Abstract
Introduction
The purpose of this monocentric study was to assess the long-term outcome of a group of severe traumatic brain-injured patients and explore the prognostic values of some clinical and paraclinical parameters available at the initial stage.
Methodology
The patients included were victims of severe traumatic brain injuries in 2007 or 2008. A standardized assessment was performed for each patient including clinical, radiological, and electrophysiological data collected at the initial stage, The outcomes were assessed at least 2 years after injury. Depending on the patients’ availability and ability to communicate, the assessments included measures of dependency for activities of daily living (ADL), cognitive functions, behaviour, mood, and quality of life.
Results
Eighteen patients were included, of whom ten were autonomous for ADL at the time of assessment. Memory complaints, attentional deficits, anxiety, and irritability were the main long-term impairments observed. A correlation analysis showed significant correlations between the dependency level (as rated by the Functional Independence Measure) and each of length of coma, length of the post-traumatic amnesia, and the N100 auditory evoked potentials.
Discussion
These results confirm the uniqueness of each patient regarding the long-term consequences of a traumatic brain injury and the multi-determined nature of each prognosis.
Résumé
Introduction
L’objectif de cette étude monocentrique était d’explorer le devenir à long terme de patients victimes d’un traumatisme crânien grave et de rechercher des marqueurs pronostiques.
Méthodologie
Le devenir des patients victimes d’un TC sévère a été évalué au moins 2 ans après le traumatisme. Ces patients ont tous bénéficié d’une évaluation standardisée, comprenant des paramètres cliniques, radiologiques et électrophysiologiques recueillis à la phase aiguë, au cours des années 2007 et 2008. Selon la disponibilité et le niveau de communication de chaque patient, l’évaluation à distance comportait des échelles mesurant la dépendance, les fonctions cognitives, le comportement, l’humeur et la qualité de vie.
Résultats
Parmi les dix-huit patients inclus, dix étaient autonomes pour les gestes de la vie quotidienne lors de l’évaluation tardive. La plainte mnésique, les troubles attentionnels, l’anxiété et l’irritabilité étaient au premier plan des séquelles à distance du TC. L’analyse de corrélation a montré que la durée du coma, la durée de l’amnésie post-traumatique, l’onde N100 des PEA étaient corrélées au niveau de la dépendance à la MIF.
Discussion
Les résultats confirment la singularité de chaque cas quant aux conséquences du TC à long terme et le caractère multi-déterminé du pronostic.
1
English version
1.1
Introduction
Traumatic brain injury is a major public health problem. In France, a study carried out in the Aquitaine region revealed an incidence of 280/100,000 habitants . The results of this study show that around 9% of patients with traumatic brain injury suffer severe long-term consequences . Cognitive and neurobehavioral sequelae are among the most frequent . While all aspects of cognition may be altered, executive function disorders and memory deficiencies are most often predominant . Thymus disorders are likewise quite frequent, with prevalence of the depressive syndrome that ranges from 25 to 40% according to the studies and the post-trauma time period . In fact, prevalence of the depressive syndrome is estimated 7.5 times greater in trauma patients than in the overall population . The thymic and behavioural disorders necessitate organization of a treatment process proceeding through social and vocational rehabilitation .
Early determination of a prognosis is of prime importance for the families and for patient orientation. With present-day knowledge, however, no single prognostic factor is specific enough to predict absence of awakening and to justify limitation of treatment in the brain-injured patient . Several clinical and paraclinical parameters, which are available during the acute phase, are of demonstrated predictive interest with regard to awakening and functional recovery. They include premorbid condition , age , pupillary response , depth of come as measured by the Glasgow Coma Scale and occurrence of secondary systemic brain insults (SBI) . A prognostic assessment score concerning probability of mortality and unfavourable outcome through association of clinical and scanning-based prognostic factors was established by Hukkelhoven in 2005 . Age higher than 65 years, bilateral abolition of pupillary response, absent motor response or extensor posturing, hypoxia or arterial hypotension, presence of a subarachnoid hemorrhage and a score of III or IV on the Marshall CT classification are indicators of an adverse outcome (death or severe handicap) at 6 months following severe or moderate brain injury. The two major prognostic factors during the period when consciousness is regained are duration of the coma and duration of post-traumatic amnesia (PTA). Coma length is correlated with mortality, neuropsychological damage and psychosocial adaptation . Brooks et al. have demonstrated a relationship between PTA duration and long-term persistence of cognitive disorders. In their study, while no cognitive sequel was detected in patients presenting with less than 2 weeks of PTA, cognitive, frequently mnesic troubles persisted in patients presenting with more than 3 months of PTA. From a paraclinical standpoint, numerous works have underscored the importance of imaging data, particularly encephalic MRI, which greatly facilitates assessment of the injuries. The total number of lesions visualized by MRI is correlated with the prognosis as evaluated by the Glasgow Outcome Scale three months after admission . When lesions are located at the level of the corpus callosum and the dorso-lateral part of the brain stem (cerebral peduncles and protuberance), this is a poor prognostic factor with regard to awakening and functional sequelae at 6 months , as is bilaterality of the lesions .
Neurophysiologic testing permits prognostic evaluation on awakening from a coma. Auditory evoked potential (AEP) and somesthetic evoked potential (SEP) facilitate study of the functional status of the sensory pathways at different levels of integration in the cortex of the peripheral nerve. The abolition or diminution in amplitude of early AEPs (EAEP) reflects damage of the brain stem and constitutes a poor prognostic factor, while the presence of EAEP confers weak prognostic value . Middle-latency evoked potentials permit evaluation of the sensory pathways to the primary somatosensory cortex. Prognostic value has been largely studied with regard to auditory and somesthetic evoked potentials. Bilateral absence of the primary cortical components of SEP (N20) is strongly correlated with unfavourable evolution (absence of awakening or severe deficit), while the presence of a primary cortical response may signal a possibility of favourable outcome . Identical prognostic value has been found for the primary cortical components (Na and Pa) of the middle-latency AEP (AEPml) . Late and cognitive evoked potentials (N100, mismatch negativity MMN and P300) are the reflection of elaborative processing of the information corresponding to activation of the associative cortical areas. The prognostic value of these components has been extensively studied in its auditory modality. While their presence signifies that the patient is in the process of awakening, it does not necessarily mean that functioning is to be restored .
Up until now, no single clinical or paraclinical parameter has allowed for prediction of patient outcome in terms of the neuropsychological, behavioural and emotional difficulties that are to condition future autonomy and socioprofessional rehabilitation.
The principal aim of our study is to research correlations between the level of gravity of the brain injury during the initial phase of the coma and the severity of the neuropsychological and neuropsychiatric sequels suffered by the patients subsequent to the head trauma. In addition, we have attempted to study the relationship between lesion location and neuropsychological damage.
1.2
Methodology
From 1997 to 2010, the clinical and electrophysiological data of all the patients admitted to the neurological intensive care unit with a Glasgow Score lower than or equal to 8 were collected in a standardized way to assess the prognostic value of several types of clinical and electrophysiological information . The present work is a descriptive and monocentric study based on the above cohort. To be included, patients had to have presented between January 2007 and July 2008 with severe traumatic brain injury defined by a Glasgow Coma Score lower than or equal to 8 during initial management or within the 48 hours following the trauma. The patients had to be at least 16 years old and initial assessment had to involve a neurophysiologic evaluation and a brain scan. They also had to be subsequently hospitalized in a physical medicine and rehabilitation unit. The patients fulfilling these criteria were extracted from the cohort so as to assess the prognostic value of evoked potentials.
Patient outcome was assessed two to four years after the brain injury, that is to say in 2010 and 2011. Clinical and behavioural analysis, neuropsychological evaluation and questioning of the families were carried out during a follow-up medical consultation in the framework of the care and treatment pathway. If it was impossible for the patient to be present, the consultation took place over the telephone.
Retrospective analysis was performed starting with the data collected during the initial phase.
Ethical considerations: as an observational study that did not modify patient management and in accordance with French law, this type of non-interventional study does not necessitate the approval of an ethics committee. All of the conscious patients involved provided their informed consent as regarding their participation in the study and use of the data. If difficulty in communication did not allow for informed consent, it was provided by a person representing the patient.
1.2.1
Data collection and analysis
1.2.1.1
Explanatory variables (initial prognostic factors)
The studied variables comprised sex, age, initial Glasgow Coma Score, pupillary response, number of secondary brain insults (SBI), duration of coma and duration of post-traumatic amnesia (PTA). In order to avoid a confounding factor pertaining to sedation treatments undergone by patients with a breathing tube, the Glasgow Coma Score taken into account was the one established during the initial intervention of emergency services, even if it was higher than 8. Pupillary response was characterized as unreactive when at least one pupil was not reactive. Several secondary brain insults were systematically researched: systolic arterial hypotension lower than 90 mmHg, hypoxia defined as oxygen saturation lower than 90%, anemia with a hematocrit value lower than 26%. The number of SBIs having occurred during the first week was calculated.
Coma duration was defined as the time interval between the brain injury (BI) and the appearance of motor response to simple orders. Post-traumatic amnesia (PTA) duration was defined as the time interval between the BI and the resumption of mnemonic fixation, which was retrospectively evaluated from the data available in the medical records and by questioning of the patient and his family. Duration is indicated in terms of number of days and a ceiling of 90d was established for patients presenting with persistent impairment of consciousness .
The neuroradiological assessment included the Marshall Score obtained during the initial CT-scan and an MRI analysis of the axial sequences T1, T2, T2 flair, gradient-echo T2 (T2*) and diffusion, as well as the sagittal sequence T1. When several MRIs were carried out in the same patient, the earliest one was analyzed. The lesion-based criteria were: presence of hydrocephalus, total number of diffuse axonal lesions (DAL) (all sequences included) and number of T2* diffuse axonal lesions (hemorrhagic DAL). For each patient, analysis of lesion location was performed by a neuroradiologist who had not been informed of patient evolution. The systematically analyzed anatomical areas were: lamina quadrigemina (tectum), midbrain, protuberance, bulb, thalamus, basal ganglia, corpus callosum, and the temporal, frontal, parietal and occipital lobes.
Electrophysiological assessment included a recording of early auditory evoked potentials (EAEP), as well as middle-latency evoked potentials (MLEP) and late auditory (N100) and cognitive evoked potentials (MMN). The presence of the P300 event-related component was indicated when it had been researched. Classification was carried out in accordance with the standardized scales used for the evoked potentials of the brain stem and middle latencies . So as to simplify analysis, classification was binomial. For early auditory evoked potentials, grades 1 (normal) and 2 (latency increase between peaks I and V without diminution in amplitude) were assigned to the “normal” category, while grades 3 (amplitude of the ratio V/I < 0,5), 4 (peak IV or V not detected) and 5 (only peak I present) were considered as “abnormal”. As regards middle-latency auditory evoked potentials, the “normal” category corresponded to the unilateral or bilateral presence of the Na and Pa waves, while the “abnormal” category was defined by bilateral abolition of the Pa wave. The late evoked potentials N100, MMN and P300 were considered present when a response or onset of response was found. In the event of asymmetry between the two hemispheres, the better response was taken into consideration.
1.2.1.2
Variables to be explained (data collected 2 to 4 years after the BI)
For each patient, a standardized evaluation of cognitive and neurobehavioral deficiencies was carried out. The neuropsychological assessment included an evaluation of the memory function with the mnenomic efficiency battery (MEB) authored by Signoret , in its abridged version of 84 tests assorted into 7 categories according to the type of memory impairment. In each category, a standardized score greater than standard deviation 1.6 was considered pathological and counted as one point. Total score could range from 0 (no mnemonic deficit) to 7 (total amnesia). In order to evaluate attentional capacities, four tests in the TEA battery (Test d’Évaluation de l’Attention) were performed: the phasic alert test, the working memory test, the Go No-Go exercise and the flexibility test. A test result was considered pathological if the test was not taken or if the median T score for reaction times, errors and/or omissions was lower than 30, corresponding to –2 standard deviations from the mean. A “pathological” result was counted as one out of 4 possible points. In addition, presence of frontal lobe disorder was examined in the Frontal Assessment Battery at Bedside (FAB) by Dubois et al. with a pathological threshold lower than or equal to 15. A more global evaluation was carried out using the revised neurobehavioral scale NRS-r with its 29 items . Finally, depressive syndrome was sought out using the depression inventory authored by Beck . A depressive syndrome was considered to exist in the event of a score higher than or equal to 20.
Evaluation of activity limitation was carried out by the same physician during the clinical examination or in questioning on patient outcome using the Glasgow Outcome Scale (GOS) : (GOS V, good recovery, low disability; GOS IV, moderate disability; GOS III, severe disability; GOS II, persistent vegetative state; GOS I, death), and the Functional Independence Measure . The patient’s participation in social life was evaluated in terms of living place as well as marital and professional status. The QOLBI questionnaire provided assessment of quality of life as perceived by the patient himself and the person accompanying him. A score out of 30 points (maximum satisfaction) was recorded in percentage for the purposes of statistical analysis.
1.2.1.3
Data analysis
Descriptive statistical analysis was carried out for each variable. Second-level analysis was performed in order to find correlations between explanatory variables and the variables to be explained. The correlations were studied with regard to explanatory clinical variables (GCS, number of SBIs, duration of coma and post-traumatic amnesia) and with regard to explanatory paraclinical, electrophysiological and radiological variables (number of total diffuse axonal and hemorrhagic lesions on MRI). These elements were correlated with data concerning long-term outcome (GOS, FIM, NRS, QOLBI, evaluation of attention and memory functions as well as depression). For quantitative variables, the Spearman rank correlation coefficient was calculated. In order to achieve correlations with qualitative data to be explained, a Wilcoxon rank sum test was applied. When the explanatory data and the data to be explained were binary, a Chi 2 test was carried out.
In addition, we wished to find out whether certain brain regions were more specifically damaged according to the nature of the cognitive disorders, for example a temporal lesion in patients with mnemonic impairment or a frontal or parietal lesion in patients with attentional problems. Correlations between frontal, parietal or temporal location and the results of attention and memory tests were expressly researched and studied.
1.3
Results
1.3.1
Population
Among the 151 patients hospitalized between January 2007 and July 2008 in neurological intensive care for severe traumatic brain injury, 36 underwent electrophysiological assessment, and 22 of them were also subjected to MRI; 2 of the patients died during the sub-acute traumatic period, one patient refused to participate in the study and one patient was lost of sight before outcome assessment. So, 18 patients fulfilled the inclusion criteria. Twelve patients were examined some time after the injury and eight of them had a neuropsychological assessment carried out. As concerning the 4 patients who were unable to benefit from cognitive assessment, one of them refused, while the other three presented with overly impaired consciousness or faculty of communication. Five patients were contacted by telephone since they could not be examined on account of geographical remoteness or they were not being able to travel. One patient died in the course of the long-term evaluation.
The main cause of brain injury was a road accident ( n = 12, 67%). In 5 cases (28%), the BI was secondary to a fall and in one case (6%), it came about through an attack with blunt impact.
During initial management, seven patients (39%) had an initial Glasgow Coma Score higher than 8, but their state of consciousness had deteriorated over the first 48 hours following traumatic injury. Eight patients (42%) had at least one unreactive pupil during initial management. Nine patients had required neurosurgical management for mass lesion evacuation ( Tables 1a and 1b ).
| n | Mean ± ET (min–max) | |
|---|---|---|
| Age | 18 | 37.3 ± 19.3 (16–78) |
| GCS | 18 | 6,8 ± 3,1 (3–14) |
| Number SBI | 18 | 0.56 ± 0.9 (0–3) |
| Marshall Score TDM | 18 | 3.1 ± 0.9 (2–4) |
| Total number DAL | 18 | 7.5 ± 9.4 (0–40) |
| Number DAL T2* | 18 | 5 ± 9.3 (0–40) |
| Coma duration (d) | 18 | 25 ± 14.6 (9–60) |
| PTA duration (d) | 18 | 66 ± 29.8 (15–90) |
| GOS | 18 | 3.8 ± 1,1 (1–5) |
| FIM | 17 | 87.4 ± 44 (20–126) |
| FAB | 9 | 15.2 ± 2.3 (12–18) |
| Beck Scale | 12 | 11.9 ± 10 (1–30) |
| QOLBI patient | 12 | 63.9 ± 22.2 (25–100) |
| QOLBI person accompanying | 14 | 60.2 ± 26.1 (19–100) |
| NRS | 12 | 17 ± 12 (3–39) |
| n | Present/Normal | Absents/Anormal | |
|---|---|---|---|
| Bi-pupillary reaction [ n (%)] | 18 | 17 (94) | 1 (6) |
| Marshall Score Surgical mass lesions > 25cc [ n (%)] | 18 | 9 (50) | 9 (50) |
| Brain stem lesion [ n (%)] | 18 | 8 (44) | 10 (56) |
| Bilateral BI lesions [ n (%)] | 18 | 4 (22) | 14 (78) |
| Diencephalic lesions [ n (%)] | 18 | 11 (61) | 7 (39) |
| Corpus callosum lesions [ n (%)] | 18 | 11 (61) | 7 (39) |
| Temporal cortex lesions [ n (%)] | 18 | 16 (89) | 2 (11) |
| Frontal cortex lesions [ n (%)] | 18 | 14 (78) | 4 (22) |
| Parietal cortex lesions [ n (%)] | 18 | 9 (50) | 9 (50) |
| Occipital cortex lesions [ n (%)] | 18 | 5 (28) | 13 (72) |
| EAEP (Norm/Abnorm) [ n (%)] | 18 | 17 (94) | 1 (6) |
| AEPml (Norm/Abnorm) [ n (%)] | 18 | 16 (89) | 2 (11) |
| N100 wave [ n (%)] | 18 | 15 (83) | 3 (17) |
| MMN [ n (%)] | 17 | 10 (59) | 7 (41) |
| P300 wave [ n (%)] | 13 | 5 (38) | 8 (62) |
| Attention test (norm/abnorm) | 8 | 3 | 5 |
| Memory test (norm/abnorm) | 8 | 2 | 6 |
MRI had been carried out over a mean time delay of 28 days; median of 10d, 1st quartile: 5d, 3rd quartile: 27d. Image analysis revealed highly disparate lesion locations, and the number of diffuse axonal lesions varied considerably from one individual to the next.
Early electrophysiological examinations (EAEP and AEPml), N100 waves and MMN had been carried out in the 18 patients, while P300 waves had been researched in 13 patients. Mean delay before AEP was 9 days, the median was 7d; 1st quartile: 3d, 3rd quartile: 12d.
Patient outcome assessment was carried out two to four years after the trauma and is reported in Tables 1a and 1b .
A characteristic depressive syndrome was detected in 2 out of the 12 patients surveyed by means of the Beck Depression Inventory.
The NRS Scale was applied with regard to 12 patients. Complaints on memory were the most frequent; they were reported in 11 out of the 12 patients interrogated. Nine patients complained about an attention deficit; more specifically, they reported difficulties in conceptual organization, planning ahead, mental flexibility and taking initiatives. Nine patients complained of mental fatigability. Anxiety, somatic preoccupations and irritability were reported in 8 out of the 12 patients, while depressive moods were described in 3 of them.
The QOLBI quality of life questionnaire was used with regard to 12 patients. The mean “satisfaction” score was 64%. The most frequently reported sources of dissatisfaction were:
- •
intellectual, with a majority of complaints pertaining to memory and attention capacities;
- •
psychological and social, with lessened capacity to express and control emotions and problems communicating with friends and family;
- •
personal, especially concerning difficulty accepting the sequelae of the accident and a new identity, with the impression of having enjoyed little or no self-fulfillment since the accident. For the 14 persons accompanying the patients and with whom the QOLBI questionnaire was used, satisfaction with quality of life was estimated at 60%.
The matrimonial and conjugal situation of the patients had not greatly changed following the brain injury; before the traumatic event, 13 patients were single and 5 were married, while after the traumatic event, 12 patients were single, 5 were married, 1 married subsequent to the accident, and 1 married patient died.
From a professional standpoint, 9 patients (53%) presented with BI sequelae incompatible with occupational activity in an ordinary environment; 2 of them were adapting to a protected environment and the 7 others were disabled. Two patients had resumed full-time employment after having engaged in “therapeutic” part-time work over a two-year period. One of them managed to recover full-time employment as a qualified heritage officer in civil service and benefited from flexible scheduling. The other had to change jobs after having been dismissed unwillingly from a position as human resources manager in a factory. The reasons put forward involved difficulties in task organization and impulsiveness during everyday work activities. New employment on a temporary basis was found for this person as a human resources adviser in municipal administration. Five patients had resumed their studies following a delay lasting 2 to 4 years. One of them was retired at the time of the brain injury.
The explanatory variables presenting the most significant correlations with regard to outcome parameters were duration of coma and duration of post-traumatic amnesia ( Tables 2a and 2b ).
| GCS n = 18 | SBI n = 18 | Total DAL n = 18 | DAL T2* n = 18 | Coma duration n = 18 | PTA duration n = 18 | |
|---|---|---|---|---|---|---|
| FIM n = 17 [r ( P )] | (0.21 (0.4)) | (–0.07 (0.8)) | (0.59 (0.01*)) | (0.39 (0.1)) | (–0.5 (0.04*)) | (–0.68 (0.003*)) |
| GOS n = 18 [r ( P )] | (0.30 (0.2)) | (–0.03 (0.8)) | (0.6 (0.008*)) | (0.28 (0.3)) | (–0.54 (0.01*)) | (–0.75 (0.003*)) |
| QOLBI Patient n = 12 [r ( P )] | (0.66 (0.2*)) | (–0.15 (0.6)) | (0.52 (0.1)) | (0.26 (0.4)) | (–0.36 (0.2)) | (–0.76 (0.004*)) |
| QOLBI Person accompanying n = 14 [r ( P )] | (0.31 (0.3)) | (0.12 (0.68)) | (0.59 (0.02*)) | (0.35 (0.2)) | (–0.57 (0.03*)) | (–0.73 (0.003*)) |
| NRS n = 12 [r ( P )] | (–0.22 (0.49)) | (0.29 (0.36)) | (–0.20 (0.5)) | (–0.04 (0.9)) | (0.51 (0.1)) | (0.36 (0.2)) |
| BECK n = 12 [r ( P )] | (0.06 (0.84)) | (–0.49 (0.11)) | (0.07 (0.8)) | (0.21 (0.5)) | (0.03 (0.9)) | (0.19 (0.5)) |
| FAB n = 9 [r ( P )] | (–0.27 (0.5)) | (–0.10 (0.8)) | (0.06 (0.8)) | (–0.15 (0.7)) | (–0.38 (0.4)) | (–0.38 (0.9)) |
| Attention test n = 8 [r ( P )] | (0.3 (0.7)) | (–1.2 (0.2)) | (0.8 (0.5)) | (–0.4 (0.6)) | (–1.6 (0.1)) | (–0.3 (0.7)) |
| Memory test n = 8 [r ( P )] | (1.3 (0.2)) | (–1.2 (0.4)) | (1 (0.3)) | (0.3 (0.7)) | (–1.3 (0.2)) | (–1 (0.3)) |
| EAEP n = 18 Normal = 17 Abnormal = 1 | AEPml n = 18 Normal = 16 Abnormal = 2 | N100 n = 18 Present = 15 Absent = 3 | MMN n = 17 Present = 10 Absent = 7 | P300 n = 13 Present = 5 Absent = 8 | |
|---|---|---|---|---|---|
| MIF n = 17 [r ( P )] | (0.9 (0.3)) | (1.6 (0.2)) | (2.3 (0.022*)) | (0.9 (0.3)) | (0.9 (0.3)) |
| GOS n = 18 [r ( P )] | (–1.2 (0.2)) | (1.3 (0.1)) | (1.7 (0.1)) | (0.7 (0.5)) | (0.5 (0.6)) |
| QOLBI Patient n = 12 [r ( P )] | (–1.3 (0.2)) | (1.1 (0.2)) | (1.2 (0.2)) | (0.5 (0.6)) | (1.3 (0.2)) |
| QOLBI person accompanying n = 14 [r ( P )] | (–1.3 (0.2)) | (1 (0.1)) | (1.6 (0.1)) | (0 (1)) | (0 (1)) |
| NRS n = 12 [r ( P )] | (0 (1)) | (–1.5 (0.1)) | (–1.6 (0.1)) | (–0.6 (0.5)) | (–0.9 (0.4)) |
| BECK n = 12 [r ( P )] | (0.4 (0.5)) | (–1.3 (0.2)) | (–1.3 (0.2)) | (–0.3 (0.7)) | (0.4 (0.6)) |
| FAB n = 9 [r ( P )] | (0.5 (0.5)) | (0.3 (0.7)) | (0.3 (0.7)) | (–0.9 (0.4)) | (0 (1)) |
As regards to the imaging data, number of diffuse axonal lesions or levels of dependence or types of long-term neuropsychological sequelae were not correlated or were correlated in ways that were contrary to expectations (between total DAL number and the attendant FIM, GOS and QOLBI). Paradoxically enough, the patient with highest number of diffuse axonal lesions ( n = 40) managed to resume his studies and obtain a baccalaureate with honours, albeit with the right to take a third more time than a non-disabled examinee. And as for the three patients presenting with the lowest number of diffuse axonal lesions, they still had severe neurological impairment, with scores equal to III on the GOS. Total DAL number was not correlated with the neuropsychological attention and memory tests.
More specifically, the presence of a frontal or a parietal lesion was not correlated with attention deficit severity as assessed by the TEA battery ( Table 3 ). Out of the 7 patients having undergone that evaluation, four presented with at least one set of “pathological” results; in 3 of them, there existed a frontal and/or parietal lesion. Only one patient had no lesion of the frontal-parietal network, but there existed a lesion of the basal ganglion and the right caudate nucleus. As regards to the 3 patients presenting with a normal TEA battery, frontal lesions were found in two of them, and a frontal and parietal lesion in the third. The temporal lesions were not correlated with a mnemonic disorder evaluated by the MEB (mnemonic efficiency battery). Five patients out of the 7 presenting with a temporal lesion suffered from memory impairment with at least 4 pathological MEB results; for the other 2, the latter were normal. Only one patient with pathological MEB results did not present with a temporal lesion.
| Frontal lesions | Parietal lesions | Temporal lesions | |
|---|---|---|---|
| FAB n = 9 | Chi 2 = 0.03 P = 0.5 | Chi 2 = 0.09 P = 0.5 | Chi 2 = 1.3 P = 0.3 |
| Attention test n = 7 | Chi 2 = 1.6 P = 0.2 | Chi 2 = 0.035 P = 0.8 | Chi 2 = 0.68 P = 0.4 |
| Memory test n = 8 | Chi 2 = 0.88 P = 0.3 | Chi 2 = 0.17 P = 0.7 | Chi 2 = 0.38 P = 0.5 |
The FAB was not correlated with the presence of a frontal lesion.
Study of correlations between data from the initial electrophysiological evaluation and long-term outcome revealed a significant correlation between the N100 wave and the functional independence measure (FIM). None of the other correlations between initial electrophysiology results and outcome were significant ( Table 2b ).
1.4
Discussion
Given the relatively small dimensions of the sample, this preliminary work is essentially limited to interrogation of initial prognoses with regard to the neuropsychological and neuropsychiatric consequences corresponding to the complaints most frequently voiced by patients and their families’ years after the traumatic event .
In the present study, the majority of the patients who could be evaluated still suffered from memory impairment and attentional difficulties, were present in 75 and 33% respectively. While dysexecutive syndrome was apparently less frequent, it was probably underestimated in the tests taken, especially the FAB.
The originality of our study consisted of researching possible correlation between on the one hand, the many clinical, radiological and neurophysiologic parameters available during the acute phase and, on the other hand, long-term outcomes, particularly those involving cognition, dysthymia and quality of life.
Study of correlations confirmed the prognostic value of coma and post-traumatic amnesia duration with regard to long-term cognitive and functional sequelae .
From a neuroradiological standpoint, the pejorative prognostic value of brain stem lesions is well-known, particularly when the lesions are bilateral . However, and contrarily to the study published by Firshing et al. in 1998 , our study showed that a pejorative outcome cannot be assumed simply on account of bilateral brain stem lesions; some patients with the latter were able to recover. A more nuanced study of the anatamo-functional correlation associating with injury to the brain stem is necessary and would necessitate thorough analysis of the lesions by an experienced radiologist.
Other studies have established correlations between the depth of lesions in the brain parenchyma, particularly the basal ganglia, and the gravity of later disability . Moreover, several studies have underscored a correlation between loss of brain volume as assessed by MRI and severity of the brain injury .
Given the involvement of the temporal regions in memory and of the frontal regions in executive functions, we sought out reciprocal correlations between damage to these anatomical structures and deficits in the corresponding cognitive functions. All but one of the patients with a frontal-parietal lesion was presented with an attention deficit in at least one of the TEA tests. While one patient presented with an attention deficit without frontal lesion, there existed a lesion of the basal ganglion and the right caudate nucleus, and it has been demonstrated that the basal ganglia also assume a key role in the executive functions . Similarly, all but two of the patients with a temporal lesion suffered from mnemonic impairment and had at least 4 pathological test results in the memory assessment battery. The patient without a temporal lesion presented with pathological results in the memory efficiency battery (MEB).
The clinical and functional consequences of diffuse axonal lesions (DAL) remain a debated subject. Lagares et al., en 2009 showed that DAL depth is correlated with severity of the coma. A correlation has also been observed between the number of DALs observed in MRI and outcome in terms of the GOS . The ischemic nature of these lesions appears to have pejorative prognostic value with regard to hemorrhagic DALs . On the other hand, the prognostic values of DALs with regard to cognitive disorders remain to be defined. In our study, we did not find any correlation between number of DALs and severity of mnemonic or dysexecutive impairments long after the traumatic event. It may be the case that the MRI sequences used for detection of DALs were lacking in sensitivity and influenced the results observed in our study. Moreover, the MRIs were carried out at various time intervals, in some cases, several weeks after the traumatic event, and the delay may have diminished the sensitivity of the diffusion sequences in detection of some axonal lesions. Use of sequences, such as susceptibility-weighted imaging (SWI) that display greater sensitivity to the microbleeding associated with DALs may facilitate detection of microlesions in strategic regions like the association areas, the “deep” areas (hypothalamic-pituitary region) and the commissural fibers (fornix fibers, anterior and posterior forceps) that may fail to be perceived in standard T2* sequences . As for the sequences of diffusion tensor imaging (DTI), they allow detection the lesions of the white matter fibers that are neither visible in standard sequences nor associated with axonal lesions; as a result, they can enhance analysis of the correlations between post-traumatic lesions and cognitive disorders and prediction of these types of correlations in patients suffering from brain injury of moderate severity . And yet, in spite of these highly promising results, SWI and DTI sequences are not readily available for clinical practice. Use of these sequences necessitates the presence of staff with experience in re-reading data, and they may be difficult to carry out in the early phases after the initial trauma due to the patient’s agitated condition, to artifacts from the material used in intensive care and to the need to have a high Field 3 Tesla machine ensuring high-quality images.
On an overall basis and in view of apprehending the cognitive sequelae, our results confirm the importance of a precise and quantified evaluation of brain imaging in terms of the location and nature of the lesions. Network functioning and brain plasticity must be taken into account so as to avoid overinterpreting the prognostic value of the imaging data. In our series, one noteworthy point is that the patient presenting with the largest number of DALs (40 lesions) recovered fully enough to earn a science baccalaureate just 2 years after the accident; this particular case underscores the multifactorial nature of the prognosis following brain injury; lesions alone are not necessarily of decisive importance.
From an electrophysiological standpoint, two patients in our study presented with bilateral abolition of middle-latency AEP. In both cases, there existed a brain stem lesion. Three years after the traumatic event, they were conscious but severely disabled. While bilateral AEPml abolition is consequently a marker of evolution towards grave impairment, it is not necessarily associated with non-wakefulness in the BI patient, and this finding is in contradiction with observations in patients with non-traumatic comas .
As in several previous studies, we found a significant correlation between the N100 component and long-term outcomes . In the present case, absence of the N100 wave was correlated with a less satisfactory degree of autonomy according to the Functional Independence Measure (FIM). However, it should be noted that in most of the subjects, only one significant correlation between the electrophysiological data and the outcomes was found. Study of cognitive potentials showed that among the 10 patients in whom MMN was present, none evolved towards a vegetative state. This result is in agreement with several preceding studies showing that MMN presence meant that the patient was in the process of awakening . More recently, new protocols have been developed to study the electrophysiological response of the brain when the patient’s first name is pronounced or during “active” tasks in which the patient is asked to detect violations of the sequencing rules for standard and deviant stimuli . The results confirm the interest of both imagery and electrophysiology as a complement to the clinical examination, while etiology of the coma should also be taken into account. The attendant tasks are relatively simple and may be carried out at bedside and repeated according to the patient’s evolution.
1.5
Conclusion
In this study, there existed large-scale heterogeneity in the long-term clinical outcomes of the brain-injured patients. The differences pertained to the patient’s state of health and social status before the traumatic event, age, presence of other comorbidities and the nature of the psychological profile: anxiety, phobias, addictions… The differences in etiologies and lesion mechanisms, the isolated or associated character of the extracerebral lesions and the existing or non-existing neurosurgical character of the lesions all need to be taken into account during prognostic assessment.
Years after the traumatic event, each situation becomes singular and unique given the nature of the impairments, the limits to activity, and the restrictions to participation with which the individual patient has got to cope. The family context and the overall environment are of increasingly essential importance. The biopsychosocial model of the disability becomes fully relevant as a means of understanding the actual consequences of the severe traumatic brain injury. It consequently appears that the individual prognosis involves multiple important determinants, each of which need to be taken into consideration at every stage of care and treatment.
Disclosure of interest
The authors declare that they have no conflicts of interest in relation to this article.
2
Version française
2.1
Introduction
Le traumatisme crânien est un problème majeur de santé publique. En France, l’étude réalisée en Aquitaine retrouvait une incidence de 280/100 000 habitants . Les résultats de cette enquête montrent qu’environ 9 % des patients victimes d’un traumatisme crânien conservent des séquelles graves . Les séquelles cognitives et neurocomportementales font partie des séquelles les plus fréquentes à long terme . Tous les secteurs cognitifs peuvent être altérés mais ce sont les troubles des fonctions exécutives et les déficiences mnésiques qui sont le plus souvent au premier plan . Les troubles thymiques sont également très fréquents avec une prévalence du syndrome dépressif qui peut varier entre 25 et 40 % selon les études et le délai post-traumatique . La prévalence du syndrome dépressif chez le patient traumatisé crânien est estimée 7,5 fois plus importante que celle de la population générale . Ces troubles thymiques et comportementaux justifient un parcours de soins organisé jusqu’à la réinsertion socio-professionnelle .
Déterminer le pronostic à un stade précoce demeure un enjeu important pour les familles et l’orientation des patients. Néanmoins, dans l’état actuel des connaissances, aucun facteur pronostique isolé n’est suffisamment spécifique pour prévoir l’absence d’éveil et envisager des limitations de soins chez le patient TC . Plusieurs paramètres cliniques et paracliniques, disponibles à la phase aiguë, ont montré leur intérêt pour prévoir l’éveil et la récupération fonctionnelle. Il s’agit notamment de l’état prémorbide , de l’âge , de la réactivité pupillaire , de la profondeur du coma évaluée par le score de coma de Glasgow, de la survenue d’agression cérébrale secondaires d’origine systémique (ACSOS) . Un score d’évaluation pronostique probabiliste de la mortalité et d’une évolution défavorable associant les facteurs pronostiques cliniques et scannographiques a été établi par Hukkelhoven en 2005 . Un âge supérieur à 65 ans, une abolition bilatérale de la réactivité pupillaire, une réponse motrice absente ou en extension, une hypoxie ou une hypotension artérielle, la présence d’une hémorragie sous-arachnoïdienne et un score évalué à III ou IV sur la classification scannographique de Marshall sont les indicateurs d’une évolution péjorative (décès ou handicap sévère) à 6 mois d’un traumatisme crânien sévère ou modéré. Les deux facteurs pronostiques majeurs lors de la période de reprise de conscience sont la durée du coma et la durée de l’amnésie post-traumatique (APT). La durée du coma est corrélée à la mortalité, l’atteinte neuropsychologique et l’adaptation psychosociale . Brooks et al., ont montré une relation entre la durée d’APT et la persistance de troubles cognitifs à long terme. Dans cette étude, il n’a pas été décelé de séquelle cognitive chez les patients qui présentaient une APT de moins de 2 semaines ; alors qu’il persistait des troubles cognitifs, notamment mnésiques chez les patients qui présentaient une APT supérieure à 3 mois. Au plan paraclinique, de nombreux travaux ont montré l’importance des données d’imagerie et notamment de l’IRM encéphalique qui permet de préciser le bilan des lésions. Le nombre total de lésions visualisées sur l’IRM est corrélé́ au pronostic évalué́ par la Glasgow Outcome Scale trois mois après l’admission . La localisation lésionnelle au niveau du corps calleux, de la partie dorso-latérale du tronc cérébral (pédoncules cérébraux et protubérance) est un facteur de mauvais pronostic de l’éveil et des séquelles fonctionnelles à 6 mois ; de même que le caractère bilatéral des lésions .
Les explorations neurophysiologiques permettent une évaluation pronostique d’un éveil de coma. Les potentiels évoqués auditifs (PEA) et somesthésiques (PES) permettent d’étudier l’état fonctionnel des voies sensorielles à différents niveaux d’intégration du nerf périphérique au cortex. L’abolition ou la diminution d’amplitudes des PEA précoces (PEAp) traduit une atteinte du tronc cérébral et constitue un facteur de mauvais pronostique alors que la présence des PEAp confère une faible valeur pronostique . Les potentiels évoqués de latence moyenne évaluent les voies sensorielles jusqu’au cortex primaire. La valeur pronostique a surtout été étudiée pour les potentiels évoqués somesthésiques et auditifs. L’absence bilatérale des composantes corticales primaires des PES (N20) est fortement corrélée à une mauvaise évolution (absence d’éveil ou déficit sévère) alors que la présence d’une réponse corticale primaire est en faveur d’une évolution favorable . Une valeur pronostique identique a été retrouvée pour les composantes corticales primaires (Na et Pa) des PEA de latence moyenne (PEAlm) . Les potentiels évoqués tardifs et cognitifs (N100, Mismatch negativity, MMN, et P300) sont le reflet d’un traitement élaboré de l’information correspondant à l’activation de régions corticales associatives. La valeur pronostique de ces composantes a surtout été étudiée dans la modalité auditive. Leur présence signifie que le patient est dans le processus d’éveil mais ne préjuge pas du devenir fonctionnel .
À ce jour, aucun paramètre clinique ou paraclinique isolé ne permet de prévoir l’évolution en termes de déficiences neuropsychologiques, comportementales et affectives, qui conditionnent pourtant l’autonomie des patients et leur réinsertion socio-professionnelle.
Notre étude a pour objectif principal de rechercher des corrélations entre le niveau de gravité du traumatisme crânien établi à la phase initiale du coma et la sévérité des séquelles neuropsychologiques et neuropsychiatriques des patients à distance du traumatisme crânien. Par ailleurs, nous avons cherché à étudier les relations entre la localisation lésionnelle et l’atteinte neuropsychologique.
2.2
Méthodologie
Entre 1997 et 2010, les données cliniques et électrophysiologiques de tous les patients avec score de coma de Glasgow inférieur ou égal à 8 admis dans l’unité de réanimation neurologique ont été recueillies de façon standardisée afin d’évaluer la valeur pronostique de plusieurs données cliniques et électrophysiologiques . L’étude actuelle est une étude descriptive et monocentrique basée sur cette cohorte. Pour être inclus, les patients devaient avoir présenté entre le mois de janvier 2007 et le mois de juillet 2008 un traumatisme crânien grave défini par un score de coma de Glasgow inférieur ou égal à 8 lors de la prise en charge initiale ou dans les 48 heures suivant le traumatisme. Les patients devaient avoir plus de 16 ans et le bilan initial devait comprendre une évaluation neurophysiologique et une IRM encéphalique. Ils devaient avoir été hospitalisés dans un service de médecine physique et de réadaptation dans les suites. Les patients répondant à ces critères ont été extraits de la cohorte pour évaluer la valeur pronostique des potentiels évoqués.
Le devenir des patients a été évalué deux à quatre ans après le traumatisme crânien ; au cours des années 2010 et 2011. L’analyse clinique, comportementale, le bilan neuropsychologique, ainsi que l’interrogatoire des familles ont été effectués au cours d’une consultation médicale de suivi dans le cadre du parcours de soin. Si la présence du patient était impossible, une consultation téléphonique a été effectuée.
Une analyse rétrospective du pronostic a été réalisée à partir des données recueillies à la phase initiale.
Considération éthique : il s’agit d’une étude observationnelle ne modifiant pas la prise en charge des patients et en accord avec la loi Française, ce type d’étude non interventionnelle ne nécessite pas l’approbation d’un comité d’éthique. Tous les patients conscients ont donné leur consentement éclairé pour participer à l’étude et pour l’utilisation des données. En cas de trouble de la communication ne permettant pas un consentement éclairé, celui-ci a été donné par le représentant du patient.
2.2.1
Recueil et analyse des données
2.2.1.1
Variables explicatives (facteurs pronostiques initiaux)
Les variables étudiées comprenaient le sexe, l’âge, le score de coma de Glasgow initial, l’état pupillaire, le nombre d’ACSOS, la durée du coma, la durée de l’amnésie post-traumatique. Afin d’éviter un facteur de confusion relatif aux traitements sédatifs des patients sous intubation, le score de coma de Glasgow pris en compte était celui établi lors de la prise en charge par le SAMU ; même si celui-ci était supérieur à 8. L’état pupillaire était qualifié d’aréactif lorsqu’au moins une pupille n’était pas réactive. Plusieurs ACSOS ont été recherchées de façon systématique : hypotension artérielle systolique inférieure à 90 mmHg ; hypoxie définie par une saturation en oxygène inférieure à 90 % ; anémie aiguë avec un taux d’hématocrite inférieur à 26 %. Le nombre d’ACSOS survenues au cours de la première semaine a été calculé.
La durée du coma était définie par le délai entre le TC et l’apparition d’une réponse motrice aux ordres simples. La durée de l’amnésie post-traumatique était définie par le délai entre le TC et la reprise d’une fixation mnésique, évaluée de façon rétrospective à partir des données disponibles dans le dossier médical, par l’interrogatoire du patient et de son entourage. Celle-ci est renseignée en nombre de jours et un seuil maximal à 90j a été établi pour les patients présentant un trouble de conscience persistant .
L’évaluation neuroradiologique comprenait le score de Marshall sur le scanner réalisé à la période initiale et une analyse en IRM sur les séquences axiales T1, T2, T2 Flair, T2 écho de gradient (T2*), diffusion et sur la séquence sagittale T1. Lorsque plusieurs IRM ont été réalisés chez le même patient, l’IRM la plus précoce a été analysée. Les critères lésionnels retenus étaient la présence d’une hydrocéphalie, le nombre de lésions axonales diffuses (LAD) total (toutes séquences confondues) et en T2* (LAD hémorragiques). Pour chaque patient, une analyse de la localisation des lésions a été réalisée par un neuroradiologue, ne connaissant pas le devenir des patients. Les régions anatomiques suivantes ont été analysées de façon systématique : lame tectale, mésencéphale, protubérance, bulbe, thalami, noyaux lenticulaires, corps calleux, lobes temporaux, frontaux, pariétaux et occipitaux.
Le bilan électrophysiologique comprenait un enregistrement des potentiels évoqués auditifs précoces (PEAp), de latences moyennes (PEAlm), tardifs et cognitifs avec l’onde N100 et la MMN. La présence de la composante P300 attentionnelle était précisée lorsque celle-ci a été recherchée. La classification a été effectuée à partir des graduations standardisées utilisées pour les potentiels évoqués du tronc cérébral et de latences moyennes . De façon à simplifier les analyses, la classification utilisée était binomiale : pour les potentiels évoqués auditifs précoces, les grades 1 (normal) et 2 (augmentation de la latence entre les pics I et V sans diminution d’amplitude) ont été classés dans la catégorie « normal » ; les grades 3 (amplitude du ratio V/I < 0,5), 4 (pic IV ou V non détecté) et 5 (seul le pic I est présent) dans la catégorie « anormal ». Pour les potentiels auditifs de latences moyennes, la catégorie « normal » correspondait à la présence de l’onde Na et Pa de façon uni- ou bilatérale ; la catégorie « anormal » était définie par l’abolition de l’onde Pa de façon bilatérale. Les potentiels évoqués tardifs N100, MMN et P300 étaient considérés comme présents lorsqu’une réponse ou une ébauche de réponse était retrouvée. En cas d’asymétrie entre les deux hémisphères, la meilleure réponse était retenue.
2.2.1.2
Variables à expliquer (données recueillies entre 2 et 4 ans après le TC)
Pour chaque patient, une évaluation standardisée des déficiences cognitives et neurocomportementales a été réalisée. Le bilan neuropsychologique comprenait une évaluation de la mémoire avec la batterie d’efficience mnésique (BEM) de Signoret version courte de 84 épreuves regroupées en 7 catégories selon le type d’atteinte mnésique. Pour chaque catégorie, si le score normalisé était supérieur à 1,6 écarts-types, il était considéré comme pathologique et comptabilisait un point. Le score total pouvait donc varier de 0 (absence de déficit mnésique) à 7 (amnésie totale). Pour évaluer les capacités attentionnelles, quatre épreuves de la batterie TEA (test d’évaluation de l’attention) ont été effectuées : le test d’alerte phasique, le test de mémoire de travail, l’épreuve du Go No-Go, le test de flexibilité. Chaque épreuve était considérée comme pathologique si elle n’était pas réalisée ou si le score T de la médiane des temps de réaction, des erreurs et/ou des omissions était inférieur à 30 ; correspondant à –2 écarts-types par rapport à la médiane. Une épreuve pathologique comptabilisait un point ; le score total était donc sur 4. L’existence d’un syndrome frontal était recherchée avec la batterie rapide d’efficience frontale (BREF) de Dubois et al., avec un seuil pathologique inférieur ou égal à 15. Une évaluation plus globale a été effectuée en utilisant l’échelle neurocomportementale révisée, NRS-r comportant 29 items . La recherche d’un syndrome dépressif par l’inventaire de dépression de Beck était effectuée. Un syndrome dépressif était retenu si le score sur l’inventaire de Beck était supérieur ou égal à 20.
L’évaluation des limitations d’activité a été effectuée par le même médecin lors de l’examen clinique ou de l’interrogatoire du devenir en utilisant le Glasgow Outcome Scale (GOS) : GOS V, bonne récupération ; GOS IV, incapacité modérée ; GOS III, incapacité sévère ; GOS II, état végétatif ; GOS I, décès ; et la mesure de l’indépendance fonctionnelle . La participation sociale du patient a été évaluée en recherchant son lieu de vie, son statut marital et professionnel. Le questionnaire QOLBI a permis d’évaluer la qualité de vie du patient estimée par le patient lui-même et par son accompagnant. Le score obtenu sur un total de 350 points (satisfaction maximale) a été rapporté en pourcentage pour l’analyse statistique.
2.2.1.3
Analyse des données
Une analyse statistique descriptive a été effectuée pour chaque variable. Une analyse de deuxième niveau a été réalisée pour rechercher des corrélations entre variables explicatives et à expliquer. Les corrélations ont été étudiées pour les variables explicatives cliniques (GCS, nombre d’ACSOS, durée du coma et de l’amnésie post-traumatique) et pour les variables explicatives paracliniques, électrophysiologiques et radiologiques (nombre de lésions axonales diffuses totales et hémorragiques à l’IRM). Ces éléments ont été corrélés aux données de l’évolution à long terme (GOS, MIF, NRS, QOLBI, évaluation des fonctions attentionnelles et mnésiques et de la dépression). Pour les variables quantitatives, le coefficient de corrélation de rang de Spearman a été calculé. Pour effectuer des corrélations avec les données à expliquer qualitatives, un test de rang de type Wilcoxon–Mann–Whitney a été utilisé. Lorsque les données explicatives et à expliquer étaient binaires, un test de Chi 2 a été effectué.
En outre, nous avons recherché si certaines régions cérébrales étaient plus spécifiquement lésées selon la nature des troubles cognitifs : recherche d’une lésion temporale chez les patients ayant un trouble mnésique ; recherche d’une lésion frontale ou pariétale chez les patients ayant un trouble attentionnel. Les corrélations ont été spécifiquement recherchées et étudiées entre la localisation lésionnelle frontale, pariétale, temporale et les résultats des tests d’attention et de mémoire.
2.3
Résultats
2.3.1
Population
Parmi les 151 patients ayant été hospitalisés en réanimation neurologique pour un traumatisme crânien sévère de janvier 2007 à juillet 2008, 36 ont bénéficié d’une évaluation électrophysiologique ; 22 d’entre eux avaient également effectué une imagerie par IRM ; 2 de ces patients sont décédés au cours de la période sub-aiguë post-traumatique ; un patient a refusé de participer à l’étude ; un patient a été perdu de vue avant l’évaluation du devenir. Dix-huit patients répondaient donc aux critères d’inclusion. Douze patients ont pu être examinés à distance du traumatisme et huit d’entre eux ont réalisé le bilan neuropsychologique. Parmi les 4 patients qui n’ont pas pu bénéficier de l’évaluation cognitive, un patient a refusé l’évaluation alors que les trois autres présentaient un trouble de conscience ou un handicap de communication trop important. Cinq patients ont été contactés par téléphone car ils n’ont pas pu être examinés en raison de l’éloignement géographique ou de l’impossibilité de déplacement. Un patient était décédé lors de l’évaluation à long terme.
La cause principale du traumatisme était un accident de la voie publique ( n = 12, 67 %). Le TC était secondaire à une chute dans 5 cas (28 %) et dans un cas (6 %), il s’agissait d’une agression par un objet contondant.
Lors de la prise en charge initiale, sept patients (39 %) avaient un score de coma de Glasgow initial supérieur à 8 mais l’état de conscience s’était dégradé au cours des 48 premières heures suivant le traumatisme crânien. Huit patients (42 %) avaient au moins une pupille aréactive lors de la prise en charge initiale. Neuf patients avaient nécessité une prise en charge neurochirurgicale pour l’évacuation d’une lésion de masse ( Tableaux 1a et 1b ).
Related posts:
The Annals of PRMkeep on moving forward
Superimposed electrical stimulation improves mobility of pre-stiff thumbs after ulnar collateral ligament injury of the metacarpophalangeal joint: A randomized study
Contribution of behavioral and cognitive therapy to managing functional urinary disorders in woman
Feasibility of the Cognitive Assessment scale for Stroke Patients (CASP) vs. MMSE and MoCA in aphasic left hemispheric stroke patients
Does adolescent idiopathic scoliosis relate to vestibular disorders? A systematic review
The Functional Reach Test: Strategies, performance and the influence of age
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
