Evaluating the Hierarchy of Rating Scale Categories for the Coma Recovery Scale-Revised in Non-Traumatic Brain Injury: A Rasch Analysis

Abstract

Background

The Coma Recovery Scale-Revised (CRS-R) is the reference standard for diagnosing disorders of consciousness after severe brain injury. Rating scale categories for the 6 CRS-R items have been operationalized to diagnostic criteria for states of consciousness, but the validity of these diagnostic categories has not been examined in non-traumatic brain injury.

Objective:

This study evaluates the hierarchy of CRS-R rating scale categories (RSCs) in individuals with disorders of consciousness due to non-traumatic brain injury.

Methods:

We analyzed 4562 CRS-R assessments from 410 individuals using a partial credit Rasch model. We assessed reproducibility, structural validity, measurement accuracy, and conceptual validity by examining RSC alignment with the Aspen Consensus Criteria.

Results:

All CRS-R items fit the Rasch model, with high Wright’s person separation reliability (0.94) and strata (3.8), indicating strong measurement precision. The Visual and Motor items exhibited disordered rating scale thresholds. Several RSCs currently aligned with the unresponsive wakefulness syndrome showed comparable mean category measures to RSCs aligned with the minimally conscious state.

Conclusions:

The CRS-R demonstrated strong reproducibility and validity in patients with non-traumatic brain injury, but may require refinement due to disordered thresholds. Consistent with literature in traumatic brain injury, our findings suggest that diagnostic criteria may need to be revised to better align with the constellation of behavioral features that are actually observed at different levels of neurorecovery. Specifically, RSC 4 on Auditory (consistent command following) and 3 on Arousal (Attention) may indicate emergence from the minimally conscious state.