Accuracy of phenotyping of autistic children based on Internet implemented parent report.
ABSTRACT While strong familial evidence supports a substantial genetic contribution to the etiology of autism spectrum disorders (ASD), specific genetic abnormalities have been identified in only a small minority of all cases. In order to comprehensively delineate the genetic components of autism including the identification of rare and common variants, overall sample sizes an order of magnitude larger than those currently under study are critically needed. This will require rapid and scalable subject assessment paradigms that obviate clinic-based time-intensive behavioral phenotyping, which is a rate-limiting step. Here, we test the accuracy of a web-based approach to autism phenotyping implemented within the Interactive Autism Network (IAN). Families who were registered with the IAN and resided near one of the three study sites were eligible for the study. One hundred seven children ascertained from this pool who were verbal, age 4-17 years, and had Social Communication Questionnaire (SCQ) scores > or =12 (a profile that characterizes a majority of ASD-affected children in IAN) underwent a clinical confirmation battery. One hundred five of the 107 children were ASD positive (98%) by clinician's best estimate. One hundred four of these individuals (99%) were ASD positive by developmental history using the Autism Diagnostic Interview-Revised (ADI-R) and 97 (93%) were positive for ASD by developmental history and direct observational assessment (Autism Diagnostic Observational Schedule or expert clinician observation). These data support the reliability and feasibility of the IAN-implemented parent-report paradigms for the ascertainment of clinical ASD for large-scale genetic research.
- [Show abstract] [Hide abstract]
ABSTRACT: Objective: Little is known about accessibility to health care transition (HCT) services (HCT) for youth with autism spectrum disorders (ASD). This study examined how often youth with ASD receive HCT services and how access varied by individual, family, and health system characteristics. Method: Questionnaires were completed by 101 parents of youth with ASD (ages 12-17 years) enrolled in a national online autism registry. Descriptive statistics and bivariate analysis were used to examine a composite HCT variable and its components. Results: Fewer than 15% of youth received HCT services. Although 41% received at least 1 HCT discussion, only 3% received all 3. One-quarter had a discussion with their health care provider about transitioning to an adult provider, adult health care needs, or insurance retention, and 31% of providers encouraged youth to take on more responsibilities. Most caregivers reported not needing 1 or more of the discussions. Results varied significantly when the sample was divided by age, with older youth more likely to have received transition services than younger adolescents. Conclusions: These findings indicate a significant disparity in access to HCT services for youth with ASD. Further research is needed to understand this disparity and develop interventions to improve HCT both for youth with ASD and those with other disabling health conditions. Additionally, many caregivers do not recognize the importance of HCT services. Education and training for caregivers, youth, and providers is essential to ensure all parties are working together to address transition issues early and often. (PsycINFO Database Record (c) 2014 APA, all rights reserved).Rehabilitation Psychology 07/2014; · 1.91 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Although it is suspected that anxiety modifies the clinical presentation of autism in fragile X syndrome (FXS), neuropsychiatric co-morbidity profiles of these two disorders have not been extensively studied. The National Fragile X Survey was completed for 1,027 males with FXS, for whom yes/no information regarding the presence of several disorders is provided. Although the survey exhibited limited depth and lacked validation by standardized measures, this exploratory study was conducted to take advantage of the data as an opportunity for identifying future lines of inquiry. We addressed the following questions: (i) how do the co-morbidity profiles of FXS males with both autism and anxiety compare to those without anxiety?; (ii) do individuals with autism exhibit specific co-morbidity profiles compared to FXS males with anxiety only, or without either autism or anxiety?; (iii) how do co-morbidity profiles in children ages 3–11 differ from profiles of individuals >12 years? The group with autism and anxiety reported the highest prevalence of attention problems, hyperactivity/impulsivity, self-injurious behavior and aggressiveness. In addition, the lowest prevalence rates of these conditions were often observed in non-anxious groups regardless of autism status. Overall, this exploratory analysis generated several hypotheses for further study: (i) anxiety increases the severity of autism in FXS, particularly through additional behavioral abnormalities; (ii) some neuropsychiatric and behavioral conditions (i.e., attention problems, hyperactivity/impulsivity, aggressiveness) are primarily related to comorbid anxiety, not autism; (iii) prevalence of behavioral abnormalities increases with age. Future studies evaluating these hypotheses should incorporate validated neurobehavioral assessments, and control for cognitive level. © 2014 Wiley Periodicals, Inc.American Journal of Medical Genetics Part A 02/2014; · 2.30 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Machine learning has immense potential to enhance diagnostic and intervention research in the behavioral sciences, and may be especially useful in investigations involving the highly prevalent and heterogeneous syndrome of autism spectrum disorder. However, use of machine learning in the absence of clinical domain expertise can be tenuous and lead to misinformed conclusions. To illustrate this concern, the current paper critically evaluates and attempts to reproduce results from two studies (Wall et al. in Transl Psychiatry 2(4):e100, 2012a; PloS One 7(8), 2012b) that claim to drastically reduce time to diagnose autism using machine learning. Our failure to generate comparable findings to those reported by Wall and colleagues using larger and more balanced data underscores several conceptual and methodological problems associated with these studies. We conclude with proposed best-practices when using machine learning in autism research, and highlight some especially promising areas for collaborative work at the intersection of computational and behavioral science.Journal of Autism and Developmental Disorders 10/2014; · 3.34 Impact Factor