Senior Medical Director, Clinical Adoption Cognoa, Inc. Mission Viejo, California, United States
Full Description:
Background: The lack of diagnostic tools for Autism Spectrum Disorder (ASD) in primary care settings and long wait lists for specialist assessment contribute to an average delay of 3 years between first parental concern and diagnosis. Delays are even longer for children who are non-white, female, of lower socioeconomic status, or rural residing. A diagnosis aid for primary care physician use which harnesses the power of artificial intelligence (AI) to aid primary care physicians in the identification of ASD has the potential to reduce the lag between screening and diagnosis, possibly enabling timelier ASD-specific interventions during the critical period in the child's neurodevelopment.
Objectives: This study examined the performance of an artificial intelligence-based Software as a Medical Device intended to aid primary care providers (PCPs) in the diagnosis of ASD.
Methods: This was a prospective multi-site pivotal study conducted in 6 states using a double-blind active comparator design with 425 completed subjects (36% female) ages 18-72 months with concern for developmental delay. Previous research developed, tuned, and tested a device that uses gradient boosted decision trees machine learning algorithm which analyzes behavioral features from 3 distinct inputs: 1) A caregiver questionnaire 2) two, 90 second minimum home videos analyzed by trained video analysts 3) A health care provider questionnaire. Device results were compared to diagnosis by independent agreement of specialist clinicians based on clinical assessment, including a modified CARS-2 and DSM-5 criteria. Specialists were child psychiatrists, child psychologists, pediatric neurologists, and developmental behavioral pediatricians experienced in diagnosing ASD.
Results: The Device rendered a determinate output (ASD positive or negative) for 31.8% of participants. Comparison of device results to specialist diagnosis found the PPV: 80.8% [95%CI, 70.3%-88.8%], NPV: 98.3% [90.6%-100%], sensitivity: 98.4% [91.6%-100%], specificity: 78.9% [67.6%-87.7%] for subjects with determinate (ASD positive or negative) device results. No significant differences in Device performance were found across participants’ sex, race/ethnicity, income, or education level. No evidence of performance degradation was found when PCP assessments were performed remotely versus in-person.
Conclusions: Using this device, PCPs can potentially efficiently, accurately, and equitably diagnose a subset of children 18-72 months old, thereby streamlining specialist referrals and facilitating earlier ASD diagnosis and interventions. The results further provide preliminary evidence that PCP evaluation of the child can be done via telemedicine or in-person with no degradation in device performance.
Abbreviated Description: Lack of diagnostic tools for Autism Spectrum Disorder(ASD) in primary care settings and long wait lists for specialist assessment contribute to an average delay of 3 years between first parental concern and diagnosis. This prospective multi-site double-blind active comparator study compared the performance of an artificial intelligence-based device intended to aid primary care providers (PCPs) diagnose ASD, to specialist clinician diagnosis. Participants were 18-72-month-olds with developmental delay concern(425 completers). Comparison of device results to specialist diagnosis found the PPV: 80.8%, NPV: 98.3%, sensitivity: 98.4%, specificity: 78.9% for subjects with determinate device results. Device use could potentially expand timely, accurate and equitable ASD diagnosis in primary care.
*CONGRATULATIONS! You have found 1 of the 4 participating Posters*