PRIMARY OBJECTIVE: The aim of this study was to determine the validity of the Polish version of the Coma Recovery Scale-Revised (CRS-R).

METHODS AND DESIGN: Two trained raters, A and B, administered CRS-R on a group of 20 patients with severe brain injury (median age ± SD, 38.0 ± 14.39 years). Both rater A and rater B completed their assessment on day 1, and rater A repeated their assessment on day 2. Inter-rater and test-retest reliability were evaluated with an intra-class correlation coefficient and Spearman rank correlation. Internal consistency was estimated with Cronbach's α. Agreement in diagnostic impression was determined using Cohen's κ.

RESULTS: Inter-rater reliability for CRS-R total scores and test-retest reliability was excellent: (ρ = 0.76, p < 0.001) and (ρ = 0.92, p < 0.001), respectively. Inter-rater diagnostic agreement was good (κ = 0.72, p < 0.001). Inter-rater reliability for subscales was fair to excellent. Internal consistency was excellent (Cronbach's α = 0.85).

CONCLUSION: The Polish version of CRS-R can be administered reliably by trained raters and can successfully differentiate between vegetative state (VS), minimally conscious (MCS), and patients emerging from a minimally conscious state (EMCS).