Evaluation of the Psychometric and Measurement Properties of the SCAT5 and Child SCAT5
Abstract
Concussions are one of the most complex conditions to manage in sport medicine due to the individualized clinical presentation caused by the complex neurometabolic cascade that occurs. The identification, assessment and management of concussion requires clinicians to employ a multifaceted approach including the subjective disclosure of symptoms by patents. In order to aid in this, symptom checklists are commonly used as they provide a standardized method for collecting the severity of concussion symptoms. One of the most common symptom checklists is the 5th Sport Concussion Assessment Tool (SCAT5) symptom evaluation which is available in an adult and pediatric version. In order to better understand the psychometric and measurement properties of both versions of the SCAT5 multiple studies were conducted. Both versions of the SCAT5 were analyzed using Rasch analysis to determine the reliability and construct validity of the tools, the items that comprise the adult SCAT5 were examined using a qualitative survey to investigate how individuals interpret the underlying meaning of the symptoms and using a clinical-anchor approach the responsiveness and minimal clinically important difference (MCID) for all of the items, total number of symptoms endorsed and total symptom score for both versions were estimated. Both the adult and child SCAT5 were found to be reliable but a poor fit to the Rasch model due to the multidimensionality and redundancy within the items and the presence of response dependency between multiple pairs of items. Through an examination of the themes identified form the qualitative survey, 6 items that comprise the SCAT5 were being interpreted in a manor inconsistent with the original wording of the item. Finally, MCID estimates and measures of responsiveness were calculated and all of the MCID estimates for the adult SCAT5 were found to be significant however the child SCAT5 only had 12 of the items from the child section and 15 items from the adult section that were significant. Overall, both tools were found to have serious psychometric issues and require further refinement using a systematic test development methodology in order to guarantee the validity, reliably and clinical utility of the tools.