Related Experiment Video
Updated: Oct 15, 2025
08:13
Development and Implementation of a Multi-Disciplinary Technology Enhanced Care Pathway for Youth and Adults with Concussion
Published on: January 20, 2019
6.8K
Immediate Post-Concussion and Cognitive Testing (ImPACT): Effects of Data Integration Strategies on Classification
1Department of Psychology, University of Rhode Island, Kingston, and Department of Psychiatry and Human Behavior, Alpert Medical School of Brown University, Providence, Rhode Island.
The Journal of Head Trauma Rehabilitation
|October 26, 2021
Summary
This study found that while discriminant function analyses (DFAs) improved predictive accuracy for concussion, they still missed many cases. Standard methods were more sensitive but less precise, indicating limitations in concussion assessment tools.
Area of Science:
- Sports Medicine
- Neuroscience
- Psychometrics
Background:
- The Immediate Post-Concussion Assessment and Cognitive Testing (ImPACT) is widely used for concussion evaluation.
- Assessing the accuracy of ImPACT's classification is crucial for effective concussion management.
- Previous studies have explored various methods to enhance ImPACT's diagnostic utility.
Purpose of the Study:
- To evaluate the impact of different data integration procedures on the classification accuracy of ImPACT.
- To compare the diagnostic performance of discriminant function analyses (DFAs) with the standard interpretive procedure for ImPACT scores.
Main Methods:
- Utilized ImPACT data from 300 secondary school students (grades 9-12) between 2010-2015.
- Formulated standardized regression-based equations to compute reliable change index scores.
- Conducted DFAs using various combinations of ImPACT composite scores (Verbal Memory, Visual Memory, etc.) and compared results to the standard method.
Main Results:
- DFAs showed higher positive predictive values (PPVs) (61-83%) but lower sensitivity (31-49%) compared to the standard method.
- The standard interpretive procedure had higher sensitivity (73%) but a lower PPV (45%) and poor specificity (43%).
- Cognitive composite scores did not significantly improve classification accuracy when combined with symptom self-report.
Conclusions:
- Both DFA and standard interpretive procedures for ImPACT have limitations in accurately classifying concussion.
- Base rate conditions and psychometric factors likely contribute to ImPACT's restricted classification accuracy.
- Further research is needed to refine concussion assessment protocols and improve diagnostic precision.
