Improvement on the Coma Recovery Scale-Revised During the First Four Weeks of Hospital Stay Predicts Outcome at Discharge in Intensive Rehabilitation After Severe Brain Injury

Arch Phys Med Rehabil. 2018 May;99(5):914-919. doi: 10.1016/j.apmr.2018.01.015. Epub 2018 Feb 8.


Objectives: To evaluate the prognostic utility of serial assessment on the Coma Recovery Scale-Revised (CRS-R) during the first 4 weeks of intensive rehabilitation in patients surviving a severe brain injury.

Design: Prospective cohort study.

Setting: An intensive rehabilitation unit.

Participants: Patients (N=110) consecutively admitted to the intensive rehabilitation unit. Inclusion criteria were (1) a diagnosis of unresponsive wakefulness syndrome (UWS) or minimally conscious state (MCS) caused by an acquired brain injury, and (2) aged >18 years.

Interventions: All patients underwent clinical evaluations using the Italian version of the CRS-R during the first month of hospital stay.

Main outcome measures: Behavioral classification on the CRS-R and the score on the Glasgow Outcome Scale (GOS) at final discharge. Patients transitioning from UWS to MCS or emergence from MCS (E-MCS), and from MCS to E-MCS were classified as patients with improved responsiveness (IR).

Results: After a mean ± SD hospital stay of 5.3±2.7 months, 59 of 110 patients (53.6%) achieved IR. In the multivariable analysis, a higher CRS-R score change at week 4 (odds ratio =1.99; 95% confidence interval [CI], 1.49-2.66; P<.001) was the only significant predictor of IR at discharge. Fifty-three patients (48.2%) were classified as severely impaired at discharge (GOS=3). In the multivariable analysis, higher GOS scores were related to a higher CRS-R score at admission (B=.051; 95% CI, .027-.074; P<.001), a higher CRS-R score change at week 4 (B=.087; 95% CI, .064-.110; P<.001), and an absence of severe infections (B=-.477; 95% CI, -.778 to -.176; P=.002).

Conclusions: An improvement on the total CRS-R score and on different subscales across the first 4 weeks of inpatient rehabilitation discriminates patients who will have a better outcome at discharge, providing information for rehabilitation planning and for communication with patients and their caregivers.

Keywords: Brain injury; Coma; Prognosis; Rehabilitation.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Aged
  • Brain Injuries / complications
  • Brain Injuries / rehabilitation*
  • Coma / etiology
  • Coma / rehabilitation*
  • Disability Evaluation*
  • Female
  • Glasgow Outcome Scale / statistics & numerical data*
  • Humans
  • Length of Stay
  • Male
  • Middle Aged
  • Patient Discharge
  • Persistent Vegetative State / etiology
  • Persistent Vegetative State / rehabilitation*
  • Predictive Value of Tests
  • Prognosis
  • Prospective Studies
  • Recovery of Function
  • Time Factors
  • Treatment Outcome