Improving Accuracy and Accessibility of Early Autism Screening

DESCRIPTION (provided by applicant): The aims of this proposal are to complete development and evaluation of a computerized adaptation of a new tool to more accurately identify toddlers with signs of Autism Spectrum Disorders (ASD) and to make this tool accessible to varied populations. Early screening for ASD has recently become standard for pediatric practice following evidence of the efficacy of early intervention. The most widely recommended screen is the Modified Checklist for Autism in Toddlers (MCHAT; Robins, et. al., 1999). While the MCHAT has good psychometric characteristics, a major obstacle to its use in primary care is its large over-referral rate. While a structured interview (MCHAT Follow Up; Kleinman, et. al., 2007) was developed that reduces over-referral, it is impractical in its original form for brief well child visits. The platform for this project is a web-based decision support system, CHADIS that delivers pre-visit questionnaires such as the MCHAT for completion online, scores them and presents results to the clinician. CHADIS also provides decision support in an electronic textbook linked to screening results and resources in handouts for parents. A computerized version of the MCHAT Follow Up completed by clinician interview has been created in CHADIS (CHADIS MCHAT F/U). CHADIS MCHAT F/U is practical because it prompts doctors to ask the parents only a few questions to follow up on the answers causing the positive MCHAT score, with the test automatically rescored after each clinician input until a positive or negative result is obtained. It is anticipated that the computerized MCHAT and CHADIS MCHAT F/U will cut the over-referral rate by more than half, saving costs and parent agony. CHADIS is being used nationally and in Maryland for a sizable network of practices in an autism screening program. However, use in settings with large numbers of low income families has been limited because they are unable or do not remember to complete questionnaires online prior to visits. Phase I of this project will involve development of waiting room alternatives for administering the MCHAT including Interactive Voice Response (IVR) phone system, touch screen kiosk, and laptop computer that will allow results to be linked to CHADIS MCHAT F/U. A reminder system will also be created. These methods will then be assessed in parent and physician focus groups in 3 populations. Phase II will involve: assessing satisfaction with the waiting room methods in 160 parents in the 3 populations, and testing positive predictive validity of the CHADIS MCHAT F/U method. Validity will be determined by screening 4500 children in the Maryland network using the MCHAT, and completing the CHADIS MCHAT F/U and complete evaluations for ASD on the expected 300 positive children. Finally, relative costs and savings of these methods will be estimated. It is hypothesized that the new system will show advantages, savings, and acceptability that support widespread Phase III dissemination. PUBLIC HEALTH RELEVANCE: The aims of this proposal are to complete development and evaluation of a computerized adaptation of a new tool to more accurately identify toddlers with signs of Autism Spectrum Disorders (ASD), now affecting 1 in 150 children, and to make this tool accessible to varied populations. Early screening for ASD is a new standard for pediatric practice but the most widely recommended screen, MCHAT, has a large over- referral rate and low income families are less able to or do not remember to complete it prior to their visits. This project will create methods for using the MCHAT in the waiting room by phone (IVR), touch screen kiosk and laptop and a phone reminder system; pilot them for acceptability; then measure preference, satisfaction and costs of these methods in 3 populations. A computerized version of MCHAT and of a Follow Up prompted interview designed to reduce over-referral has been created in an existing web-based decision support system (CHADIS) and will be validated by evaluating 300 children for ASD who have positive MCHAT screens regardless of their MCHAT Follow Up results. It is anticipated that the computerized MCHAT and CHADIS MCHAT F/U will improve rates of valid screening for ASD, important to early identification and intervention, but cut the over-referral rate by more than half, saving costs and parent agony.