19080
Identifying Interventions for Dissemination and Implementation Research Using the National Database for Autism Research (NDAR): Promises and Pitfalls
The National Database for Autism Research (NDAR) is an NIH-funded research data repository created with the express purpose of accelerating autism research by integrating heterogeneous datasets through data sharing agreements with the National Institute of Health.
Objectives:
This study summarizes a user-developed catalog of interventions shared in NDAR that would assist health services researchers and implementation scientists interested in autism studies that could be targeted for dissemination and implementation.
Methods:
This descriptive study was conducted through October 15, 2014 to review all shared NDAR data using the “data from labs” query. The term “intervention” refers to any treatments or services provided to children with autism to improve core impairments and other associated symptoms. Each NDAR project was reviewed using its research description in the NIH Research Portfolio Online Reporting Tools to identify if any interventions were involved. A catalog of interventions was then created in an Excel spreadsheet and shared data were downloaded for further investigation.
Results:
A total of 184 research projects have shared data with NDAR involving 108,386 subjects by age (54,053 affected subjects by age and 54,333 control subjects by age) whose ages ranged from toddlers to adults. Sixty-three projects (34%) involving interventions were identified. These projects were categorized further as observational (7), effectiveness (10), health services research (11), and efficacy (35) studies. However, 30 projects appearing in NDAR have not yet released their data and at least 20 projects are still ongoing, resulting in 11,972 subjects with information currently available. Interventions include pharmacotherapy (3), early intensive behavioral therapy (5), cognitive behavioral techniques (3), community health services (3), and others (e.g., peer-engagement, sensory integration based therapy, school integration service, language integration).
Conclusions:
The NDAR has high promises for data aggregation and secondary analyses to guide dissemination and implementation of interventions. However, missing data seems to be a prominent problem in those previously sharing data. In addition, researchers require clinical background knowledge on a large number of measures used in autism to enable data aggregation. Despite some inconvenience associated with finding needed data, downloading time, and data cleaning, NDAR provides rich data for the autism research community without charge. Future improvements to NDAR and those providing data can be expected that will greatly improve functionality for services research to extract information on treatments or interventions provided to individuals with autism.