<div>This systematic review with meta-analysis investigates the accuracy of index tests for diagnosing autism in children and adolescents.</div>
Italian Ministry of Health
From database inception to March 2023
<div>Cross-sectional and case-control studies</div>
32
<div>Structured tools showed the following pooled sensitivities and specificities:</div>
<div>
<ul>
<li>Autism Diagnostic Interview-Revised (ADI-R)=0.60 and 0.84;</li>
<li>Autism Diagnostic Observation Schedule (ADOS)=0.91 and 0.74;</li>
<li>ADOS-2=0.90 and 0.69;</li>
<li>Autism Spectrum Rating Scales (ASRS)=0.66 and 0.86;</li>
<li>Childhood Autism Rating Scale (CARS)=0.85 and 0.94;</li>
<li>Checklist for Autism Spectrum Disorder (CASD)=0.91 and 0.98;</li>
<li>Diagnostic Interview for Social and Communication Disorder (DISCO)=0.96 and 0.50; and</li>
<li>Gilliam Autism Rating Scale (GARS)=0.50 and 0.78.</li>
</ul>
</div>
<div>The authors conclude that no single diagnostic tool provides enough information. They highlight that standardized tests, observation from a multidisciplinary team, and clinical judgment complement one another during the autism diagnosis process.</div>
<div> </div>
<div>Limitations of this review include the lack of data from existing test manuals and a lack of studies on the application of new technologies (e.g., artificial intelligence, machine learning, sensor-based systems). Additional research is needed to investigate the applicability of test manual information across contexts, the accuracy of combining tools, and the reliability of new technology to diagnose autism in children and adolescents.</div>
