A Systematic Review of Telehealth Screening, Assessment, and Diagnosis of Autism Spectrum Disorder
Child and Adolescent Psychiatry and Mental Health
Liu, M., & Ma, Z. (2022).
Child and Adolescent Psychiatry and Mental Health, 16(1), 79.
This systematic review explores the diagnostic accuracy of telepractice screening and evaluation tools used to identify individuals on the autism spectrum.
No funding received
2000 to June 2022
Original, peer-reviewed research with empirical study designs
26
<div>Generally, telepractice screening applications used to identify children on the autism spectrum had high sensitivity and low specificity. Specific screening apps demonstrated the following psychometrics: </div>
<ul>
<li><span style="color: #333333;"><strong>ASDTests</strong>: sensitivity = 0.92-0.98; specificity = 0.85-0.98; </span></li>
<li><span style="color: #333333;"><strong>AutismAI</strong>: sensitivity = 0.96, specificity = 0.99; </span></li>
<li><span style="color: #333333;"><strong>Autism&Beyond</strong>: sensitivity = 0.92-0.98, specificity = 0.38-1.0; </span></li>
<li><span style="color: #333333;"><strong>Cognoa</strong>: sensitivity = 0.75-0.98, specificity = 0.62-0.79, positive predictive value (PPV) = 0.81-0.83, negative predictive value (NPV) = 0.5-0.98;</span></li>
<li><span style="color: #333333;"><strong>Language Environment Analysis System (Lena)</strong>: sensitivity = 0.75, specificity = 0.98; </span></li>
<li><span style="color: #333333;"><strong>Systematic Observation of Red Flags (SORF)</strong>: sensitivity = 0.70, specificity = 0.67, PPV = 0.55, NPV = 0.79; </span></li>
<li><span style="color: #333333;"><strong>Video-Referenced Infant Rating System for Autism (VIRSA)</strong>: sensitivity = 1, specificity = 0.53, PPV = 1, NPV = 0.19; and </span></li>
<li><span style="color: #333333;"><strong>Wingfield’s mobile application</strong>: sensitivity = 0.88, specificity= 0.96. </span></li>
</ul>
<div><span style="color: #333333;">The Telehealth Evaluation of Development for Infants (TEDI) and the Child Health Improvement through Computer Automation (CHICA) did not have any reported psychometrics. Of note, there was an overall low number of studies (i.e., one to four studies) investigating each tool. Additional limitation to this review included limited availability of participant characteristics and small sample sizes within some of the included studies.</span></div>
<div>This review reported the following psychometrics for telepractice diagnostic applications designed to identify children on the autism spectrum:</div>
<ul>
<li><span style="color: #333333;"><strong>Brief Observation of Symptoms of Autism (BOSA)</strong>: sensitivity = 0.86-0.96, specificity = 0.70-1.00; </span></li>
<li><span style="color: #333333;"><strong>the NODA application</strong>: sensitivity = 0.85, specificity = 0.94; and </span></li>
<li><span style="color: #333333;"><strong>real-time video conferencing</strong>: sensitivity = 0.79-0.84, specificity = 0.88. </span></li>
</ul>
<div><span style="color: #333333;">The TeleNP and the Tele-ASD-PEDs did not have any reported psychometrics. </span><span style="color: #333333;">Further research is needed to determine the overall diagnostic accuracy of these tools due to a limited number of included studies and predominately male samples (72.8%-80%).</span></div>
