A multi-center analysis of rehospitalizations five years after brain injury.

OBJECTIVE: To investigate the incidence and cause of rehospitalizations 1 and 5 years after traumatic brain injury. DESIGN: Descriptive statistics were computed in a prospective study of the cause and incidence of rehospitalizations at years 1 and 5 after injury. Analysis of variance and chi-square tests were used to identify factors relating to rehospitalization; factors included length of stay, admission and discharge functional status, payer source, medical complications, injury severity, and demographics. SETTING: Seventeen medical centers in the federally sponsored Traumatic Brain Injury Model Systems. In each setting, the continuum of care includes emergency medical services, intensive and acute medical care, inpatient rehabilitation, and a spectrum of community rehabilitation services. PARTICIPANTS: Eight hundred ninety-five rehabilitation patients admitted to acute care within 24 hours of traumatic brain injury between 1989 and 1999 were examined at 1-year follow-up. MAIN OUTCOME MEASURES: Incidence and cause of rehospitalization at 1 and 5 years after injury. RESULTS: The incidence of rehospitalization ranged from 22.9% 1 year after injury to 17.0% at 5 years after injury. At 1 year after injury, a third of the rehospitalizations were for elective reasons. At 5 years after injury, the incidence of readmissions for seizures and psychiatric difficulties and general health maintenance increased substantially. T-tests and chi-square analyses were performed on the 5-year follow-up data to compare those rehospitalized for unplanned reasons with those not rehospitalized. CONCLUSIONS: There remains a relatively high rate of rehospitalization in the long term after traumatic brain injury. The costs of rehospitalization should be considered when evaluating the long-term consequences of injury.

Long-term neuropsychological outcome after traumatic brain injury.

OBJECTIVE: To describe neuropsychological outcome 5 years after injury in persons with traumatic brain injury (TBI) who received inpatient medical rehabilitation. To determine the magnitude and pattern neuropsychological recovery from 1 year to 5 years after injury. DESIGN: Longitudinal cohort study with inclusion based on the availability of neuropsychological data at 1 year and 5 years after injury. SETTING: National Institute on Disability and Rehabilitation Research Traumatic Brain Injury Model Systems of Care. PARTICIPANTS: One hundred eighty-two persons with complicated mild to severe traumatic brain injury. PRIMARY OUTCOME MEASURES: Digits Forward and Backward, Logical Memory I and II, Token Test, Controlled Oral Word Association Test, Symbol Digit Modalities Test, Trail Making Test, Rey Auditory Verbal Learning Test, Visual Form Discrimination, Block Design, Wisconsin Card Sorting Test, and Grooved Pegboard. RESULTS: Significant variability in outcome was found 5 years after TBI, ranging from no measurable impairment to severe impairment on neuropsychological tests. Improvement from 1 year after injury to 5 years was also variable. Using the Reliable Change Index, 22.2% improved, 15.2% declined, and 62.6% were unchanged on test measures. CONCLUSIONS: Neuropsychological recovery after TBI is not uniform across individuals and neuropsychological domains. For a subset of persons with moderate to severe TBI, neuropsychological recovery may continue several years after injury with substantial recovery. For other persons, measurable impairment remains 5 years after injury. Improvement was most apparent on measures of cognitive speed, visuoconstruction, and verbal memory.

Long-term maintenance of gains obtained in postacute rehabilitation by persons with traumatic brain injury.

OBJECTIVE: To investigate maintenance of gains after discharge from a postacute rehabilitation program. DESIGN: Longitudinal cohort study, with inclusion based on availability of subjects at three time points. SETTING: Comprehensive postacute rehabilitation program in the Southern United States. PARTICIPANTS: Thirty-four persons with medically documented complicated mild to severe traumatic brain injury; primarily male Caucasians with some college. MAIN OUTCOME MEASURES: Disability Rating Scale and the Community Integration Questionnaire completed at admission, discharge, and two follow-up time points. RESULTS: Repeated measures analyses, using time from injury to discharge as a covariate, revealed significant improvements on all measures from admission to discharge, with no significant change from discharge to either of the follow-up periods. However, substantial changes were noted in individual cases. CONCLUSIONS: The results indicate that gains made by persons with traumatic brain injury during postacute rehabilitation are generally maintained at long-term follow-up, but changes occur in individual cases. Long-term services may help prevent decline in individual cases.

Using early neuropsychologic testing to predict long-term productivity outcome from traumatic brain injury.

OBJECTIVE: To evaluate whether early neuropsychologic testing is useful in predicting long-term productivity outcome after traumatic brain injury (TBI). DESIGN: Validation cohort prediction study. SETTING: Four inpatient brain injury rehabilitation programs participating in the Traumatic Brain Injury Model Systems project. PARTICIPANTS: A total of 293 adults with nonpenetrating TBI. MAIN OUTCOME MEASURES: Fifteen neuropsychologic tests were administered to patients who emerged from posttraumatic amnesia before rehabilitation discharge. Test scores were classified in the normal range or impaired range, using objective criteria. Outcome was defined as productive if the patient was competitively employed or enrolled full time in regular education. RESULTS: Productivity at follow-up was predicted by completion of at least 1 neuropsychologic test before discharge, by an injury-test interval of less than 2 months, and by normal range scores on 10 of the 15 neuropsychologic tests. Normal range scores on these tests increased the probability of a productive outcome by 40% to 130%. CONCLUSIONS: Neuropsychologic testing can help predict long-term productivity even when performed before discharge from inpatient rehabilitation and at variable injury-test intervals. Early testing should be interpreted in relation to injury-test interval. Because tests of multiple neuropsychologic domains predicted outcome, comprehensive evaluations might be more useful in predicting outcome.

Neuropsychological assessment and the Disability Rating Scale (DRS): a concurrent validity study.

Assessment of current level of functioning among clients with traumatic brain injury (TBI) often guides the establishment of realistic outcome goals for post-acute rehabilitation. Further, data generated from neuropsychological testing provide a clinician with a better understanding of a client's pattern of cognitive strengths and weaknesses. The Disability Rating Scale (DRS) is commonly used by TBI rehabilitation facilities to assess a client's general level of functioning in terms of impairment, disability, and handicap. Previous studies have used clients' neuropsychological test results to predict future level of functioning. These studies have shown mixed results regarding the predictive validity of the test findings; however, they usually employ only a limited number of tests for prediction representing a limited number of cognitive domains. Using a concurrent validity design, the present study investigated the bivariate associations between various neuropsychological testing domains (i.e. intellectual, academic, language, visuoperceptual, memory, and executive functioning) and level of functioning as indexed by the DRS. Participants were administered the DRS and the neuropsychological evaluation during the initial part of post-acute rehabilitation. Composite scores were derived for each of the neuropsychological domains. Most participants were categorized as sustaining a severe TBI. The mean age and education of this predominately male sample was 28.84 years (SD = 9.13) and 11.83 years (SD = 1.7), respectively. Results revealed a significant positive relationship between performances in intellectual, executive, academic, and visuoperceptual domains and level of functioning.

Predictive validity of the Neurobehavioural Cognitive Status Examination (NCSE) in a post-acute rehabilitation setting.

Within the context of a post-acute rehabilitation setting, association and agreement between results from the Neurobehavioural Cognitive Status Examination (NCSE) and from the neuropsychological (NP) evaluation are examined. All participants (n = 48) had sustained a severe traumatic brain injury and NCSE testing preceded NP testing by an average of 1 month. A significant relationship and fair classification agreement (i.e. presence or absence of cognitive impairment) was found between the overall results from NCSE and NP evaluation. Significant relationships were also observed between most NCSE subtests and paired NP tests thought to be assessing the same cognitive domains. However, the classification agreement (i.e. the presence or absence of deficient performance) between most NCSE subtests and paired NP tests was poor. The findings are discussed from the standpoint of individual treatment planning.

The Community Integration Questionnaire revisited: an assessment of factor structure and validity.

OBJECTIVE: To investigate the factor structure and concurrent validity of the Community Integration Questionnaire (CIQ), using a large sample of persons with traumatic brain injury (TBI). DESIGN: Principal components analysis with varimax rotation was performed on CIQ items completed through interview with patients at 1 year after injury. Correlational analyses compared CIQ scores to scores on other widely used outcome measures. SETTING: Outpatient clinics affiliated with four TBI Model System rehabilitation centers funded by the National Institute on Disability and Rehabilitation Research. PARTICIPANTS: Three hundred twelve patients with medically documented TBI who were enrolled in the TBI Model Systems Project. The majority of patients were Caucasian males with severe TBI. MAIN OUTCOME MEASURES: CIQ; Functional Independence Measure (FIM); Functional Assessment Measure (FAM); Disability Rating Scale (DRS). RESULTS: Three factors emerged: Home Competency, Social Integration, and Productive Activity. The financial management item was moved from Social Integration to Home Competency, and the travel item was moved from Productive Activity to Social Integration. Each CIQ scale score showed significant correlations in the expected direction with the FIM+FAM and DRS items. CONCLUSIONS: The results provide further evidence for the validity of the CIQ and improve the scoring system. The factor structure is clinically and theoretically meaningful. The subscale and total scores show significant relationships with other widely used measures of outcome. Future research should focus on increasing the range of questions, accounting for changes from preinjury functioning, and obtaining normative data on the new factors.

Etiology and incidence of rehospitalization after traumatic brain injury: a multicenter analysis.

OBJECTIVE: To investigate incidence and etiology of rehospitalizations at 1, 2, and 3 years after traumatic brain injury. DESIGN: Descriptive statistics were computed in a prospective study of etiology and incidence of rehospitalization at years 1, 2, and 3 postinjury. Analysis of variance (ANOVA) and chi2 were used to identify factors relating to rehospitalization; factors included length of stay, admission and discharge functional status, payer source, medical complications, injury severity, and demographics. SETTING: Four medical centers in the federally sponsored Traumatic Brain Injury Model Systems. In each setting, the continuum of care includes emergency medical services, intensive and acute medical care, inpatient rehabilitation, and a spectrum of community rehabilitation services. PARTICIPANTS: Six hundred sixty-five rehabilitation patients admitted to acute care within 24 hours of traumatic brain injury between 1989 and 1996. MAIN OUTCOME MEASURES: Annual incidence and etiology of rehospitalization. RESULTS: The annual incidence of rehospitalization ranged from 20% to 22.5%. Approximately half the rehospitalizations were for elective reasons. The most common reason for rehospitalization was for orthopedic or reconstructive surgery, followed by infectious disorders and general health maintenance. After the first year, the incidence of readmissions for seizures and psychiatric difficulties increased substantially. ANOVA and chi2 analyses were performed on data from the first year postinjury. No statistically significant associations were noted between incidence and etiology of rehospitalization and: demographics; injury severity; payer source for rehabilitation; concurrent injuries; acute care and rehabilitation length of stays; discharge Functional Assessment Measure; and discharge residence (p > .05). CONCLUSIONS: There is a relatively stable but high rate of rehospitalization for at least 3 years after injury. The costs of rehospitalization should be considered when evaluating the long-term consequences of injury.

Impaired awareness and employment outcome after traumatic brain injury.

Inaccurate self-awareness is a common finding after traumatic brain injury. Such impaired awareness has been hypothesized to limit patients' eventual functional outcomes by decreasing motivation for treatment and resulting in selection of inappropriate long-term goals. Previous investigations of the association between impaired awareness and employment outcome have produced inconsistent results. The present article reviews these studies and presents the results of our new investigation of this issue. In addition, we studied the comparability of two methods of measuring impaired awareness. Results provided strong support for a positive relationship between accurate self-awareness and favorable employment outcome at follow-up.

Characteristics of impaired awareness after traumatic brain injury.

Impaired awareness of the effects of brain injury is a commonly observed and poorly understood finding in traumatic brain injury survivors. Nonetheless, impaired awareness has been identified as a major factor in determining outcome for traumatic brain injury survivors. Review of previous studies of impaired awareness in this patient population revealed a number of preliminary findings regarding the nature of this phenomenon. The present paper presents the results of 2 new studies with a total of 111 traumatic brain injury patients conducted to bring further clarity to this area. Findings confirmed and extended many results of previous investigations. Specific findings included patient overestimation of functioning as compared to family member ratings, patient report of greater physical than nonphysical impairment, greater patient-family agreement on specific ratings of patient functioning than on general ratings, greater agreement of family and clinician ratings of patient functioning with each other than with patient self-ratings, and partial disagreement of different methods of measuring impaired awareness.

Assessment and treatment of impaired awareness after brain injury: implications for community re-integration.

Impaired self-awareness of deficits is a common finding in patients who have suffered traumatic brain injury. Impaired awareness can limit motivation for treatment and contribute to poor outcome. Consequently, it is important for brain injury rehabilitation professionals to understand this phenomenon and utilize treatment approaches that may improve patient awareness. The present article reviews the existing literature on measurement of impaired awareness, characteristics of impaired awareness, the relationship of impaired awareness to functional outcome, possible treatment approaches for impaired awareness and empirical investigations of interventions to improve awareness. The treatment strategies we use to address impaired awareness in our community re-integration program for brain injury survivors are described in detail. These approaches include: establishment of the therapeutic alliance, family interventions, peer feedback, education, roleplaying, videotape feedback, real world experiences, therapeutic milieu and psychotherapy.

Agreement between persons with traumatic brain injury and their relatives regarding psychosocial outcome using the Community Integration Questionnaire.

OBJECTIVE: To extend psychometric research on the Community Integration Questionnaire (CIQ) by comparing patients' reports with those of family members. DESIGN: Reports on community integration were obtained from family members and patients 1 year after brain injury. Kappa reliability coefficients were used to examine agreement for individual items on the questionnaire. Total scores for the three CIQ scales were compared between groups using independent samples t tests. SETTING: Outpatient clinics affiliated with the four NIDRR traumatic brain injury Model System rehabilitation centers. PARTICIPANTS: CIQ data were obtained for 259 adult patients with traumatic brain injury. For 122 cases, patients and family members provided corresponding data. In 101 cases, only self-reported data were available. In 36 cases, information was provided solely by family members. All patients received inpatient rehabilitation and were at least 16 years of age. The full range of brain injury severity was represented. MAIN OUTCOME MEASURE: The 15-item CIQ, comprised of three scales, Home Integration, Productive Activity, and Social Integration. RESULTS: Analysis of kappa values found "moderate" to "almost perfect" agreement levels for all 15 items. Almost-perfect ratings were found for three of the four items on the Productive Activity Scale. Comparisons were made between scale scores derived from patients and families. The Home Integration scale showed differences, with patient scores higher than those reported by family members (t = 3.51, p < .01). However, the difference in scores was attributable to small discrepancies on two items and was not considered clinically meaningful. The Total CIQ score also showed a difference, with patients reporting significantly higher levels of integration relative to family members (t = 2.30, p < .05). This difference was also attributable to discrepancies on two items of the Home Integration Scale. CONCLUSIONS: The results have important implications for researchers who may be limited to one source of data. The hypothesis that patients and family members have very different perspectives of patients' community activities was not supported. Rather, similar ratings of individual activities were given with two exceptions, meal preparation and housekeeping. More research is needed to delineate relationships between agreement ratings and outcome categories.

Predictors of psychological health in caregivers of patients with closed head injury.

The purpose of the current study was to investigate the contribution of coping strategies, subjective burden, and social support to psychological health in caregivers. The sample included 69 primary caregivers of patients with severe closed head injuries. There were three groups of caregivers: acute (0-6 months post-injury); intermediate (6 months-1.5 years); and long-term (> 1.5 years). All had participated in rehabilitation. Caregivers completed an interview and series of questionnaires, including the Ways of Coping Questionnaire, Social Support Questionnaire, Subjective Burden Measure, and General Health Questionnaire. The Disability Rating Scale was completed by staff to assess patients' level of functioning at the time of caregivers' assessment. ANOVA revealed no between-group differences in coping style or social support. Multiple regression revealed that greater use of emotion-focused coping was associated with greater emotional distress. Coping style contributed to a greater proportion of the variance in caregivers' psychological health that did patients' level of functioning. Increased satisfaction with social support was associated with less emotional distress. The full model, including group, caregiver gender, emotion-focused coping, social support, patient level of recovery, burden, and the burden x coping interaction accounted for over half of the variance in psychological health. Results support a multidimensional model for explaining caregivers' adjustment.

Functional outcome from traumatic brain injury: unidimensional or multidimensional?

This study compared multidimensional and unidimensional measures of functional outcome from traumatic brain injury in a referred sample of 67 persons with moderate to severe traumatic brain injury who resided in the community or in a nonhospital facility. Four outcome indicators (independence in self-care, independence in travel, paid employment, and friendship) were correlated with the patients' ratings on two unidimensional outcome measures, the Disability Rating Scale (DRS) and Glasgow Outcome Scale (GOS), and on a multidimensional outcome measure, the Craig Handicap Assessment and Reporting Technique (CHART). The results showed that the CHART Occupation and Social Integration scales were associated with different outcome indicators, controlling for overall outcome as measured by the DRS and GOS. DRS and GOS ratings were associated more strongly with the self-care and travel indicators than with the employment or friendship indicators. The results support a multidimensional model of functional outcome from brain injury, suggest that the CHART may measure some outcome dimensions more specifically than do the DRS and GOS, and point to limitations of unidimensional outcome scales.

A comparison of psychiatric symptoms in vascular dementia and Alzheimer's disease.

OBJECTIVE: Psychiatric symptoms account for much of the morbidity of vascular dementia and Alzheimer's disease. The goals of this study were to extend previous observations of the psychopathology and behavioral problems associated with vascular dementia and to compare the profile of symptoms in patients with vascular dementia to that in patients with Alzheimer's disease. METHOD: Twenty-eight pairs of patients (one with vascular dementia and one with Alzheimer's disease) were matched with respect to education, age, and severity of dementia. Their psychiatric symptoms were assessed with the Neurobehavioral Rating Scale, a 28-item observer-rated instrument, and the Hamilton Depression Rating Scale, and the symptoms in the two diagnostic groups were compared. RESULTS: Blunted affect, depressed mood, emotional withdrawal, motor retardation, low motivation, anxiety, unusual thoughts, and somatic concerns occurred in more than one-third of the patients with vascular dementia. There was no significant relation between severity of cognitive impairment and severity of these noncognitive symptoms. The patients with vascular dementia had more impairment than the patients with Alzheimer's disease, as indicated by the Neurobehavioral Rating Scale total scores and scores on the behavioral retardation, anxiety/depression, and verbal output disturbance factors. They also had a higher total score on the Hamilton depression scale and higher scores on 14 of the 17 Hamilton depression items. CONCLUSIONS: Patients with vascular dementia have more severe behavioral retardation, depression, and anxiety than those with Alzheimer's disease when the groups have similar levels of cognitive impairment. This probably reflects the contrasting brain regions typically involved in the two disorders.

Assessment of cognitive, psychiatric, and behavioral disturbances in patients with dementia: the Neurobehavioral Rating Scale.

OBJECTIVE: To assess the validity of the Neurobehavioral Rating Scale (NRS) in patients with Alzheimer's disease (AD) or multi-infarct dementia (MID) and to characterize the cognitive, psychiatric, and behavioral disturbances that occur in these patients. DESIGN: Cross-sectional evaluation. SETTING: West Los Angeles VAMC Geropsychiatry Inpatient Unit, Neurobehavior Inpatient Unit, and Dementia Clinic; UCLA Alzheimer's Disease Clinic. PATIENTS: Convenience sample of 61 patients with AD and 22 patients with MID. MAIN OUTCOME MEASURE: The NRS, a 27-item observer-rated instrument that measures cognitive, psychiatric, and behavioral disturbances. RESULTS: The NRS demonstrated content and convergent validity in this patient group. Principal components analysis of the NRS ratings identified a six-factor solution, and each factor contained clinically related symptoms. The factors were Cognition/Insight, Agitation/Disinhibition, Behavioral Retardation, Anxiety/Depression, Verbal Output Disturbance, and Psychosis. Among the patients with AD, agitation, disinhibition, hostility, poor insight, poor motivation, suspiciousness, and delusions were more severe in patients with more advanced dementia. Depressive symptoms occurred with equal severity in patients with mild and advanced dementia, but depressed mood was more severe in patients with earlier age of onset of AD. CONCLUSION: The NRS is a useful instrument for structured assessment of a broad range of cognitive, psychiatric, and behavioral disturbances in patients with dementia.

Serial MRI and neurobehavioural findings after mild to moderate closed head injury.

Fifty patients who sustained mild to moderate closed head injury (CHI) underwent a CT scan, MRI, and neurobehavioural testing. At baseline 40 patients had intracranial hyperintensities detected by MRI which predominated in the frontal and temporal regions, whereas 10 patients had lesions detected by CT. Neurobehavioural data obtained during the first admission to hospital disclosed no distinctive pattern in subgroups of patients characterised by lesions confined to the frontal, temporal, or frontotemporal regions, whereas all three groups exhibited pervasive deficits in relation to normal control subjects. The size of extraparenchymal lesion was significantly related to the initial Glasgow Coma Scale score, whereas this relation was not present in parenchymal lesions. One and three month follow up MRI findings showed substantial resolution of lesion while neuropsychological data reflected impressive recovery. The follow up data disclosed a trend from pervasive deficits to more specific impairments which were inconsistently related to the site of brain lesion. These results corroborate and extend previous findings, indicating that intracranial lesions detected by MRI are present in most patients hospitalised after mild to moderate CHI. Individual differences in the relation between site of lesion and the pattern of neuropsychological findings, which persist over one to three months after mild to moderate CHI, remain unexplained.

Neurobehavioral outcome 1 year after severe head injury. Experience of the Traumatic Coma Data Bank.

The outcome 1 year after they had sustained a severe head injury was investigated in patients who were admitted to the neurosurgery service at one of four centers participating in the Traumatic Coma Data Bank (TCDB). Of 300 eligible survivors, the quality of recovery 1 year after injury was assessed by at least the Glasgow Outcome Scale (GOS) in 263 patients (87%), whereas complete neuropsychological assessment was performed in 127 (42%) of the eligible survivors. The capacity of the patients to undergo neuropsychological testing 1 year after injury was a criterion of recovery as reflected by a significant relationship to neurological indices of acute injury and the GOS score at the time of hospital discharge. The neurobehavioral data at 1 year after injury were generally comparable across the four samples of patients and characterized by impairment of memory and slowed information processing. In contrast, language and visuospatial ability recovered to within the normal range. The lowest postresuscitation Glasgow Coma Scale (GCS) score and pupillary reactivity were predictive of the 1-year GOS score and neuropsychological performance. The lowest GCS score was especially predictive of neuropsychological performance 1 year postinjury in patients who had at least one nonreactive pupil following resuscitation. Notwithstanding limitations related to the scope of the TCDB and attrition in follow-up material, the results indicate a characteristic pattern of neurobehavioral recovery from severe head injury and encourage the use of neurobehavioral outcome measurements in clinical trials to evaluate interventions for head-injured patients.

Recovery of orientation following closed-head injury.

The pattern of recovery of orientation to person, place, and time was investigated in 84 patients who were initially disoriented while hospitalized on the neurosurgery service after sustaining a closed-head injury (CHI) of varying severity. Results of daily administration of the Galveston Orientation and Amnesia Test revealed that the most common sequence of recovery of orientation was person, place, and time: accounting for about 70% of the patients. Return of orientation to time preceded reorientation to place in 13% of the patients while other orderings of reorientation were present in 11% of the cases. Temporal disorientation was initially characterized by backward displacement of the date from the actual date in 68% of the patients. The magnitude of this displacement progressively shrank as the patients became more oriented. Patients who exhibited the greatest backward displacement of the date had more severe and persistent impaired consciousness, were older and had longer durations of posttraumatic amnesia. These data support Ribot's hypothesis that older memories are relatively resistant to cerebral insult.

Dichotic listening and manual performance in relation to magnetic resonance imaging after closed head injury.

In order to investigate post-traumatic hemispheric disconnection effects, dichotic listening and intermanual tasks were administered to 69 patients who had sustained a closed head injury of varying severity. The manual tasks consisted of naming objects palpated in either hand, transfer of postures from one hand to the other and writing. Consistent with predictions, the degree of ear asymmetry in dichotic listening performance was directly related to the severity of the head injury as reflected by the degree of impaired consciousness. Depth and localisation of parenchymal lesion characterised by magnetic resonance imaging were also related to the degree of ear asymmetry. Parenchymal lesions situated in sites which could potentially interfere with callosal auditory or geniculocortical pathways produced a greater disparity in response to left versus right ear inputs as compared with parenchymal lesions in areas such as the frontal lobes which are purportedly unrelated to asymmetries in dichotic listening performance. The results provide further evidence for the effects of multifocal brain lesions involving the white matter on tasks which require intra and/or interhemispheric integration.