CLINICIANS AND INVESTIGATORS have long recognized

Transcription

1 1788 ORIGINAL ARTICLE Co-Occurring Traumatic Brain Injury and Acute Spinal Cord Injury Rehabilitation Outcomes Stephen Macciocchi, hd, AB, Ronald T. Seel, hd, Adam Warshowsky, hd, Nicole Thompson, MH, Kimether Barlow, MH ABSTRACT. Macciocchi S, Seel RT, Warshowsky A, Thompson N, Barlow K. Co-occurring traumatic brain injury and acute spinal cord injury rehabilitation outcomes. Arch hys Med Rehabil 2012;93: Objective: To determine the impact of co-occurring traumatic brain injury (TBI) on functional motor outcome and cognition during acute spinal cord injury (SCI) rehabilitation. Design: rospective, longitudinal cohort. Setting: Single-center National Institute of Disability and Rehabilitation Research SCI Model System. articipants: ersons aged 16 to 59 years (N 189) admitted for acute SCI rehabilitation during the 18-month recruitment window who met inclusion criteria. Interventions: Not applicable. Main Outcome Measures: FIM Motor Scale (Rasch transformed) and acute rehabilitation length of stay (LOS). Results: In the tetraplegia sample, co-occurring TBI was not related to FIM Motor Scale scores or acute rehabilitation LOS despite having negative impacts on memory and problem solving. ersons with paraplegia who sustained co-occurring severe TBI had lower admission and discharge FIM Motor Scale scores and longer acute rehabilitation LOS than did persons with paraplegia and either no TBI or mild TBI. ersons with paraplegia and severe TBI had lower functional comprehension, problem solving, and memory and impairments on tests of processing speed compared with persons with paraplegia and no TBI, mild TBI, and moderate TBI. ersons with paraplegia and co-occurring mild and moderate TBI had equivalent acute rehabilitation motor outcomes and cognitive functioning compared with persons with paraplegia and no TBI. Conclusions: This study provides evidence that persons aged 16 to 59 years with paraplegia and co-occurring severe TBI had worse motor outcomes and longer acute rehabilitation LOS than did persons with paraplegia and no TBI. Impairments in processing speed, comprehension, memory, and problem solving may explain suboptimal motor skill acquisition. Research with larger samples is required to determine whether mild and moderate TBI impact acute rehabilitation motor outcomes and LOS. Key Words: Cognition; rognosis; Rehabilitation; Spinal cord injuries; Traumatic brain injuries by the American Congress of Rehabilitation Medicine CLINICIA AND INVESTIGATORS have long recognized that persons who sustain traumatic spinal cord injury (SCI) may also sustain co-occurring traumatic brain injury (TBI), which could limit acute rehabilitation functional gains. 1-5 Investigators initially focused on documenting the incidence of SCI and co-occurring TBI by examining markers for TBI such as International Classification of Diseases codes, neuroimaging results, loss of consciousness (LOC), and posttraumatic amnesia (TA). The reported incidence of SCI and co-occurring TBI varies on the basis of the diagnostic markers used. For instance, studies using International Classification of Diseases codes or positive imaging have found a 16% to 24% incidence of co-occurring TBI, but when TA is used as a marker for TBI, the incidence increases to 42% to 50%. 6 A recent prospective study 6 by using a diagnostic algorithm with multiple TBI markers found that 60% of persons with SCI sustained a co-occurring TBI. In this study, most co-occurring TBIs were mild in severity, but about 25% of persons with SCI sustain mild complicated, moderate, or severe TBI. Several investigations have examined the neuropsychological test performance of persons during acute SCI rehabilitation. 1,3,4,7-10 These studies documented impaired attention, processing speed, problem solving, learning, and memory in persons undergoing acute SCI rehabilitation. Three studies compared persons with SCI and co-occurring TBI with persons with SCI alone. The earliest study 1 (n 30) found a trend for impaired Halstead Category Test performance by persons with SCI and self-reported TBI compared with persons with SCI alone. A subsequent study 4 (n 150) found that persons with SCI who reported LOC had equivalent neuropsychological test scores compared with persons not reporting LOC. The most recent study 10 (n 18) found that persons with SCI and cooccurring TBI documented by positive neuroimaging had more impaired attention, processing speed, and memory than did persons with SCI alone. Two retrospective, matched, case-control studies 10,11 compared SCI acute rehabilitation functional outcomes between persons with and without co-occurring TBI. The first study 11 (n 82) matched pairs of individuals with SCI and co-occurring TBI (51% mild, 28% moderate, 21% severe) with persons with SCI alone on age, sex, education, and SCI level. ersons with SCI and TBI had less FIM Motor Scale From the Shepherd Center, Atlanta, GA(Macciocchi, Seel, Warshowsky, Thompson, Barlow); and Department of Kinesiology, University of Georgia, Athens, GA (Macciocchi). Supported by the National Institute on Disability and Rehabilitation Research, US Department of Education (grant no. H113G030004). 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. Reprint requests to Stephen Macciocchi, hd, AB, Shepherd Center, 2020 eachtree Rd, Atlanta, GA 30309, stephen_macciocchi@shepherd.org. In-press corrected proof published online on Apr 4, 2012, at /12/ $36.00/0 doi: /j.apmr ANCOVA AMS ASIA GCS LOC LOS TA SCI TBI List of Abbreviations analysis of covariance American Spinal Injury Association Motor Score American Spinal Injury Association Glasgow Coma Scale loss of consciousness length of stay posttraumatic amnesia spinal cord injury traumatic brain injury

2 SINAL CORD INJURY AND CO-OCCURRING TBI, Macciocchi 1789 improvement scores than did persons with SCI alone. 11 A second study 10 (n 20, admitted 2 6 months postinjury) matched persons with SCI and co-occurring TBI (positive neuroimaging) with persons with SCI alone on the general SCI level (eg, cervical, thoracic, or lumbar). ersons with SCI and TBI had equivalent FIM Motor Scale outcomes compared with persons with SCI alone. 10 While results from previous studies are mixed, the primary objective of our single-center study was to prospectively examine the relation between co-occurring TBI severity and acute SCI rehabilitation motor outcomes and length of stay (LOS). Secondary analyses were conducted on motor skills that require new leaning and sequencing, which could be sensitive to the effects of TBI. Last, we examined specific cognitive impairments related to TBI severity, which could explain suboptimal motor outcomes. METHODS articipants Our study was conducted in a National Institute on Disability and Rehabilitation Research SCI Model System Center that is designated as a long-term care hospital. ersons with SCI received individualized daily therapies from multidisciplinary teams composed of health care professionals in medicine, nursing, case management, physical therapy, occupational therapy, speech/language pathology, assistive technology, health education, therapeutic recreation, and psychology. ersons with SCI and co-occurring TBI were admitted to the SCI acute rehabilitation program once they had cleared TA and received cognitive therapies commensurate with their impairments. Consecutive persons with SCI aged 16 to 59 years admitted for acute rehabilitation secondary to trauma over an 18-month period ( ) were eligible for inclusion (n 266). Because of the unavailability of bilingual examiners and neuropsychological test norms, non English-speaking persons were not included. Fifty-four persons declined to participate, yielding an 80% consent rate. Of the 212 participants who enrolled, 14 persons withdrew prior to baseline assessment and 3 were discharged early and primary outcome variables could not be collected. An additional 6 participants were identified as outliers ( 3.5SD) in the number of days from injury to rehabilitation admission or acute rehabilitation LOS and were removed from analyses. The final sample (N 189) represented 89% of enrolled persons. rocedure Data collection. The host institution s Internal Review Board approved the study, and participants informed consent was obtained prior to data collection. articipants were enrolled within 1 week of SCI acute rehabilitation admission. Research coordinators interviewed participants and collected demographic and pre- and postinjury data. Acute medical surgical records were reviewed for TBI relevant data. Interdisciplinary rehabilitation team members who were blinded to participants baseline interview data completed admission and discharge FIM assessments. hysical therapists rated the admission American Spinal Injury Association (ASIA) Motor scores (AMSs). sychometrists blinded to study objectives administered neuropsychological tests approximately 1 to 4 weeks postrehabilitation admission (mean d). Measures. TBI was classified 6 by using estimated days to emerge from TA, worst Glasgow Coma Scale (GCS) score, and neuroimaging findings as follows: No evidence of TA or lesion No evidence of lesion and either 24 hours TA or GCS total score 13 to 14 or 11T Mild complicated TBI ositive neuroimaging and either TA 24 hours or GCS total score 13 to 14 or 11T TA 1 to 6 days or GCS total score of 9 to 12, 8T to 10T Severe TA 7 days or GCS total score of 3 to 8, 3T to 7T All persons with documented LOC also had documented evidence of TA. Mild complicated and moderate TBI classifications were combined and labeled moderate because of small cell sizes and research indicating similarity in postinjury outcomes. 12,13 The TBI classification system has criterion validity in the SCI population. 6 The acute rehabilitation AMS was used to quantify neurologic impairment. 14 The AMS is composed of 20 items rated on a 0- to 5-point scale that measures the functioning of 10 key upper and lower extremity muscles. Higher scores denote greater levels of preservation. The upper extremity AMS was used as a covariate in the tetraplegia sample for functional motor outcome analyses, and the lower extremity AMS was used as a covariate for the paraplegia sample analyses. 15 The FIM was used to measure levels of independence For statistical analyses, FIM Motor and Cognitive scales scores were converted to Rasch-transformed scores, which range from 0 to 100, with higher scores denoting greater levels of independence. 19 FIM Motor Scale change scores were calculated by subtracting the admission motor score from the discharge motor score. FIM Motor Scale efficiency scores were calculated by dividing FIM Motor Scale change scores by the acute rehabilitation LOS. Six FIM Motor Scale items bathing, lower extremity dressing, toileting, and chair, toilet, and tub/shower transfers were identified as tasks in which skill acquisition required sequencing. Three FIM items comprehension, problem solving, and memory were identified as cognitive functions most likely to affect motor skill acquisition. Neuropsychological tests not dependent on upper extremity motor skills were used to assess attention, working memory, learning, processing speed, and problem solving. These instruments included the Wechsler Adult Intelligence Scale-3rd edition Digit Span and Letter-Number Sequencing Tests, Continuous Visual Memory Test, Hopkins Verbal Learning Test (2nd Ed.), Symbol Digit Modalities Test Oral, and the Short Category Test The 6 tests are standardized, possess good psychometric properties, and are routinely used in clinical practice and research. Data Analyses SSS (version 15.0) a was used to analyze all data. The tetraplegia and paraplegia samples have different functional motor outcomes and were described and analyzed separately. Between-sample differences were analyzed by using t tests for continuous variables and Cramer s V for qualitative variables. Admission upper and lower extremity AMS scores were used to control for neurologic impairment in all motor outcome analyses and when identifying other potential covariates. Analysis of covariance (ANCOVA) was used to examine age, sex, race, education, risk indicators of preinjury cognitive impairment (brain injury, neurological disorder, learning problem, and held back in school), and health insurance payer type (managed care, Medicaid/Medicare, and Worker s Compensation) as potential covariates for motor outcomes and LOS. ANCOVA was used to examine TBI classification on days

3 1790 SINAL CORD INJURY AND CO-OCCURRING TBI, Macciocchi Table 1: Characteristics of Tetraplegia (n 96) and araplegia (n 93) Samples Characteristics Tetraplegia araplegia Age (y) Sex (men) 81% 75% Race/ethnicity White 65% 68% African American 34% 28% Other 1% 4% Health insurance payer type Managed care 59% 57% Medicaid/Medicare 32% 29% Worker s Compensation 6% 10% Charity 2% 4% Injury etiology.002 Motor vehicle collision 66% 59% Sporting injury 20% 7% Violence 8% 22% Falls/flying object 6% 13% Days from injury to rehabilitation admission Rehabilitation LOS AMS (admission) Rasch FIM Motor Scale scores Admission Discharge Change Efficiency Education (highest year completed) Risk indicators of preinjury cognitive impairment reviously sustained brain injury 7% 10% Diagnosed neurological disorder 5% 5% Learning problem 14% 14% Held back in school 29% 26% Brain injury presence/severity Not present 37% 45% Mild 34% 32% Mild complicated 14% 8% Moderate 6% 5% Severe 9% 10% Rasch FIM Cognitive Scale score (admission) Neuropsychological test scores Digit Span (SS) Letter-Number Sequencing (SS) Symbol Digit-Oral (T50) Hopkins Verbal Learning-Total Recall (T50) Hopkins Verbal Learning-Delayed Recall (T50) Continuous Visual Memory Test-Total (T50) Category Test (T50) NOTES. All reported statistics are mean SD or % of sample. All neuropsychological test scores are standardized: SS scale score (mean 10, 50th percentile; SD 3); T50 T score (mean 50, 50th percentile; SD 10). Bold indicates significant findings. Abbreviation:, not significant. from injury to acute rehabilitation admission, acute rehabilitation LOS, Rasch-transformed FIM Motor Scale total, change, and efficiency scores, and FIM Motor Scale items. Analysis of variance was used to examine TBI classification on FIM Cognitive Scale items and neuropsychological test performance. Findings were considered significant using an alpha level of.05. artial eta squared was computed to determine effect size, and Edgeworth series expansion was computed to determine power. Bonferroni post hoc tests, which reduce the risk of type I error associated with multiple comparisons, were used to identify between-group differences. RESULTS Demographic, Injury, Neurologic, and Cognitive Variables Table 1 provides descriptive data on participants age, sex, ethnicity, health insurance payer type, injury mechanism, risk indicators of preinjury cognitive impairment, and TBI classification. Means are provided for days from injury to rehabilitation admission, rehabilitation LOS, total admission AMS, FIM scores, education, and standard scores on neuropsychological tests. In the tetraplegia sample, 26% had a motor injury

4 SINAL CORD INJURY AND CO-OCCURRING TBI, Macciocchi 1791 Cognitive Measures Table 2: Effects of TBI on Cognition in Tetraplegia Sample artial (n 35) (n 33) (n 19) FIM items Comprehension (admission) a ( ) 6.6 ( ) 6.3 a ( ) 6.1 ( ) Comprehension (discharge) ( ) 6.9 ( ) 6.6 ( ) 6.9 ( ) roblem solving (admission) a ( ) 6.3 ( ) 6.0 ( ) 5.2 a ( ) roblem solving (discharge) ( ) 6.7 ( ) 6.4 ( ) 6.3 ( ) Memory (admission) a,b ( ) 6.6 c,d ( ) 5.6 a,c ( ) 4.9 b,d ( ) Memory (discharge) a ( ) 6.8 ( ) 6.3 a ( ) 6.7 ( ) Neuropsychological Attention (WAIS-3 Digit Span) ( ) 9.0 ( ) 8.7 ( ) 10.1 ( ) Working Memory (WAIS-3 L-N Seq.) ( ) 8.6 ( ) 8.3 ( ) 8.6 ( ) Auditory learning (HVLT Total Recall) ( ) 39.3 ( ) 40.3 ( ) 42.4 ( ) Auditory recall (HVLT Delayed Recall) ( ) 38.9 ( ) 38.8 ( ) 40.9 ( ) rocessing speed (SDMT-Oral) ( ) 34.7 ( ) 33.3 ( ) 32.7 ( ) NOTES. Analysis of variance used to determine main effects. Means (95% confidence intervals) are reported for the 4 TBI groups. a, b, c, d indicate that each TBI group with the same letter was different based on Bonferroni s test for multiple comparisons. Bold indicates significant findings. Abbreviations: HVLT, Hopkins Verbal Learning Test; L-N Seq., Letter-Number Sequencing; SDMT, Symbol Digit Modalities Test; WAIS-3, Wechsler Adult Intelligence Scale-3rd edition. level of C1-4 and 63% had an ASIA impairment score of A or B. In the paraplegia sample, 57% had a motor injury level of T2-8, 33% had a motor injury level of T9-12, and 10% had a motor injury level of L1-3; 72% had an ASIA impairment score of A or B. All participants were oriented on the basis of Orientation Log scores greater than While the mean education level was slightly above the 12th grade for both samples, participants mean performance on all neuropsychological tests was approximately 1 SD below age- and educationcorrected normative scores. Co-Occurring TBI and Tetraplegia Acute Rehabilitation Outcomes Covariates. Admission upper extremity AMS was entered as a covariate in ANCOVAs for days from injury to rehabilitation admission, rehabilitation LOS, and FIM Motor Scale scores and was statistically significant for all FIM Motor Scale score analyses. Medicaid/Medicare payer type was entered as a covariate and was significant in the ANCOVA for acute rehabilitation LOS. Upper extremity AMS and ASIA impairment scores did not differ between TBI classification groups. Main effects. For persons with tetraplegia, TBI was not related to time from injury to acute rehabilitation admission; acute rehabilitation LOS; and admission and discharge FIM Motor Scale total, change, and efficiency scores. TBI was related to admission FIM comprehension and problem-solving scores and admission and discharge FIM memory scores (see table 2). On neuropsychological tests, TBI was related to lower attention and working memory scores (see table 2). ost hoc analyses. ersons with tetraplegia and severe TBI had impaired functional problem solving and memory at admission compared with persons with no TBI. ersons with tetraplegia and severe TBI also had impaired functional memory at admission compared with persons with mild TBI. No differences were identified on discharge functional cognition or neuropsychological test scores. ersons with tetraplegia and moderate TBI had lower functional comprehension at admission and lower functional memory at admission and discharge compared with persons with no TBI. ersons with tetraplegia and moderate TBI had functional memory at admission compared with persons with mild TBI. ersons with tetraplegia and moderate TBI had a trend for Outcomes Table 3: Association Between TBI and Acute Rehabilitation Outcomes in araplegia Sample artial (n 42) (n 30) (n 12) Days to admit a,b ( ) 28.4 c,d ( ) 46.5 a,c ( ) 56.5 b,d ( ) Rehabilitation LOS a ( ) 34.2 b ( ) 39.7 ( ) 48.9 a,b ( ) FIM Motor Scale admission score a ( ) 37.9 b ( ) 37.1 ( ) 29.5 a,b ( ) FIM Motor Scale discharge score a ( ) 55.9 b ( ) 56.6 c ( ) 49.0 a,b,c ( ) FIM Motor Scale change score ( ) 18.0 ( ) 19.5 ( ) 19.5 ( ) FIM Motor Scale efficiency score ( ) 0.57 ( ) 0.55 ( ) 0.41 ( ) NOTES. ANCOVA used to determine main effects, with admission, lower extremity AMS used as a covariate for days to admit and rehabilitation LOS, and admission, lower extremity AMS, and sex used as covariates for Rasch-transformed FIM Motor Scale scores. Estimated marginal means (95% confidence intervals) are reported for the 4 TBI groups. a, b, c, d indicate that each TBI group with the same letter was different based on Bonferroni s test for multiple comparisons. Bold indicates significant findings.

5 1792 SINAL CORD INJURY AND CO-OCCURRING TBI, Macciocchi Motor Tasks Table 4: Association Between TBI and Motor Tasks Requiring Sequencing in araplegia Sample artial (n 42) (n 30) (n 12) Toileting a ( ) 4.6 b ( ) 5.1 c ( ) 2.5 a,b,c ( ) Dressing lower a ( ) 5.5 b ( ) 5.5 c ( ) 4.0 a,b,c ( ) Bathing a ( ) 5.0 ( ) 5.1 ( ) 4.0 a ( ) Toilet transfer a ( ) 4.4 b ( ) 4.3 ( ) 3.2 a,b ( ) Chair transfer a,b ( ) 4.5 a ( ) 4.6 ( ) 3.9 b ( ) Tub/shower transfer a ( ) 4.0 ( ) 3.8 ( ) 3.2 a ( ) NOTES. ANCOVA used to determine main effects, with admission, lower extremity AMS, and sex used as covariates for FIM Motor Scale items. Estimated marginal means (95% confidence intervals) are reported for the 4 TBI groups. a, b, c indicates that each TBI group with the same letter was different based on Bonferroni s test for multiple comparisons. Bold indicates significant findings. decreased attention (.062) and working memory (.063) compared with persons with no TBI. ersons with tetraplegia and mild TBI had equivalent functional comprehension, problem solving, and memory at admission and discharge compared with persons with tetraplegia and no TBI. ersons with tetraplegia and mild TBI had a trend for less attention (.053) and working memory (.056) compared with persons with no TBI. Co-Occurring TBI and araplegia Acute Rehabilitation Outcomes Covariates. Admission lower extremity AMS was entered as a covariate in ANCOVAs for days from injury to rehabilitation admission, acute rehabilitation LOS, and all FIM Motor Scale score analyses, and was significant for acute rehabilitation LOS, FIM Motor Scale total scores at admission and discharge, and toilet, chair, and tub/shower transfers items at discharge. Sex was entered as a covariate in ANCOVAs for acute rehabilitation LOS and all FIM Motor Scale score analyses. Sex was significant for acute rehabilitation LOS, FIM Motor Scale total scores at admission and discharge, FIM Motor Scale efficiency scores, and toilet, chair, and tub/shower transfers items at discharge. Lower extremity AMS and ASIA impairment scores did not differ between TBI classification groups. Main effects. For persons with paraplegia, TBI was related to time from injury to acute rehabilitation admission, acute rehabilitation LOS, and admission and discharge FIM Motor Scale scores (see table 3). TBI also was related to all 6 motor tasks requiring sequencing (see table 4). Motor differences had moderate to strong effect sizes and statistical power typically exceeded.90. For persons with paraplegia, TBI was related to admission and discharge FIM comprehension, problem solving, and memory items (see table 5). On neuropsychological tests, TBI was related to impaired performance on processing speed, auditory learning, and delayed auditory recall tests. ost hoc analyses. ersons with paraplegia and severe TBI had a greater number of days from injury to SCI rehabilitation admission, longer rehabilitation LOS, and lower admission and discharge FIM Motor Scale scores compared with persons with paraplegia and either no TBI or mild TBI (see table 3). ersons with paraplegia and severe TBI had worse outcomes on all 6 motor functional tasks requiring sequencing compared with persons with paraplegia and no TBI (see table 4). ersons with paraplegia and severe TBI had lower admission and discharge functional comprehension, problem solving, and memory compared with persons with no TBI, mild TBI, and moderate TBI on all but 1 comparison (see table 5). On neuropsychological tests in the paraplegia sample, persons with severe TBI had impaired processing speed compared with persons with no TBI, mild TBI, and moderate TBI. ersons with severe TBI had a trend for impaired delayed recall (.082) compared with persons with no TBI. ersons with paraplegia and moderate TBI were admitted to acute rehabilitation on average 22 days later than persons with no TBI and 18 days later than persons with mild TBI (see table Cognitive Measures Table 5: Effects of TBI on Cognition in araplegia Sample (n 93) artial (n 42) (n 30) (n 12) FIM items Comprehension (admission) a ( ) 6.8 b ( ) 6.8 c ( ) 4.6 a,b,c ( ) Comprehension (discharge) a ( ) 7.0 b ( ) 6.8 c ( ) 6.3 a,b,c ( ) roblem solving (admission) a ( ) 6.5 b ( ) 6.0 c ( ) 3.7 a,b,c ( ) roblem solving (discharge) a ( ) 6.7 b ( ) 6.3 ( ) 5.6 a,b ( ) Memory (admission) a ( ) 6.6 b ( ) 6.7 c ( ) 3.7 a,b,c ( ) Memory (discharge) a ( ) 6.8 b ( ) 6.7 c ( ) 5.9 a,b,c ( ) Neuropsychological Attention (WAIS-3 Digit Span) ( ) 9.2 ( ) 9.9 ( ) 7.4 ( ) Working memory (WAIS-3 L-N Seq.) ( ) 8.9 ( ) 9.0 ( ) 6.8 ( ) Auditory learning (HVLT Total Recall) ( ) 41.7 ( ) 51.2 a ( ) 32.6 a ( ) Auditory recall (HVLT Delayed Recall) ( ) 39.7 ( ) 47.9 a ( ) 30.3 a ( ) rocessing speed (SDMT-Oral) a ( ) 43.7 b ( ) 36.0 c ( ) 23.4 a,b,c ( ) NOTE. Analysis of variance used to determine main effects. Means (95% confidence intervals) are reported for the 4 TBI groups. a, b, c indicate that each TBI group with the same letter was different based on Bonferroni s test for multiple comparisons. Bold indicates significant findings. Abbreviations: HVLT, Hopkins Verbal Learning Test; L-N Seq., Letter-Number Sequencing; SDMT, Symbol Digit Modalities Test; WAIS-3, Wechsler Adult Intelligence Scale-3rd edition.

6 SINAL CORD INJURY AND CO-OCCURRING TBI, Macciocchi ). ersons with paraplegia and moderate TBI had equivalent acute rehabilitation LOS, functional motor outcomes, and performance on motor tasks requiring sequencing compared with persons with paraplegia and no TBI (see tables 3 and 4). ersons with paraplegia and moderate TBI had equivalent ratings on admission and discharge FIM comprehension, problem solving, and memory items and their neuropsychological test performance was equivalent to that of persons with paraplegia and no TBI (see table 5). ersons with paraplegia and mild TBI had equivalent days from injury to rehabilitation admission, admission and discharge FIM Motor Scale scores, and acute rehabilitation LOS (see table 2). ersons with paraplegia and mild TBI performed worse on chair transfers than did persons with paraplegia and no TBI (see table 4). ersons with paraplegia and mild TBI had equivalent cognitive performance on admission and discharge FIM cognitive items and all neuropsychological tests compared with persons with paraplegia and no TBI (see table 5). DISCUSSION Our prospective, single-center, cohort study provides evidence that co-occurring TBI negatively affects functional outcomes following paraplegia even when controlling for the severity of motor impairment and sex. ersons with paraplegia and severe TBI had lower discharge functioning in almost all FIM motor skills and longer rehabilitation LOS compared with persons with paraplegia and no TBI. ersons with paraplegia and severe TBI also had impaired FIM comprehension, problem solving, and memory scores as well as impaired processing speed and memory performance on neuropsychological tests. ersons with paraplegia and mild or moderate TBI had highly similar mean admission and discharge FIM Motor Scale total, change, and efficiency scores and scores on all 6 functional motor tasks requiring sequencing. Relatively fewer persons with paraplegia sustained moderate TBI in our sample, which reduced statistical power for identifying differences compared with the no-tbi group. ersons with paraplegia and moderate TBI were also admitted several weeks later than were persons with mild or no TBI, which may have resulted in the resolution of brain-injury related functional cognitive deficits prior to rehabilitation admission. In the tetraplegia sample, co-occurring TBI was not related to acute rehabilitation motor outcomes, but moderate and severe TBI negatively affected functional cognition and neuropsychological test performance. Admission upper extremity ASIA motor scores explained a substantial proportion of discharge FIM Motor variance in the tetraplegia sample, which most likely obscured the relationship between co-occurring TBI and motor skill acquisition. Sample size prevented analysis of the relationship between co-occurring TBI on completeness and level of cervical SCI, but future studies should examine this relationship as well as consider using motor-related compensatory tasks such as managing attendants and use of adaptive equipment. Using newer and potentially more sensitive SCI measures such as the Spinal Cord Independece Measure or the National Institute of Health Toolbox also merits consideration. On the basis of current data, persons with paraplegia and severe TBI should be provided modified therapeutic interventions that focus on improving processing speed, comprehension, learning, memory, and problem solving. Staff, patients, and families also should be educated regarding the unique health care needs of persons with paraplegia and co-occurring severe TBI. Moreover, health care payers should be prepared to extend the duration of hospitalization for persons with SCI and co-occurring severe TBI to achieve functional skill levels similar to those of peers without co-occurring TBI. Study Limitations A number of methodological limitations merit mention. Our sample was younger and less likely to sustain SCI as a result of falling compared with the general SCI population. ersons admitted many months from injury or with acute rehabilitation LOS greater than 100 days were not analyzed. Also, while brain injury diagnoses were based on a detailed review of acute medical-surgical records, a few participants may have sustained brief, undocumented LOC. Also, the effect of medical severity on acute rehabilitation LOS and the effect of rehabilitation service type and intensity on outcome were not examined. In addition, because our rehabilitation center is classified as a long-term care hospital, LOS and efficiency results may not generalize to independent rehabilitation facilities and effort tests were not utilized, and so the extent to which participant engagement played a role in below-expected test performance could not be determined. Most important, relatively few participants in our sample sustained moderate or severe TBI, and so a larger, multicenter study will likely be required to provide definitive evidence on how TBI affects functional skill acquisition. CONCLUSIO This study provides evidence that persons aged 16 to 59 years with paraplegia and co-occurring severe TBI had worse motor outcomes and longer acute rehabilitation LOS than did persons with paraplegia and no TBI. Impairments in processing speed, comprehension, learning, memory, and problem solving may explain suboptimal motor skill acquisition, and therapeutic interventions that target these impairments are recommended. Research with larger SCI samples is required to determine whether mild and moderate TBI is related to acute rehabilitation motor outcomes. References 1. Davidoff G, Morris J, Roth E, Bleiberg J. Cognitive dysfunction and mild closed head injury in traumatic spinal cord injury. Arch hys Med Rehabil 1985;66: Davidoff G, Thomas, Johnson M, Berent S, Dijkers M, Doljanac R. Closed head injury in acute traumatic spinal cord injury: incidence and risk factors. Arch hys Med Rehabil 1988;69: Roth E, Davidoff GN, Thomas, et al. A controlled study of neuropsychological deficits in acute spinal cord injury patients. araplegia 1988;27: Richards JS, Brown L, Hagglund K, Gua G, Reeder K. Spinal cord injury and concomitant traumatic brain injury: results of a longitudinal investigation. Am J hys Med Rehabil 1988;67: Richards JS, Osuna FJ, Jaworski O, et al. The effectiveness of different methods of defining traumatic brain injury in predicting post-discharge adjustment in a spinal cord injury population. Arch hys Med Rehabil 1991;72: Macciocchi S, Seel R, Thompson N, Byams R, Bowman B. Spinal cord injury and co-occurring traumatic brain injury: assessment and incidence. Arch hys Med Rehabil 2008;89: Davidoff GN, Roth EJ, Haughton JS, Ardner MS. Cognitive dysfunction in spinal cord injury patients: sensitivity of the Functional Independence Measure. Arch hys Med Rehabil 1990;71: Stutts M, Kreutzer JS, Barth JT, et al. Cognitive impairment in persons with recent spinal cord injury: findings and implications for clinical practice. NeuroRehabil 1991;3:79-85.

7 1794 SINAL CORD INJURY AND CO-OCCURRING TBI, Macciocchi 9. Hess DW, Marwitz JH, Kreutzer JS. Neuropsychological impairments after spinal cord injury: a comparative study with mild traumatic brain injury. Rehabil sychol 2003;48: Bradbury CL, Wodchis W, Mikulis DJ, et al. Traumatic brain injury in patients with spinal cord injury: clinical and economic consequences. Arch hys Med Rehabil 2008;89:S Macciocchi SN, Bowman B, Coker J, Apple D, Leslie D. Effect of co-morbid traumatic brain injury on functional outcome of persons with spinal cord injuries. Am J hys Med Rehabil 2004;83: Williams DH, Levin HS, Eisenberg HM. Mild head injury classification. Neurosurgery 1990;27: Kashluba S, Hanks RA, Casey JE, Millis SR. Neuropsychologic and functional outcome after complicated mild traumatic brain injury. Arch hys Med Rehabil 2008;89: Marino RJ, Barros T, Biering-Sorensen F, et al; ASIA Neurological Standards Committee International standards for neurological classification of spinal cord injury. J Spinal Cord Med 2003;26:S Graves DE, Frankiewicz RG, Donovan WH. Construct validity and dimensional structure of the ASIA motor scale. J Spinal Cord Med 2006;29: Heinemann AW, Linacre JM, Wright BD, Hamilton BB, Granger C. Relationships between impairment and physical disability as measured by the Functional Independence Measure. Arch hys Med Rehabil 1993;74: Linacre JM, Heinemann AW, Wright BD, Granger CV, Hamilton BB. The structure and stability of the Functional Independence Measure. Arch hys Med Rehabil 1994;75: Ottenbacher KJ, Hsu Y, Granger CV, Fiedler RC. The reliability of the Functional Independence Measure: a quantitative review. Arch hys Med Rehabil 1996;77: Heinemann AW, Linacre JM, Wright BD, Hamilton BB, Granger CV. Measurement characteristics of the Functional Independence Measure. Top Stroke Rehabil 1994;1: Trahan DE, Larrabee GJ. Continuous Visual Memory Test professional manual. Odessa: sychological Assessment Resources; Wechsler D. Administration and scoring manual for the Wechsler Adult Intelligence Scale 3rd edition. San Antonio: The sychological Corporation; Brandt J, Benedict RH. Hopkins Verbal Learning Test-Revised manual. Lutz: sychological Assessment Resources; Smith A. Symbol Digit Modalities Test. Los Angeles: Western sychological Services; Wetzel L, Boll TJ. Short Category Test booklet. Los Angeles: Western sychological Services; Novack TA, Dowler RN, Bus BA, Glen ET, Schneider JJ. Validity of the Orientation Log relative to the Galveston Orientation and Amnesia Test. J Head Trauma Rehabil 2000;15: Suppliers a. SSS Inc, 233 S Wacker Dr, 11th Fl, Chicago, IL