, et al. Diagnostic tests, Users’ guides to the medical literature: a manual for evidence-based practice. New York, The McGraw-Hill Companies, Inc. 2014;
    Google Scholar

  • First published: 22 March 2016
  • Online issue publication: 21 April 2016

Article versions

Article metrics
Altmetricopen-url
Dimensionsopen-url