Just a moment, the page is loading...

Item factor analysis of the SNAP-IV Parent Report Form: Generating item parameter estimates








Item factor analysis of the SNAP-IV Parent Report Form: Generating item parameter estimates


Alexandra Sturm, Ph.D.


University of California, Los Angeles


None


Cai receives software royalties from Scientific Software International, Inc. Cai has received consultant income from Shire Pharmaceuticals and Janssen Pharmaceuticals through his company Vector Psychometric Group, LLC.
McCracken has received consultant income from Roche, Dart Neuroscience, and Think Now, Inc, and clinical trial contracts from Roche. He has also received study drug (Intuniv) and matching placebo from Shire.


26 September 2016


Attention-deficit/hyperactivity disorder (ADHD) is a neurodevelopmental disorder that affects between 5 and 10% of children and adolescents. A number of rating scales exist for the purpose of measuring symptoms of ADHD for children and adolescence including the SNAP-IV and the ADHD-RS-IV. Rating scales serve as a necessary tool for clinicians who wish to determine presence and severity of ADHD symptoms, in addition to monitoring response to treatment. Because the best method for detecting the effectiveness of a treatment is through perceived changes in symptoms, accurate measurement of symptomatology is crucial.

Item response theory (IRT), a latent variable model, presents an opportunity to improve the way we measure baseline clinical symptoms of ADHD and treatment response. A prior project approved by clinicalstudydatarequest.org successfully modeled the multidimensional structure of the ADHD-RS-IV and subsequently estimated treatment response across several randomized controlled trials (RCT) using a longitudinal item factor model (Sturm et al., under review). Instead of change monitored across summed scores as is traditional in clinical trials, change was estimated across latent dimensions representing inhibition, sustained attention, and hyperactivity. The findings were unique as estimating latent change of factors produced a more nuanced understanding of how standard ADHD treatments exert effects across the refined symptom model.

The analytic plan was successful using the ADHD-RS-IV and now we hope to analyze another RCT comparing multiple ADHD treatments that used the SNAP-IV as a primary endpoint. There are two core data analytic procedures that must be conducted in order to successfully model treatment effect using IRT. The first step is to estimate the item parameters of the identified measurement instrument using, in this case, a multidimensional item factor model. The second is to estimate latent change of the factors in a longitudinal IRT model, fixing the item parameters (factor loadings and thresholds) to the values found in part 1. This two-step procedure is essential because latent variable models require large samples to accurately identify item parameter estimates, and treatment samples are often not large enough to simultaneously estimate treatment effect and a number of item parameters.

Data from clinicaldatastudyrequest.org would be used to complete only part 1 of the study. Therefore, this study aims to use IRT to determine the psychometric properties of the SNAP-IV in children and adolescents with ADHD. The results of this IRT analysis are predicted to confirm the multidimensional structure of ADHD symptoms using the SNAP-IV that was originally identified using the ADHD-RS-IV. This study may produce not only a better understanding of ADHD symptoms, but will be instrumental in successful modelling of treatment effect using IRT. Results will be submitted for publication in psychological methods journals.



[{ "PostingID": 2001, "Title": "LILLY-B4Z-SB-LYDW", "Description": "A Randomized, Double-Blind Comparison of Atomoxetine versus Placebo in Children and Adolescents with Attention-Deficit/Hyperactivity Disorder and Comorbid Oppositional Defiant Disorder

Medicine: Atomoxetine, Condition: Attention Deficit Hyperactivity Disorder; Oppositional Defiant Disorder, Phase: 4, Clinical Study ID: B4Z-SB-LYDW, Sponsor: Lilly" },{ "PostingID": 3738, "Title": "LILLY-B4Z-MC-LYBX", "Description": "A Randomized, Double-Blind Comparison of Atomoxetine Hydrochloride and Placebo in Child and Adolescent Outpatients with Attention-Deficit/Hyperactivity Disorder and Comorbid Oppositional Defiant Disorder

Medicine: Atomoxetine, Condition: Attention Deficit Hyperactivity Disorder; Oppositional Defiant Disorder, Phase: 3, Clinical Study ID: B4Z-MC-LYBX, Sponsor: Lilly" }]

Statistical Analysis Plan


The publication citation will be added after the research is published.