If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Noé E, Olaya J, Navarro MD, Noguera P, Colomer C, García-Panach J, Rivero S, Moliner B, Ferri J. Behavioral recovery in disorders of consciousness: a prospective study with the Spanish version of the Coma Recovery Scale–Revised.
To describe the clinical characteristics and short-term pattern of evolution of a sample of patients within 1 year after acquiring a brain injury that led to a vegetative state (VS) or a minimally conscious state (MCS).
Inpatient brain injury rehabilitation program.
Patients with acquired brain injury (N=32; 47% traumatic, 37.5% hemorrhagic, 15.5% anoxic) who were in a VS or an MCS according to Coma Recovery Scale–Revised (CRS-R) scores.
Integrative multisensory program comprising daily physical rehabilitation procedures and multimodal sensory stimulation.
Main Outcome Measure
All patients were assessed with a Spanish version of the CRS-R at admission and then monthly for at least 6 months or until emergence from MCS.
At the time of admission, 12 patients were diagnosed as being in a VS and 20 as being in an MCS. Eight patients were able to emerge from their MCS during follow-up. Seven of these 8 patients were diagnosed as being in an MCS at inclusion, and only 1 was diagnosed as being in a VS. Emergence from an MCS was mostly associated with improvement in both the communication and motor function scales (n=4). Lesser chronicity (P=.01) and the presence of more than visual behavioral responses at admission (P=.05) were both significant predictors of emergence from an MCS.
The CRS-R seems appropriate for establishing an immediate prognosis in this population. A quick referral of these patients for specialized assessment and rehabilitation facilities is recommended.
DISORDERS OF CONSCIOUSNESS represent a neurologic challenge from a diagnostic, prognostic, and therapeutic point of view. Misdiagnosis reaches up to 15% to 43% of these cases, prognosis is uncertain, and no treatment has been empirically shown to be effective. For almost 20 years, from the first description of the vegetative state (VS) as a “wakefulness state without awareness,”
(p734) to the early 1990s, patients with disorders of consciousness were almost ignored in the scientific literature. Conversely, during the last 2 decades, the increased survival and the prolonged life expectancy of these patients have generated an increase in case incidence and prevalence, and consequently a renewed clinical and scientific concern. As a result of this interest, there has been a recent increase in the number of descriptions, including a subgroup of patients with severe brain injury appearing to retain some, albeit limited, capacity for conscious behavior.
are 3 of the most common predictors of outcome analyzed in previous investigations. However, since an MCS is a condition that has only recently been defined, there are few studies that speak to prognosis in this condition. In recent years, some longitudinal studies
including patients in an MCS have been published; however, the categorization of this state in most of these studies has been made based on a retrospective analysis of clinical data, with a significant effort by the authors to match any clinical sign observed along the evolution of the disease with actual standardized clinical criteria. Data derived from all of these studies suggest that, in contrast to patients in a VS, those in an MCS have better short- and long-term outcomes.
This prospective study describes the clinical characteristics and short-term pattern of evolution of a sample of patients within 1 year after acquiring a brain injury that led to a VS or an MCS according to their scores on the Spanish version of the Coma Recovery Scale–Revised (CRS-R). Our aim is to provide practitioners with a useful tool for prospective evaluation of those neurologic aspects that are of clinical and prognostic relevance in this population.
All patients with an acquired brain injury who were classified as Rancho Los Amigos Scale of Cognitive Functioning (RLA-S) level III or lower on admission to a specialized rehabilitation service between March 2008 and July 2010 were eligible to participate in this study. Since age and chronicity are factors influencing the evolution of these states, patients younger than 16 years and those with chronicities lower than a month or greater than 1 year were excluded from the initial sample. Eight of the 43 initial patients who met the inclusion criteria were lost to follow-up, and 1 patient died 2 months after admission. Additionally, 2 patients who had severe communication problems (aphasia-anarthria) that interfered with the assessment were excluded.
The present study population was composed of 32 patients, 10 women (31.2%) and 22 men (68.8%) with a mean age ± SD of 39.9±13.9 years (range, 16–64y) and a mean ± SD of 10±3.7 years of education (range, 5–21y). The mean interval ± SD in days from injury to the date of inclusion was 144.9±81.6 days (range, 38–360d). Most patients sustained a traumatic brain injury (TBI; n=15, 47%), with hemorrhagic stroke (n=12, 37.5%) and hypoxic-ischemic brain injury (n=5, 15.5%) accounting for the remaining cases. All patients were prospectively followed up for at least 6 months after admission or until emergence from the MCS. Those patients who failed to emerge from the MCS after the initial 6-month follow-up period continued under treatment a mean ± SD of 143.6±156.5 days (range, 1–517d), and any significant change in their clinical status according to their monthly CRS-R scores was recorded.
Spanish version of the CRS-R
The original CRS-R scale consists of 23 items that comprise 6 subscales addressing auditory, visual, motor, oromotor, communication, and arousal functions.
A summary of the psychometric properties of CRS-R (reliability, internal consistency, and validity) are available through http://www.tbims.org/combi/crs/crsprop.html and also in the special article of the Brain Injury–Interdisciplinary Special Interest Group of the American Congress of Rehabilitation Medicine.
The Spanish translation of the CRS-R was made for this study using a back-translation method, adapted to the peculiarities of the Spanish. The Spanish version of the CRS-R mirrors the structure (6 subscales) and the content of the original described previously. Specifically, all authors agreed on a version translated from 4 different versions made by 4 of the authors. The selected version was back-translated to detect errors and ensure that there were no problems of interpretation in the administration and scoring procedures. A consensus meeting was held to agree on a fully comprehensible and accurate Spanish translation that was consistent with the original English text (supplemental appendix 1).
All patients were assessed with the CRS-R by 2 expert neuropsychologists before inclusion in an integrative multisensory program comprising daily physical rehabilitation procedures and multimodal sensory stimulation. The CRS-R was then administered monthly for at least 6 months of follow-up or until emergence from the MCS. Each assessment included (1) the CRS-R total score, (2) the CRS-R subscale scores, (3) the patient's neurologic condition (VS vs MCS vs emergence from MCS), and (4) the number of subscales with scores above the cutoff for the diagnosis of an MCS (range, 1–5) or emergence from an MCS (range, 1–2).
Descriptive statistics were used for patient characteristics. A comparative analysis of demographic and clinical characteristics of patients in a VS and those in an MCS at admission was performed using t tests for continuous demographic/injury variables and chi-square or Fisher exact probability tests for categorical variables. Longitudinal analysis included comparisons of demographic and clinical variables between those patients who emerged from an MCS during the rehabilitation process and those who did not (Student t test and chi-square or Fisher exact test when needed). Logistic regression modeling was used to assess the prognostic significance of clinical variables (neurologic condition at admission, number of subscales with scores above the cutoff score for the diagnosis of an MCS, CRS-R total score, etc) and demographic variables (age, sex, chronicity, etiology, years of education) on emerging from an MCS. Chronicity and age were introduced as continuous centered variables and then dichotomized using the mean age and chronicity of those patients who emerged from an MCS (age, ≤33y or >33y; chronicity, ≤94d or >94d). Etiology (traumatic or nontraumatic) and the number of subscales with scores above the cutoff for the diagnosis of an MCS (visual alone or visual associated with another subscale) were also dichotomized to increase the clinical utility of the results. Variables were included in the multivariate analysis if the univariate P value was 0.1 or less. A backward selection procedure was used to select the subset of predictors from this group of variables.
The initial neurologic condition, according to the CRS-R criteria discussed previously, showed that 12 patients at admission were in a VS and 20 patients in an MCS. The difference in CRS-R total scores was significant (P<.001) when comparing the VS and MCS patients, but there were no other significant differences in clinical and demographic variables, with the exception of a slight but significant difference (P=.03) in the Disability Rating Scale total score (table 1). Specifically, statistical analysis did not reveal any significant difference in time postinjury either when comparing VS and MCS patients (P=0.5), or when comparing TBI versus non-TBI subjects (P=0.4). Eleven of the 20 patients in an MCS were categorized in this state exclusively by their score on the visual subscale, 4 patients by their scores on the visual and motor subscale, 3 by their scores on the visual, motor, and communication subscale, and 2 by their scores on the visual, motor, communication, and auditory subscale.
Table 1Comparison of Characteristics Between Patients at Admission
Disability Rating Scale
NOTE. Values are mean ± SD, n (%), or as otherwise indicated.
Of the 32 patients admitted in a VS or an MCS, 8 (25.8%) emerged from their MCS during follow-up. Patients admitted in a VS were less likely to emerge from an MCS than patients admitted in an MCS (7 patients [35%] admitted in an MCS and only 1 patient [8.3%] admitted in a VS emerged from an MCS). The difference was nearly significant (Fisher exact test, P=.09). Specifically, the single patient admitted in a VS who emerged from an MCS was the one with the least chronicity of the entire sample (38d). This patient was already in an MCS 2 weeks after admission into our service through behavioral relevant responses in both the communication subscale and the motor function subscale of the CRS-R scale simultaneously. Nine (75%) of the 12 patients admitted in a VS persisted in a similar neurologic condition at the time of maximum follow-up, and only 2 patients were categorized as in an MCS. These 2 patients persisted in an MCS 1 year after admission.
Regarding clinical and demographic variables, time from injury to enrollment was slightly but significantly (P=.04) shorter in those patients who emerged from an MCS compared with those who did not (94±36.4d vs 161.8±85.9d). Six (54.6%) of the 11 patients with a time from injury to enrollment shorter than 3 months and 2 (14.3%) of 14 patients with a time from injury to enrollment shorter than 6 months could emerge from an MCS during follow-up. None of the 12 patients with a time from injury to enrollment longer than 6 months could emerge from an MCS during follow-up. Regarding etiology, only 2 (11.7%) of 17 patients who had sustained a non-TBI could emerge from their MCS compared with 6 (40%) of 15 of those admitted after a TBI (Fisher exact test, P=.06) (table 2).
Table 2Comparison of Characteristics Between Patients at Follow-up
Emerged From MCS (n=8)
Not Emerged From MCS (n=24)
Neurologic status at admission
NOTE. Values are mean ± SD, n (%), or as otherwise indicated.
Considering only the sample of MCS patients at admission (n=20), 7 patients (35%) emerged from this state across follow-up. Regarding etiology, 2 patients in an MCS after a non-TBI emerged from this state 2 months after admission. None of the 9 patients admitted in an MCS after a non-TBI emerged from the MCS after the second month of follow-up, while 5 of the 11 traumatic MCS patients emerged from this state between the second and sixth month after admission (fig 1) . Five (71.4%) of the 7 patients who emerged from an MCS showed more than visual responses at admission, while most of those who did not emerge (69.2%) showed only visual responses at admission (Fisher exact test, P=0.1). Three of these 7 patients emerged showing a maximum score only on the communication subscale of the CRS-R, 3 more patients emerged obtaining maximum scores on both the communication subscale and the motor function subscale, and finally, just 1 patient emerged from an MCS through maximum behavioral responses only in the motor subscale. The patients' individual trajectory of change, in terms of number of CRS-R subscales with scores above the cutoff for the diagnosis of an MCS, from baseline to the moment of emerging MCS was not uniform, with some patients showing a slow and progressive increase and others an acute change.
The sample of MCS subjects who did not emerge from their MCS after the 6-month follow-up period (n=13) was reassessed an average of 489.5±184.1 days after the injury, and no significant changes in their neurologic situation were observed in any of them.
Chronicity dichotomized (P=.01), and the presence of more than visual behavioral responses at admission (P=.05) were all significant predictors of emergence from an MCS in the univariate logistic regression model. Age dichotomized (P=.06) and etiology dichotomized (P=.08) showed a trend toward significance and were also included in the multivariate analysis. The final multivariate model included the presence of more than visual behavioral responses at admission (P=.07) and chronicity dichotomized (P=.01). The final regression model correctly classified 87% of the patients included in the analysis (table 3).
Table 3Emergence From MCS at Follow-up: Comparison of OR From Univariate and Multivariate Logistic Regression Models
B ± SE
Univariate OR (95% CI)
B ± SE
Multivariate OR (95% CI)
Chronicity (d) (≤94 vs >94)
Age (y) (≤33 vs >33)
Sex (male vs female)
Etiology (traumatic vs nontraumatic)
CRS–R at admission (total score)
Initial status (VS vs MCS)
Number of subscales (visual alone vs visual + other)
Abbreviations: B, standardized regression coefficient; CI, confidence interval; NA, not applicable; NS, not significant; OR, odds ratio; SE, standard error.
During the last decade, a renewed attention has been paid to defining the diagnosis of patients with disorders of consciousness and estimating the prognosis for return of consciousness. Our results tend to confirm the already suggested difference in outcome between patients in an MCS and a VS,
few standardized rating scales have been adapted to these criteria, and few outcome predictors specific to an MCS have been identified. From all the assessment batteries already published, the CRS-R has an extensive psychometric analysis and also has the advantage of collecting all Aspen Workgroup criteria.
However, since the CRS-R was not developed until 2004, most previous studies including this scale have preferentially focused on its diagnostic utility, and longitudinal studies about its validity for outcome prediction are still needed. A recent evidence-based recommendation of the Brain Injury–Interdisciplinary Special Interest Group of the American Congress of Rehabilitation Medicine supported the use of the CRS-R to assess these disorders with just some minor reservations because of a lack of prognostic validity studies.
Most of the previous prognostic studies before the publication of the CRS-R used a retrospective analysis of clinical findings collected using standardized and nonstandardized methodologies to determine diagnosis and functional outcome at different moments postinjury.
According to these studies, the original CRS total score, the current diagnosis at admission (VS vs MCS), and the rate of acute functional change have been significant predictors of subacute functional disability. Using a retrospective analysis, Giacino and Kalmar
retrospectively reviewed 36 patients admitted in an MCS or a VS and analyzed functional outcome at 1 to 4 years postinjury. According to their results, patients admitted in a VS were less likely to emerge from an MCS than patients admitted in an MCS (45% admitted in a VS and 80% admitted in an MCS emerged from their MCS). Finally, Luauté et al
have recently reported a high percentage (33%) of those patients in an MCS, at least 1 year after a brain insult, recovering consciousness during a 5-year follow-up period, compared with those in a VS (0%). However initial patients’ classifications in this study was based on a modified version of the Glasgow Outcome Scale, which may be insensitive to detect those behavioral signs that characterize MCS. The absence of prospective studies comparing patients categorized in an MCS versus a VS according to standardized scales is particularly striking considering the high percentage of diagnostic confounds when these tools are not used.
In contrast to previous investigations, all patients included in our study were categorized at admission according to actual standardized criteria based on their scores on the CRS-R. Additionally, all of our patients were prospectively followed up and then assessed monthly with the same instrument. Behavioral signs of consciousness detected by the CRS-R in our sample of MCS patients included purposeful eye movements alone (n=11) or in combination with auditory, motor, communicative, or verbal signs of awareness (n=9). A number of studies have reported the natural recovery of consciousness from the VS, starting with the reemergence of visual pursuit and continuing with the gradual appearance of some kind of meaningful interaction with the environment.
Our data suggest that there are several presentations of the MCS, with those patients presenting preferentially with complex behavioral responses showing better outcomes than those showing only visual features. It seems clear that regaining consciousness progresses in a continuous fashion, with those patients with better outcomes showing more frequent, extensive, and intense spontaneous or induced behavioral responses. In cases of severe TBI, the process of recovery of consciousness is likely a continuum, with recovery of behavioral responses corresponding to recovery of connectivity from primary to high-level associative cortical regions disrupted after severe diffuse brain damage. According to our data, recent functional neuroimaging studies have shown that the progressive recovery of behavioral responses in these states is associated with a functional restoration of cortico-cortical and cortico-subcortical networks.
Characteristically, evolution of the MCS in our sample did not follow a constant rate or a uniform pattern of change from one state to another, with some patients showing an increment in the number of behavioral responses and others showing an increasing intensity of initial behavioral responses. Whether these differences result from discrepancies in etiology, demographics, or other unrecognized factors should be addressed in further investigations that include a greater number of patients followed up along the recovery process.
Because diagnostic criteria for emergence from an MCS were only recently defined, there is a lack of empirical evidence at this stage.
To clarify this clinically relevant moment, the Aspen Workgroup proposed the presence of reliable and consistent evidence of either functional communication or functional object use for MCS emergence, and proposed the CRS-R as a valid tool to monitor this period.
Our results showed that 3 patients who emerged from an MCS in our study showed a maximum score only on the communication subscale of the CRS-R, 1 patient on the motor subscale, and 4 patients emerged obtaining maximum scores on both the communication subscale and the motor function subscale. Our data are consistent with those reported by Taylor et al,
who showed that object manipulation occurred along with components of functional communication in most of their sample (56%).
Finally, our results are in agreement with the 3 major factors that traditionally have been proposed to affect the prognosis of patients with disorders of consciousness: chronicity, age, and type of brain injury.
Characteristically, the only patient of our sample admitted in a VS who emerged from an MCS was the one with the least chronicity in the entire sample. Regarding etiology, patients with TBI had greater prospects of emergence from an MCS at longer times postinjury than those with non-TBI etiologies in this study sample and other studies.
Our results should be interpreted cautiously since the specific characteristics of the sample and the analysis used here may preclude the replication of our findings under different situations. Also, the study sample is relatively small, although similar in size to the samples in some recent reports, and the proportion who emerged from an MCS was somewhat less than reported in other similar studies. Discrepancies among studies relating chronicity and follow-up may account for this difference. Future studies that include more homogeneous and larger samples with prolonged follow-up periods may help to confirm the behavioral responses and the neurologic outcomes reported in our study.
Our results suggest that the information provided by a standardized assessment of behavioral functions, as proposed by the CRS-R, in combination with time from the moment of injury to assessment, is decisive data for the establishment of a medium-term prognosis in this population. According to our data, the CRS-R seems to be a valid measure to assess and monitor behavioral signs that are seen in patients with these disorders all along the recovery process. The clinical relevance of our results is clear because the diagnosis of an MCS appears to be associated with a more favorable prognosis for consciousness recovery, particularly when it is diagnosed in the early course of recovery and it is caused by a traumatic compared to a nontraumatic injury.
Supplemental Appendix 1: The Spanish Version of the CRS-R
No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit on the authors or on any organization with which the authors are associated.