Comparing the Different Sets of Item-Level Diagnostic Criteria of the Coma Recovery Scale-Revised (CRS-R): A...

Area of Science:

• Neuroscience
• Psychometrics
• Rehabilitation Medicine

Background:

• The Coma Recovery Scale-Revised (CRS-R) is crucial for assessing disorders of consciousness.
• Internal construct validity of the CRS-R requires robust statistical validation.
• Rasch Measurement Theory provides a framework for evaluating measurement properties.

Purpose of the Study:

• To validate the CRS-R's internal construct using Rasch analysis with a large, multicenter sample.
• To compare diagnostic criteria against a Rasch-derived measurement ruler.
• To enhance understanding of consciousness levels and diagnostic criteria.

Main Methods:

• Multicenter retrospective study involving 380 inpatients with disorders of consciousness.
• Rasch analyses were performed on validation and confirmation subsamples.
• Internal construct validity was assessed for monotonicity, unidimensionality, local independence, and invariance.

Main Results:

• Rasch analyses confirmed adequate model fit and sufficient reliability (Person Separation Index > 0.870).
• Five distinct levels of performance ability were identified.
• No significant differential item functioning was observed across patient factors, including etiology.
• CRS-R rulers facilitated comparison of diagnostic criteria, suggesting refinement possibilities.

Conclusions:

• The study provides a stable Rasch calibration of the CRS-R using the largest multicenter sample to date.
• The measurement-based approach offers insights into CRS-R score categories.
• "Automatic motor response" and "object recognition" may indicate MCS+ and eMCS, respectively.