Advertisement
Archives of Physical Medicine and Rehabilitation
Advanced SearchSave search
Original article| Volume 91, ISSUE 12, P1904-1913, December 01, 2010

Cognitive Impairment 3 Months After Moderate and Severe Traumatic Brain Injury: A Prospective Follow-Up Study

DOI:https://doi.org/10.1016/j.apmr.2010.08.021
Cognitive Impairment 3 Months After Moderate and Severe Traumatic Brain Injury: A Prospective Follow-Up Study
Previous ArticleDynamic Versus Fixed Equinus Deformity in Children With Cerebral Palsy: How Does the Triceps Surae Muscle Work?
Next ArticleBalance, Balance Confidence, and Health-Related Quality of Life in Persons With Chronic Stroke After Body Weight–Supported Treadmill Training

      Abstract

      Skandsen T, Finnanger TG, Andersson S, Lydersen S, Brunner JF, Vik A. Cognitive impairment 3 months after moderate and severe traumatic brain injury: a prospective follow-up study.

      Objective

      To explore the magnitude and frequency of cognitive impairment 3 months after moderate to severe traumatic brain injury (TBI), and to evaluate its relationship to disability at 1-year follow-up.

      Design

      Prospective follow-up study.

      Setting

      Regional level I trauma center.

      Participants

      Patients aged 15 to 65 years with definite TBI, defined as Glasgow Coma Scale score of 3 to 13 and injury documented by magnetic resonance imaging (n=59) or computed tomography (n=2); healthy volunteers (n=47) served as controls.

      Interventions

      Not applicable.

      Main Outcome Measures

      Neuropsychological assessment 3 months postinjury and Glasgow Outcome Scale Extended (GOSE) at 3 and 12 months postinjury.

      Results

      Patients with TBI performed worse than controls, most consistently in terms of information processing speed and verbal memory. However, a maximum of only 43% of patients with TBI had impaired test scores (defined as <1.5 SD below mean of normative data) on any one measure. Based on a selection of 9 tests, a 0 or 1 impaired score was seen in 46 (98%) of 47 controls, in 20 (57%) of 35 patients with moderate TBI, and in 9 (35%) of 26 patients with severe TBI. At 1 year postinjury, disability (defined as GOSE score ≤6) was present in 57% of those with 2 or more impaired test scores and in 21% of those with 0 or 1 impaired score (P=.005).

      Conclusions

      In this sample of patients with recent, definite TBI and healthy volunteers, we found that TBI affected cognition in moderate as well as severe cases. The presence of cognitive impairment was associated with future disability. However, half of the patients with moderate TBI and even one third of those with severe TBI had a normal cognitive assessment 3 months postinjury.

      Key Words

      List of Abbreviations:

      CCPT-II (Conners' Continuous Performance Test II), CT (computed tomography), CVLT-II (California Verbal Learning Test II), CVMT (Continuous Visual Memory Test), CWIT (Color-Word Interference Test), DAI (diffuse axonal injury), D-KEFS (Delis Kaplan Executive Function System), GCS (Glasgow Coma Scale), GOSE (Glasgow Outcome Scale Extended), IQ (intelligence quotient), MRI (magnetic resonance imaging), PTA (posttraumatic amnesia), SDMT (Symbol Digit Modalities Test), TBI (traumatic brain injury), TCF (Taylor Complex Figure), TMT (Trail-Making Test), WAIS-III (Wechsler Adult Intelligence Scale–Third Edition)
      COGNITIVE IMPAIRMENT is a common sequela of moderate and severe TBI,
      • Dikmen S.S.
      • Corrigan J.D.
      • Levin H.S.
      • Machamer J.
      • Stiers W.
      • Weisskopf M.G.
      Cognitive outcome following traumatic brain injury.
      J Head Trauma Rehabil. 2009; 24: 430-438
      with effects particularly prominent in terms of information processing speed and attention,
      • Mathias J.L.
      • Wheaton P.
      Changes in attention and information-processing speed following severe traumatic brain injury: a meta-analytic review.
      Neuropsychology. 2007; 21: 212-223
      , 
      • Willmott C.
      • Ponsford J.
      • Hocking C.
      • Schonberger M.
      Factors contributing to attentional impairments after traumatic brain injury.
      Neuropsychology. 2009; 23: 424-432
      memory,
      • Vakil E.
      The effect of moderate to severe traumatic brain injury (TBI) on different aspects of memory: a selective review.
      J Clin Exp Neuropsychol. 2005; 27: 977-1021
      and executive functioning.
      • Dikmen S.S.
      • Machamer J.E.
      • Winn H.R.
      • Temkin N.R.
      Neuropsychological outcome at 1-year post head-injury.
      Neuropsychology. 1995; 9: 80-90
      Several studies demonstrated, at various time intervals postinjury, that cognitive dysfunction mediates functional problems.
      • Novack T.A.
      • Bush B.A.
      • Meythaler J.M.
      • Canupp K.
      Outcome after traumatic brain injury: pathway analysis of contributions from premorbid, injury severity, and recovery variables.
      Arch Phys Med Rehabil. 2001; 82: 300-305
      , 
      • Rassovsky Y.
      • Satz P.
      • Alfano M.S.
      • et al.
      Functional outcome in TBI I: neuropsychological, emotional, and behavioral mediators.
      J Clin Exp Neuropsychol. 2006; 28: 567-580
      , 
      • Ponsford J.
      • Draper K.
      • Schonberger M.
      Functional outcome 10 years after traumatic brain injury: its relationship with demographic, injury severity, and cognitive and emotional status.
      J Int Neuropsychol Soc. 2008; 14: 233-242
      Cognition is markedly impaired around 1 month postinjury
      • Dikmen S.
      • McLean Jr, A.
      • Temkin N.R.
      • Wyler A.R.
      Neuropsychologic outcome at one-month postinjury.
      Arch Phys Med Rehabil. 1986; 67: 507-513
      or by resolution of PTA,
      • Boake C.
      • Millis S.R.
      • High Jr, W.M.
      • et al.
      Using early neuropsychologic testing to predict long-term productivity outcome from traumatic brain injury.
      Arch Phys Med Rehabil. 2001; 82: 761-768
      , 
      • Kreutzer J.S.
      • Gordon W.A.
      • Rosenthal M.
      • Marwitz J.M.
      Neuropsychological characteristics of patients with brain injury: preliminary findings from a multicenter investigation.
      J Head Trauma Rehabil. 1993; 8: 47-59
      but studies are difficult to compare because of differences regarding duration of follow-up, proportion of severe cases, and whether patients are selected from the acute care
      • Dikmen S.S.
      • Machamer J.E.
      • Winn H.R.
      • Temkin N.R.
      Neuropsychological outcome at 1-year post head-injury.
      Neuropsychology. 1995; 9: 80-90
      or rehabilitation setting.
      • Boake C.
      • Millis S.R.
      • High Jr, W.M.
      • et al.
      Using early neuropsychologic testing to predict long-term productivity outcome from traumatic brain injury.
      Arch Phys Med Rehabil. 2001; 82: 761-768
      , 
      • Kreutzer J.S.
      • Gordon W.A.
      • Rosenthal M.
      • Marwitz J.M.
      Neuropsychological characteristics of patients with brain injury: preliminary findings from a multicenter investigation.
      J Head Trauma Rehabil. 1993; 8: 47-59
      A common research design has been the comparison of mean test performances between patients with TBI and a control group. The magnitude of the effect of TBI on cognition (ie, effect size) may be expressed as Cohen's d, the standardized difference of means in 2 groups. In a meta-analysis, a large mean effect size (dpooled=.97) of TBI across studies was found for moderate to severe TBI within the first 6 months.
      • Schretlen D.J.
      • Shapiro A.M.
      A quantitative review of the effects of traumatic brain injury on cognitive functioning.
      Int Rev Psychiatry. 2003; 15: 341-349
      The authors pointed out that while cases of moderate and severe TBI typically are analyzed together in neuropsychological research, they constitute a heterogeneous group. Thus, the reported effects might be overestimated for moderate TBI or underestimated for severe TBI, or both.
      • Schretlen D.J.
      • Shapiro A.M.
      A quantitative review of the effects of traumatic brain injury on cognitive functioning.
      Int Rev Psychiatry. 2003; 15: 341-349
      In other studies of early cognitive outcome in TBI patient groups,
      • Boake C.
      • Millis S.R.
      • High Jr, W.M.
      • et al.
      Using early neuropsychologic testing to predict long-term productivity outcome from traumatic brain injury.
      Arch Phys Med Rehabil. 2001; 82: 761-768
      , 
      • Kreutzer J.S.
      • Gordon W.A.
      • Rosenthal M.
      • Marwitz J.M.
      Neuropsychological characteristics of patients with brain injury: preliminary findings from a multicenter investigation.
      J Head Trauma Rehabil. 1993; 8: 47-59
      as well as in individual clinical assessments,
      • Strauss E.
      • Sherman E.M.S.
      • Spreen O.
      A compendium of neuropsychological tests administration, norms, and commentary.
      3rd ed. Oxford University Pr, New York2006
      interpretation of neuropsychological test scores is based on comparison with normative data. In some settings, such as legal expert examinations and research, criteria defining impairment is required, but there is no such general definition.
      • Iverson G.L.
      • Brooks B.L.
      • Holdnack J.A.
      Misdiagnosis of cognitive impairment in forensic neuropsychology.
      in: Heilbronner R.L. Neuropsychology in the courtroom: expert analysis of reports and testimony. 1st ed. The Guilford Pr, New York2008: 243-266
      The issue is further complicated by the fact that healthy people perform below the normal range in some percentage of administered tests,
      • Iverson G.L.
      • Brooks B.L.
      • Holdnack J.A.
      Misdiagnosis of cognitive impairment in forensic neuropsychology.
      in: Heilbronner R.L. Neuropsychology in the courtroom: expert analysis of reports and testimony. 1st ed. The Guilford Pr, New York2008: 243-266
      , 
      • Schretlen D.J.
      • Testa S.M.
      • Winicki J.M.
      • Pearlson G.D.
      • Gordon B.
      Frequency and bases of abnormal performance by healthy adults on neuropsychological testing.
      J Int Neuropsychol Soc. 2008; 14: 436-445
      and defining “impairment” remains a matter of discussion.
      Taken together, there is still a need for studies exploring the extent of cognitive deficits in patients with TBI, because previous studies are heterogeneous regarding the time postinjury, the selection of patients, and the research design.
      For the present study we performed neuropsychological testing 3 months postinjury as part of a large follow-up study of patients admitted with moderate or severe head injury to a regional level I trauma center. In a larger subgroup of the main cohort, we previously demonstrated that virtually all patients had parenchymal lesions detected with early MRI. DAI and contusions were frequently found, often in concert.
      • Skandsen T.
      • Kvistad K.A.
      • Solheim O.
      • Strand I.H.
      • Folvik M.
      • Vik A.
      Prevalence and impact of diffuse axonal injury in patients with moderate and severe head injury: a cohort study of early magnetic resonance imaging findings and 1-year outcome.
      J Neurosurg. 2010; 113: 556-563
      The aim of the present study was to explore the magnitude and frequency of cognitive impairments 3 months after moderate or severe TBI in comparison with healthy controls and with normative data. Furthermore, we sought to relate the level of cognitive functioning at 3 months to measures of global functioning at 3 months and 1-year follow-up.

      Methods

      The Regional Committee for Medical Research Ethics and the Norwegian Social Science Data Services approved the study. Written consent was obtained from patients and from parents of patients younger than 16 years.

       Participants

      Sixty-one patients (age range, 15–65y) admitted to the Neurosurgical Department, St Olavs Hospital, Trondheim University Hospital, Norway participated in the study. The hospital, a level I trauma center, has an ongoing database that includes all patients admitted with moderate to severe head injuries as defined by the Head Injury Severity Scale criteria.
      • Stein S.C.
      • Spettell C.
      The Head Injury Severity Scale (HISS): a practical classification of closed-head injury.
      Brain Inj. 1995; 9: 437-444
      In this main database, 97% of admitted patients have consented to registration, and less than 2% have been lost to follow-up (ie, missing GOSE score).
      For the present study, patients in the main database were invited to participate in neuropsychological testing at 3 months postinjury. Inclusion criteria were (1) the ability to cooperate during testing; (2) no ongoing substance abuse, diagnosed neurologic or psychiatric condition, or previous moderate to severe head injury according to the same criteria; and (3) fluency in the Norwegian language. The main inclusion period was from October 2004 to October 2007. During this period, 52 patients were included (appendix 1); these constitute 85% of the sample in this study. For the purpose of increasing the sample, we included 9 patients who had been injured and registered in the main database after the first inclusion period and who were evaluated at 3 months follow-up. Table 1 reports patient demographic data. The control group consisted of 47 healthy persons, matched to the total sample of patients for age, sex, and education. They were recruited via advertisements, among family and friends of patients with head injury, and among acquaintances of researchers and staff.
      Table 1Demographic, Injury-Related, and Outcome Variables in Individuals With Moderate and Severe TBI
      VariableNo. of Cases (%)
      Main Cohort (n=52)Patients Enrolled Later (n=9)
      Age at testing; mean ± SD (y)30.6±14.626.4±12.8
      Male/female40 (77)/12 (23)5 (56)/4 (44)
      Days postinjury; mean ± SD (range)97.3±10.8 (75–133)103.4±9.4 (90–122)
      GCS score; median (IQR)
      Exact GCS score available in 59 cases.
      		9 (7–13)7.5 (6.25–12.75)
      Moderate/severe injury32 (62)/20 (38)3 (33)/6 (67)
      Mechanism of injury
       Fall20 (39)1 (11)
       Traffic accident28 (54)8 (89)
       Other4 (7)0
      MRI examination
      One patient not examined with MRI.
       No findings1 (2)0
       Pure DAI10 (19)3 (38)
       Cortical contusions17 (33)0
       DAI and contusions24 (46)5 (56)
      GOSE score at testing
      One patient could not be reliably assessed because of severe orthopedic injuries.
       812 (23)0
       73 (6)0
       613 (25)2 (22)
       521 (40)7 (78)
       41 (2)0
       31 (2)0
      GOSE score at follow-up
      Data not available in 2 cases.
       822 (42)2 (29)
       711 (21)1 (14)
       69 (17)3 (43)
       510 (19)1 (14)
      NOTE. Values are number of cases (%) unless otherwise stated.
      Abbreviation: IQR, interquartile range.
      low asterisk Exact GCS score available in 59 cases.
      † One patient not examined with MRI.
      ‡ One patient could not be reliably assessed because of severe orthopedic injuries.
      § Data not available in 2 cases.

       Injury-Related Variables

      Evaluated variables included mechanism of injury, MRI findings, and GCS score (scoring procedures have been described in an earlier publication
      • Skandsen T.
      • Kvistad K.A.
      • Solheim O.
      • Strand I.H.
      • Folvik M.
      • Vik A.
      Prevalence and impact of diffuse axonal injury in patients with moderate and severe head injury: a cohort study of early magnetic resonance imaging findings and 1-year outcome.
      J Neurosurg. 2010; 113: 556-563
      ), with a GCS score of 9 to 13 indicating moderate TBI and a GCS score of 8 or less reflecting severe TBI.

       Magnetic Resonance Imaging

      MRI (1.5 Tesla) was performed at a median of 10 days postinjury (range, 1–120). The scan protocol included T1- and T2-weighted sequences, a T2*-weighted gradient echo sequence, fluid-attenuated inversion recovery sequences, and diffusion-weighted imaging. MRI parameters and procedure of evaluation have been reported previously.
      • Skandsen T.
      • Kvistad K.A.
      • Solheim O.
      • Strand I.H.
      • Folvik M.
      • Vik A.
      Prevalence and impact of diffuse axonal injury in patients with moderate and severe head injury: a cohort study of early magnetic resonance imaging findings and 1-year outcome.
      J Neurosurg. 2010; 113: 556-563

       Procedures for Neuropsychological Testing and Scoring

      Neuropsychological assessment was performed at a mean of 98±10 (SD) days postinjury. Psychologists, 2 trained masters-level students, and 1 test technician at St Olav University Hospital performed all testing. To compensate for errors associated with several examiners, all were supplied with oral and written instructions regarding the protocol and the procedures. The students received training and could discuss issues with the psychologists.
      Raw scores were converted to standard scores by use of normative data provided by the manufacturers of the tests, except for the Symbol Digit Modality Test, where a normative sample quoted by Lezak et al
      • Lezak M.D.
      • Howieson D.B.
      • Loring D.W.
      Neuropsychological assessment.
      4th ed. Oxford University Pr, New York2004
      was used. For participants aged 15 years, the norms for those aged 16 years were used. Standard scores were given as T scores, scaled scores (S scores), Z scores, or percentiles. An individual's standardized test score was classified as impaired if below 1.5 SD according to the reference norms for the test (T score ≤34, S score ≤5, Z score ≤−1.5, or percentile ≤5). Data were also analyzed applying a cutoff criterion at 1 SD. In some cases 1 or more tests were not administered for various reasons; thus the number of patients evaluated with each test deviates from the total sample size.

       Neuropsychological Measures

      The 4 subtests of the Wechsler Abbreviated Scale of Intelligence
      • Wechsler D.
      Wechsler Abbreviated Scale of Intelligence Scale.
      The Psychological Corporation, San Antonio1999
      were administered to estimate general intellectual capacity. To avoid future retest effects in a planned reassessment, we used a split-half procedure, and a raw score was estimated. The control participants were tested with all items, but their IQ scores in this study were calculated as for the patients, by use of every second item procedure. The following neuropsychological methods were used to assess different domains of cognitive function:
      • 1
        Motor function: Grooved Pegboard, dominant hand,
        • Matthews C.G.
        • Klove K.
        Instruction manual for the Adult Neuropsychology Test Battery.
        University of Wisconsin Medical School, Madison1964
        TMT, condition 5 (motor speed) from the D-KEFS.
        • Delis D.C.
        • Kaplan E.
        • Kramer J.
        Delis Kaplan Executive Function System.
        The Psychological Corporation, San Antonio2001
      • 2
        Information processing: TMT, condition 1 (visual scanning), 2 (number sequencing), and 3 (letter sequencing); CWIT, condition 1 (color naming) and 2 (word reading) from D-KEFS and SDMT, oral and written versions.
        • Smith A.
        Symbol Digit Modality Test (SDMT) manual.
        Western Psyhological Services, Los Angeles2002
      • 3
        Attention and vigilance: CCPT-II.
        • Conners C.K.
        • Staff M.
        Conners' Continuous Performance Test II (CPT II) computer program for Windows, technical guide and software manual.
        Multi-Health Systems, Toronto2000
      • 4
        Visual learning and memory: CVMT,
        • Trahan D.E.
        • Larrabee G.J.
        Continous Visual Memory Test.
        Psychological Assessment Resources, Odessa1988
        TCF.
        • Taylor L.B.
        Localisation of cerebral lesions by psychological testing.
        Clin Neurosurg. 1969; 16: 269-287
        The TCF was not administered to controls. Raw scores for the TCF were converted to standard scores based on normative data for the Rey-Osterrieth Complex Figure.
        • Meyers J.E.
        • Meyers K.R.
        Rey Complex Figure Test and Recognition Trial: professional manual.
        Psychological Assessment Resources, Odessa1995
      • 5
        Verbal learning and memory: CVLT-II.
        • Delis D.C.
        • Kramer J.H.
        • Kaplan E.
        • Ober B.A.
        California Verbal Learning Test second edition (CVLT-II).
        The Psychological Corporation, San Antonio2000
      • 6
        Working memory: Digit Span Backwards from the WAIS-III,
        • Wechsler D.
        Wechsler Adult Intelligence Scale.
        3rd ed. The Psychological Corporation, San Antonio1997
        Letter-Number Sequencing from WAIS-III.
      • 7
        Executive functions: Wisconsin Card Sorting Test computer version
        • Heaton R.K.
        Wisconsin Card Sorting Test, computer version-2 research edition.
        Psychological Assessment Resources, Odessa1993
        ; Verbal Fluency Test from D-KEFS; TMT, condition 4 (letter-number switching) from D-KEFS; CWIT, condition 3 and 4 (inhibition and inhibition/switching) from D-KEFS; Tower test from D-KEFS.
      The different subscores obtained from each test are shown in Table 2, Table 3.
      Table 2Neuropsychological Test Performance in Patients and Controls
      TestnPatientsnControlsPEScontrol
      Cohen's d; mean difference divided by the pooled SD.
      		dpooled
      For tests yielding normally distributed data; Glass' d; ES is the standardized mean, using the SD of the control group. For the nonnormally distributed data, EScontrol was calculated as difference between medians divided by .75 × interquartile rangecontrol.
      Motor function
       TMT; motor
      Data not normally distributed.
      		6125 (18.5–34)4722 (18–27.5).06.42
       Grooved Pegboard; dominant hand (s)6172.8±18.34764.0±7.3.0011.20.61
       Grooved Pegboard; nondominant hand (s)6080.5±17.84770.2±8.4<.0011.22.71
      Information processing speed
       TMT; visual scanning
      Data not normally distributed.
      (s)      		6126 (23–33)4720 (17–23)<.0011.33
       TMT; numbers
      Data not normally distributed.
      (s)      		6134 (29–47)4724 (21–29)<.0011.67
       TMT; letters
      Data not normally distributed.
      (s)      		6134 (28.5–49.5)4724 (20–30)<.0011.33
       CWIT; color naming (s)5934.8±9.34729.0±4.7<.0011.23.76
       CWIT; word reading (s)5925.4±5.84722.4±3.6.002.82.60
       SDMT; oral version5956.0±17.54766.1±11.4.001.88.67
       SDMT; written version6145.2±13.34753.5±8.0<.0011.04.73
      Sustained attention
       CCPT; omissions
      Data not normally distributed.
      		573 (1–6)471 (0–2).0141.33
       CCPT; comissions
      Data not normally distributed.
      		5712 (9–21.5)479 (5–15).003.40
       CCPT; hit RT57378.7±82.247391.7±63.8.37
       CCPT; hit RT SE
      Data not normally distributed.
      		575.8 (4.3–6.0)475.0 (3.9–5.9).047.56
       CCPT; detectability57.68±.3947.92±.43.004.55.58
       CCPT; hit RT by block
      Data not normally distributed.
      		570.0 (−.01 to .01)470.0 (−.01 to .02).60
       CCPT; hit RT by block SE
      Data not normally distributed.
      		570.000 (−.033 to .033)470.005 (−.030 to .050).57
       CCPT; hit RT by ISI
      Data not normally distributed.
      		57.040 (−.000 to .070)47.040 (−.030 to .060).17
       CCPT; hit RT by ISI SE57−.014±.0947.034±.12.021.41.46
      Visual memory
       CVMT; hits
      Data not normally distributed.
      		5938 (34–39)4739 (37–41).012.33
       CVMT; total correct5973.2±9.54777.0±6.3.019.62.47
       CVMT; false
      Data not normally distributed.
      		5916 (10–20)4715 (11-19.75).85
       CVMT; delayed594.3±1.7474.7±1.4.10.34.32
      Verbal memory
       CVLT; total recall trial 1–56147.5±12.94753.4±8.9.006.67.52
       CVLT, short-delay free recall
      Data not normally distributed.
      		6110 (8–12.5)4712 (11–14).013.67
       CVLT, long-delay free recall
      Data not normally distributed.
      		6111 (7.5–14)4712 (11–14).007.44
      Working memory
       Digitspan backwards606.7±2.2476.9±2.4.68
       Letter-Number Sequencing5910.3±3.14711.4±2.8.051.41.39
      Executive function
       WCST; total errors
      Data not normally distributed.
      		5922 (14–38)4714 (10–20).0021.07
       WCST; perseverative responses
      Data not normally distributed.
      		5911 (7–19)478 (6–11).010.80
       WCST; categories achieved
      Data not normally distributed.
      		596 (6–6)476 (6–6).57
       Verbal Fluency; letter6032.4±10.84739.5±11.3.001.63.65
       Verbal Fluency; category6039.5±10.14752.6±10.9<.0011.211.26
       Verbal Fluency; category switching; tot corr6012.7±2.94713.8±2.8.036.42.41
       Verbal Fluency; category switching; tot sw6011.3±3.34711.9±3.1.29
       Tower; total achievement6117.5±3.64718.0±3.6.41
       TMT; letter-number switching
      Data not normally distributed.
      (s)      		6181 (64.5–108.5)4761 (50–76)<.0011.03
       CWIT; inhibition (s)5958.9±18.74749.8±8.0.0011.13.61
       CWIT; inhibition/switching
      Data not normally distributed.
      (s)      		5968 (56–83)4754 (51–64)<.0011.44
      NOTE. Values are mean ± SD, median (25%–75%), or as otherwise indicated.
      Abbreviations: corr, correct; ES, effect size; ISI, interstimulus interval; RT, reaction time; sw, switching; WCST, Wisconsin Card Sorting Test.
      low asterisk Cohen's d; mean difference divided by the pooled SD.
      † For tests yielding normally distributed data; Glass' d; ES is the standardized mean, using the SD of the control group. For the nonnormally distributed data, EScontrol was calculated as difference between medians divided by .75 × interquartile rangecontrol.
      ‡ Data not normally distributed.
      Table 3Number and Proportion of Scores Below 1.5 SD, All Patients and Controls
      TestTBI <1.5 SDControls <1.5 SDP
      Totaln (%)Totaln (%)
      Motor function
       PEG dominant side6112 (20)471 (2).007
       PEG nondominant side6016 (27)472 (4).52
       TMT 5615 (8)471 (2).003
      Information processing speed
       TMT; visual scanning6116 (26)471 (2).003
       TMT; numbers6113 (21)470.002
       TMT; letters6114 (23)472 (4).008
       CWIT; color naming5917 (28)472 (4).002
       CWIT; word reading5912 (20)472 (4).017
       SDMT; oral version5916 (27)470.001
       SDMT; written version6126 (43)471 (2).001
      Sustained attention/vigilance
       CCPT; omissions588 (14)473 (6).26
       CCPT; comissions585 (9)471 (2).18
       CCPT; hit RT586 (10)474 (9).78
       CCPT; hit SE586 (10)475 (11).99
       CCPT detect580 (0)471 (2).36
       CCPT; hit RT by block584 (7)472 (4).63
       CCPT; hit SE by block586 (10)475 (11).99
       CCPT; hit RT SE by ISI581 (2)473 (6).26
       CCPT; hit RT SE byISI583 (5)475 (11).36
      Visual memory
       CVMT; hits5914 (24)474 (9).040
       CVMT; false5920 (34)4714 (30).67
       CVMT; total5921 (36)4713 (28).54
       CVMT; delayed599 (15)472 (4).075
       TCF; delayed598 (14)0NAND
      Verbal memory
       CVLT; total recall trial 1–5599 (15)471 (2).022
       CVLT; short-delay free recall5915 (25)471 (2).002
       CVLT; long-delay free recall5917 (29)473 (6).004
      Working memory
       Letter-Number Sequencing604 (7)471 (2).36
      Executive function
       WCST; total errors595 (8)471 (2).19
       WCST; perseverative responses595 (8)472 (4).52
       WCST; categories achieved591 (2)471 (2).96
       Letter Fluency6010 (17)471 (2).015
       Category Fluency6010 (16)470.004
       Category switching; total correct607 (12)472 (4).20
       Category switching; total switch608 (13)472 (4).13
       CWIT inhibition5911 (19)471 (2).009
       CWIT inhibition/switching5914 (23)472 (4).006
       TMT; letter-number switching6111 (19)470.004
       Tower; total achievement591 (2)471 (2).96
      Abbreviations: ISI, interstimulus interval; PEG, Grooved Pegboard; RT, reaction time; WCST, Wisconsin Card Sorting Test; NA, not applicable; ND, no data.

       Global Outcome

      Global functioning was assessed by the first author according to the GOSE at 3 and 12 months follow-up using a structured interview.
      • Wilson J.T.
      • Pettigrew L.E.
      • Teasdale G.M.
      Structured interviews for the Glasgow Outcome Scale and the extended Glasgow Outcome Scale: guidelines for their use.
      J Neurotrauma. 1998; 15: 573-585
      The scale assesses disability, participation, and symptoms after head injury. A cutoff GOSE score of 6 was chosen, indicating the presence of severe (GOSE score, 3–4) or moderate (GOSE score, 5–6) disability. However, because head injury–related complaints may be present in patients without this level of disability, to a degree that affect daily life (GOSE score, 7), we also performed some analyses using a cutoff score of 7.

       Statistical Analysis

      Dependent variables were checked for normality in patients with TBI and control participants by inspection of Q-Q plots. Raw scores for each test are presented as mean and SD for normally distributed data, and otherwise as median with interquartile range (25th and 75th percentile). Comparisons between patients with TBI and control participants were performed using the Student t test and Mann-Whitney U test, respectively. Effect sizes were calculated both as Cohen's d based on pooled variance (dpooled) and as Glass' d, where the denominator is the SD of the control group (EScontrol).
      • Grissom R.J.
      • Kim J.J.
      Effect sizes for research A broad practical approach.
      Lawrence Erlbaum, Mahwah2005
      For the tests where data were nonnormally distributed, standardized effect size was estimated by dividing the difference between the median scores by the interquartile range of controls × .75 (EScontrol).
      • Grissom R.J.
      • Kim J.J.
      Effect sizes for research A broad practical approach.
      Lawrence Erlbaum, Mahwah2005
      Cohen defined a d of 0.8, 0.5, and 0.2 as reflecting large, medium, and small effect sizes, respectively.
      • Cohen J.
      Statistical power analysis for the behavioral sciences.
      2nd ed. Erlbaum, Hillsdale1988
      Proportions were compared using the exact z pooled test as recommended for small counts when the expected value was less than 5 for any cell.
      • Lydersen S.
      • Fagerland M.W.
      • Laake P.
      Recommended tests for association in 2 × 2 tables.
      Stat Med. 2009; 28: 1159-1175
      In the presence of missing data, we used available case analysis, using all cases where a variable is present.
      • Schafer J.L.
      • Graham J.W.
      Missing data: our view of the state of the art.
      Psychol Methods. 2002; 7: 147-177
      Thus the number of cases is different for each variable. Reported P values are 2-sided. To adjust for multiple tests, a P value of .01 or less was regarded as significant when analyzing the total battery of the neuropsychological tests, and otherwise an α level of .05 was applied. Statistical analyses were performed using the statistical software SPSS for Windows, version 16.0
      SPSS, Inc, 233 S Wacker Dr, 11th Fl, Chicago, IL 60606.
      with the exception of exact unconditional tests, which were performed using http://www.stat.ncsu.edu/exact/.

      Results

      Demographic and injury-related characteristics of the patients and control participants are presented in Table 1, Table 4.
      Table 4Intellectual Ability in Relation to Sex and Injury Severity
      VariableControls

      (n=47)      		TBI, All Patients

      (n=61)      		Moderate TBI

      (n=35)      		Severe TBI

      (n=26)      		Male, TBI

      (n=45)      		Female, TBI

      (n=16)
      VIQ117±13106±17108±12103±21102±16117±14
      PIQ115±13106±13107±13104±13104±13111±13
      Total IQ
      Three patients had no IQ scores from Wechsler Abbreviated Scale of Intelligence.
      		118±12106±16
      Significantly lower than in controls.
      		109±12103±20102±15
      Significantly lower than in women with TBI.
      		116±14
      Age (y)30.2±13.330.0±14.333.1±15.225.9±12.029.0±13.333.0±16.9
      Education (y)12.0±1.811.8±2.112.0±2.011.7±2.311.3±1.613.3±2.8
      NOTE. Values are mean ± SD.
      Abbreviations: PIQ, performance intelligence quotient; VIQ, verbal intelligence quotient.
      low asterisk Three patients had no IQ scores from Wechsler Abbreviated Scale of Intelligence.
      † Significantly lower than in controls.
      ‡ Significantly lower than in women with TBI.

       Neuroimaging

      All patients were examined by MRI except one, who had cortical contusions depicted with CT. One of the 60 patients had no lesions in the brain parenchyma when examined with MRI 21 days postinjury. This patient had a moderate injury with 4 days of PTA and subarachnoid hemorrhage on the initial CT scan. Thus, all patients had definite TBI, and 69% had DAI detected with MRI.

       Intellectual Abilities

      When tested via t test, patients with TBI had significantly lower verbal, performance, and total IQ scores than control participants (P≤.001). Women with TBI had significantly higher total IQ scores than men with TBI (P=.04) and performed in a range similar to controls. Among controls, there was no difference between sexes (P=.29) (see table 4).

       Test Performance and Effect Sizes

      There were significant differences in raw scores between patients with TBI and controls in most cognitive domains (see table 2). All measures of information processing speed and verbal memory were significantly impaired in patients compared with controls. Within the remaining domains, significant differences in mean performance were shown for many, but not all, of the measures. For working memory, no significant differences were found. Effect sizes, computed when the P value was .05 or less, were medium to large and typically larger when the variance in the control group was used for standardization.

       Frequency of impairments According to Normative Data

      Table 3 displays the proportions of persons scoring in the impaired range (−1.5 SD) according to normative data for each test. Across all tests, 0% to 43% of the patients with TBI had clinically impaired scores in contrast to 0% to 30% among controls. The tests of information processing speed, motor function assessed by the Grooved Pegboard, and the delayed recall tasks of the CVLT-II had the highest proportion of impaired scores, while they at the same time yielded very few impaired scores in the control group. On several executive tasks, as well as on the working memory task and some measures of visual memory, attention, and vigilance, no significant difference in frequency of impairment was found between groups.
      When 1.0 SD was used as the cutoff criteria for impairment, the proportion of patients with impairments increased. Those tests showing the greatest increase in the number of patients designated as impaired were the SDMT written version (43%–53%), the grooved Pegboard (20%–41%), the CVLT-II delayed recall (29%–38%), and the Letter Fluency (16%–28%). The proportions of impaired persons were only modestly increased on other tests. Additionally, more healthy controls were diagnosed as impaired on other tests. This was in the range of 10% to 15% for several tests, including the CVLT-II and the Wisconsin Card Sorting Test. For SDMT and the TMT condition 2 and 4, relatively few controls (0%–4%) were impaired.

       Cognitive Functioning in Subgroups of Patients According to Severity of Injury

      Nine tests that are frequently used in clinical practice and that in the preceding analyses were associated with large effect sizes or good ability to discriminate patients and controls were examined further: Grooved Pegboard (dominant hand), CWIT (color naming and inhibition/switching), Verbal Fluency (letters), TMT (number sequencing and number-letter switching), SDMT (written version), CVMT (hits), and CVLT-II (delayed recall). Table 5 shows the number of impaired scores (standardized test score below 1.5 SD) out of the 9 tests. In the healthy volunteers, 98% had no more than 1 impaired score, considered as reflecting a normal performance in this sample, whereas having 2 or more impaired scores was considered cognitive impairment.
      Table 5Relative Frequency of Impaired Test Scores Out of 9 Tests According to Normative Data
      No. of Low Test Scores
      Standardized test scores <1.5 SD.
      		Moderate TBI (n=35)Severe TBI (n=26)Controls (n=47)
      013 (37)5 (19)35 (75)
      17 (20)4 (15)11 (23)
      2–310 (29)10 (38)1 (2)
      4–55 (14)3 (12)0
      6–701 (4)0
      8–903 (12)0
      NOTE. Values are n (%) or as otherwise indicated.
      low asterisk Standardized test scores <1.5 SD.
      Test performances of the patients with moderate and severe TBI are shown in table 6. Mean dpooled for the normally distributed tests (Grooved Pegboard [dominant hand], VF [letters], CWIT [color naming], and SDMT [written]) was 0.61 for moderate and 0.95 for severe TBI.
      Table 6Test Performance in Relation to Injury Severity, Moderate and Severe TBI Compared With Control Subjects
      TestModerate TBI (n=35)PEScontrol
      For tests yielding normally distributed data; Glass' d; ES is the standardized mean, using the SD of the control group. For the nonnormally distributed data, EScontrol was calculated as difference between medians divided by .75 × interquartile rangecontrol.
      		dpooled
      Cohen's d; mean difference divided by the pooled SD.
      		Severe TBI (n=26)PEScontrol
      For tests yielding normally distributed data; Glass' d; ES is the standardized mean, using the SD of the control group. For the nonnormally distributed data, EScontrol was calculated as difference between medians divided by .75 × interquartile rangecontrol.
      		dpooled
      Cohen's d; mean difference divided by the pooled SD.
      Motor function
       Grooved Pegboard; dominant hand (s)68.9±11.7.032.68.5278.0±23.7.0071.91.92
      Information processing speed
       TMT; numbers
      Data not normally distributed.
      (s)      		34 (29–40)<.0011.6735 (27.5–51)<.0011.83
       CWIT; color naming (s)33.2±8.6.013.90.6936.9±1.0.0011.681.12
       SMDT; written version47.4±11.7.011.75.6242.2±15.0.0011.411.03
      Visual memory
       CMVT; hits
      Data not normally distributed.
      		39 (36–40).270036 (31–39).0011.00
      Verbal memory
       CVLT, long-delay free recall
      Data not normally distributed.
      		11 (8–14).040.4411 (3.75–14.0).013.44
      Executive function
       Verbal Fluency; letter33.3±10.6.010.58.5931.8±11.3.005.71.71
       TMT; letter-number switching
      Data not normally distributed.
      (s)      		82 (65–105)<.0011.0381.5 (63.25–127.5).0031.11
       CWIT; inhibition/switching
      Data not normally distributed.
      (s)      		65 (54.75–78.5).0041.1372 (60–89)<.0011.85
      NOTE. Values are mean ± SD, median (25%–75%), or as otherwise indicated.
      Abbreviation: ES, effect size.
      low asterisk Cohen's d; mean difference divided by the pooled SD.
      † For tests yielding normally distributed data; Glass' d; ES is the standardized mean, using the SD of the control group. For the nonnormally distributed data, EScontrol was calculated as difference between medians divided by .75 × interquartile rangecontrol.
      ‡ Data not normally distributed.
      Of those with moderate injury, 15 (43%) demonstrated cognitive impairment, as did 17 (65%) of patients with severe TBI (P=.09; via exact z pooled test for comparison of proportions).
      When applying 1.0 SD as the cutoff criterion of impairment, 88% of controls had 0 or 1 impaired score; thus 12% would have been classified as cognitively impaired. Twenty-one patients (60%) with moderate injury and 19 patients (73%) with severe injury would be classified as cognitively impaired.

       Functional Outcome at 3 Months Postinjury

      Follow-up GOSE scores at 3 months were available for 60 patients with TBI. One individual could not be reliably assessed because of severe orthopedic injury requiring lengthy institutional rehabilitation. GOSE scores concurrent with the neuropsychological assessment were 6 or less in 45 (75%) of 60 patients with GOSE scores. Seventeen (61%) of the 28 patients with no more than 1 impaired test score and 28 (88%) of the 32 patients with 2 or more impaired test scores had concurrent GOSE scores of 6 or less.

       Functional Outcome at 12 Months Postinjury

      Follow-up GOSE scores at 12 months were available for 59 patients with TBI. Of these, 23 (39%) had moderate or severe disability as defined by a GOSE score of 6 or less. Six (21%) of 29 patients with a normal neuropsychological assessment at 3 months (no more than 1 impaired score) had GOSE scores of 6 or less, and in the group with cognitive impairment (≥2 impaired test scores), 17 (57%) of 30 had GOSE scores of 6 or less (P=.006),
      When a cutoff criterion of a GOSE score of 7 or less was applied, 13 (45%) of those with 0 or 1 impaired score and 22 (73%) of those with 2 or more impaired test scores reported disability or head injury–related complaints at 12 months' follow-up (P=.030).

      Discussion

      We studied neuropsychological test performance in patients recruited from an acute care setting, 3 months after moderate or severe TBI with parenchymal lesions detected on MRI in virtually all cases. Patients with TBI performed worse than control participants in almost all domains; this could be shown for the subpopulation of patients with moderate TBI when analyzed separately. Effect sizes were largest for tests requiring manual and processing speed. However, after moderate TBI, most patients had a normal neuropsychological assessment (no more than 1 score below 1.5 SD of normative mean). Even after severe injury, normal performances were found in one third of patients. Cognitive impairment was associated with later complaints or disability.

       Effect of TBI on Cognition

      In this study, many patients with definite TBI performed in the normal range of neuropsychological assessment measures as early as 3 months postinjury. This was unexpected, and we regard this as a novel finding in contrast with some previous studies. Novack et al
      • Novack T.A.
      • Alderson A.L.
      • Bush B.A.
      • Meythaler J.M.
      • Canupp K.
      Cognitive and functional recovery at 6 and 12 months post-TBI.
      Brain Inj. 2000; 14: 987-996
      found more pronounced cognitive decline in a sample of 72 patients who were tested 6 months postinjury. Their cohort included a higher proportion of patients with severe TBI who were recruited from a rehabilitation setting. Boake et al
      • Boake C.
      • Millis S.R.
      • High Jr, W.M.
      • et al.
      Using early neuropsychologic testing to predict long-term productivity outcome from traumatic brain injury.
      Arch Phys Med Rehabil. 2001; 82: 761-768
      and Kreutzer et al
      • Kreutzer J.S.
      • Gordon W.A.
      • Rosenthal M.
      • Marwitz J.M.
      Neuropsychological characteristics of patients with brain injury: preliminary findings from a multicenter investigation.
      J Head Trauma Rehabil. 1993; 8: 47-59
      studied patients shortly after resolution of PTA; the mean days postinjury were 42 and 50, respectively. Both research groups found a higher proportion of impaired test scores compared with normative data. The difference between these studies and ours is most likely because their patients were tested earlier than ours, and their sample was drawn from the rehabilitation setting and likely included more severe cases.
      • Corrigan J.D.
      • Selassie A.W.
      • Lineberry L.A.
      • et al.
      Comparison of the Traumatic Brain Injury (TBI) Model Systems national dataset to a population-based cohort of TBI hospitalizations.
      Arch Phys Med Rehabil. 2007; 88: 418-426
      We do not, however, consider our sample to be biased by systematic exclusion or loss of severe cases. On the contrary, patients who refused testing more often had moderate TBI than those who participated, and the 9 patients who were included after the main inclusion period had severe injuries.
      We used 1.5 SD as the cutoff level for evaluation of the standardized scores, which corresponds to a 5% rate of impairment in healthy persons. The control group performed fairly well, in line with this expectation, and thus the tests demonstrated high specificity to TBI with this definition of impairment.
      Applying 1.0 SD as a cutoff did not lead to better discrimination between patients and controls. That was to be expected because even when effect sizes are large, overlap between groups is considerable.
      • Zakzanis K.K.
      Statistics to tell the truth, the whole truth, and nothing but the truth: formulae, illustrative numerical examples, and heuristic interpretation of effect size analyses for neuropsychological researchers.
      Arch Clin Neuropsychol. 2001; 16: 653-667
      Hence, there is a need for new test paradigms or other methods of examinations with higher discriminative abilities. Reitan and Wolfson
      • Reitan R.M.
      • Wolfson D.
      Can neuropsychological testing produce unequivocal evidence of brain damage? I. Testing for specific deficits.
      Appl Neuropsychol. 2008; 15: 33-38
      argued that typical neuropsychological tests with continuous score distribution should be supplemented by tests identifying specific deficits as present or absent, yielding a dichotomous distribution. They presented promising results with few false-negative tests in a sample of patients with mixed brain injuries, by use of tests designed to identify neurologic deficits. However, the applicability of this technique in patients with TBI is unknown.
      Alternatively, a substantial proportion of the patients with TBI might actually have recovered their cognitive capacity at this early stage, a phenomenon patients demonstrated in mild TBI.
      • Schretlen D.J.
      • Shapiro A.M.
      A quantitative review of the effects of traumatic brain injury on cognitive functioning.
      Int Rev Psychiatry. 2003; 15: 341-349
      , 
      • Belanger H.G.
      • Curtiss G.
      • Demery J.A.
      • Lebowitz B.K.
      • Vanderploeg R.D.
      Factors moderating neuropsychological outcomes following mild traumatic brain injury: a meta-analysis.
      J Int Neuropsychol Soc. 2005; 11: 215-227
      Still, 65% of the patients with 0 or 1 impaired score had a concurrent GOSE score of 6 or less, indicating that head injury–related disability was present. If cognitive dysfunction mediates the functional problems, which has been previously demonstrated,
      • Novack T.A.
      • Bush B.A.
      • Meythaler J.M.
      • Canupp K.
      Outcome after traumatic brain injury: pathway analysis of contributions from premorbid, injury severity, and recovery variables.
      Arch Phys Med Rehabil. 2001; 82: 300-305
      , 
      • Rassovsky Y.
      • Satz P.
      • Alfano M.S.
      • et al.
      Functional outcome in TBI I: neuropsychological, emotional, and behavioral mediators.
      J Clin Exp Neuropsychol. 2006; 28: 567-580
      this supports the concern for low ecologic validity of common neuropsychological tests.
      • Chaytor N.
      • Schmitter-Edgecombe M.
      The ecological validity of neuropsychological tests: a review of the literature on everyday cognitive skills.
      Neuropsychol Rev. 2003; 13: 181-197
      On the other hand, injury-related factors other than impaired cognition, such as pain or affective symptoms, might also contribute to their functional decline. Furthermore, GOSE scores of 7 and 8 are partly based on self-report, and may be less reliable than neuropsychological test scores.

       Test Performance in Moderate TBI

      Patients with moderate TBI had significantly lower test scores than controls. Dikmen et al
      • Dikmen S.S.
      • Corrigan J.D.
      • Levin H.S.
      • Machamer J.
      • Stiers W.
      • Weisskopf M.G.
      Cognitive outcome following traumatic brain injury.
      J Head Trauma Rehabil. 2009; 24: 430-438
      reviewed cognitive outcome more than 6 months postinjury and concluded that the evidence of cognitive dysfunction after moderate TBI was “limited and suggestive.” In the present study, we demonstrated that moderate TBI clearly affects cognition 3 months postinjury, and future studies are needed to explore the cognitive outcome in moderate TBI later in recovery. We included patients with GCS scores of 13, a score defining mild TBI in some previous studies. However, all but 1 of these patients had a PTA duration of greater than 24 hours, and thus would have been classified as having moderate TBI according to several recommendations.
      • Ruff R.M.
      • Iverson G.L.
      • Barth J.T.
      • Bush S.S.
      • Broshek D.K.
      Recommendations for diagnosing a mild traumatic brain injury: a National Academy of Neuropsychology education paper.
      Arch Clin Neuropsychol. 2009; 24: 3-10
      , 
      • Carroll L.J.
      • Cassidy J.D.
      • Holm L.
      • Kraus J.
      • Coronado V.G.
      Methodological issues and research recommendations for mild traumatic brain injury: the WHO Collaborating Centre Task Force on Mild Traumatic Brain Injury.
      J Rehabil Med. 2004; : 113-125
      Furthermore, there is an increasing trend to classify patients with GCS scores of 13 as having moderate TBI.
      • Maas A.I.
      • Stocchetti N.
      • Bullock R.
      Moderate and severe traumatic brain injury in adults.
      Lancet Neurol. 2008; 7: 728-741

       Test Performance in Relation to MRI Findings

      DAI was found in 69% of TBI cases included in this study. DAI has previously been related to impairments of information processing speed,
      • Felmingham K.L.
      • Baguley I.J.
      • Green A.M.
      Effects of diffuse axonal injury on speed of information processing following severe traumatic brain injury.
      Neuropsychology. 2004; 18: 564-571
      executive functions, and verbal memory.
      • Fork M.
      • Bartels C.
      • Ebert A.D.
      • Grubich C.
      • Synowitz H.
      • Wallesch C.W.
      Neuropsychological sequelae of diffuse traumatic brain injury.
      Brain Inj. 2005; 19: 101-108
      , 
      • Scheid R.
      • Walther K.
      • Guthke T.
      • Preul C.
      • von Cramon D.Y.
      Cognitive sequelae of diffuse axonal injury.
      Arch Neurol. 2006; 63: 418-424
      However, in some of these studies the diagnosis of DAI was based on CT finding, which is not optimal,
      • Felmingham K.L.
      • Baguley I.J.
      • Green A.M.
      Effects of diffuse axonal injury on speed of information processing following severe traumatic brain injury.
      Neuropsychology. 2004; 18: 564-571
      , 
      • Fork M.
      • Bartels C.
      • Ebert A.D.
      • Grubich C.
      • Synowitz H.
      • Wallesch C.W.
      Neuropsychological sequelae of diffuse traumatic brain injury.
      Brain Inj. 2005; 19: 101-108
      or cases of DAI were not compared with cases without DAI.
      • Scheid R.
      • Walther K.
      • Guthke T.
      • Preul C.
      • von Cramon D.Y.
      Cognitive sequelae of diffuse axonal injury.
      Arch Neurol. 2006; 63: 418-424
      The findings in our study support the hypothesis that DAI may be an important explanation for the large effects on measures of speed found in this study. Few patients had no evidence of DAI in our study, and variances in test performance were large. Thus, power was not sufficient to compare cases with and without DAI. Hopefully, this will be explored in a future, larger study.

       Test Performance in Relation to Cognitive Domains and Evaluation of Measures

      Tests measuring processing speed, like the TMT, SDMT, and CWIT, showed the largest effect sizes. Impairments demonstrated by these tests were specific to TBI, as 20% to 30% of the individuals with TBI, but very few control participants, performed in the impaired range. The most sensitive test was SDMT. However, even for this test, reported to be very sensitive to brain insults,
      • Strauss E.
      • Sherman E.M.S.
      • Spreen O.
      A compendium of neuropsychological tests administration, norms, and commentary.
      3rd ed. Oxford University Pr, New York2006
      the percentage of patients with TBI with test performance below 1.5 SD was only 30% and 43% for the oral and written version, respectively.
      The measures of working memory (Digit Span Backwards and letter-number sequencing) were surprisingly insensitive to TBI. It is possible that these tests measure aspects of working memory that are only mildly affected by TBI, and that tests requiring more simultaneous processing, such as the n-back paradigms,
      • Hill B.D.
      • Elliott E.M.
      • Shelton J.T.
      • Pella R.D.
      • O'Jile J.R.
      • Gouvier W.D.
      Can we improve the clinical assessment of working memory? An evaluation of the Wechsler Adult Intelligence Scale–Third Edition using a working memory criterion construct.
      J Clin Exp Neuropsychol. 2010; 32: 315-323
      might be more useful in future studies of working memory deficits after TBI.
      For visual memory, measured as CVMT total score, we found very similar results in both groups. This was due to a higher rate of false positives among control participants than reported in the normative sample.
      • Trahan D.E.
      • Larrabee G.J.
      Continous Visual Memory Test.
      Psychological Assessment Resources, Odessa1988
      Given the generally strong abilities of our control group, this is difficult to explain. In a study of criterion validity of the CVMT in moderate to severe TBI, the authors concluded that the CVMT was clinically useful, but they did not make any comparisons with normative data.
      • Strong C.A.
      • Donders J.
      Validity of the Continuous Visual Memory Test (CVMT) after traumatic brain injury.
      J Clin Exp Neuropsychol. 2008; 30: 885-891
      The CCPT-II also poorly discriminated between patients with TBI and control participants. Other continuous performance test measures have been found to be sensitive to dysfunction of the attention system.
      • Riccio C.A.
      • Reynolds C.R.
      • Lowe P.
      • Moore J.J.
      The continuous performance test: a window on the neural substrates for attention?.
      Arch Clin Neuropsychol. 2002; 17: 235-272
      Thus, it is possible that CCPT-II is less useful than other continuous performance tests in patients with TBI. However, our finding may also support previous research demonstrating relatively preserved vigilance after TBI.
      • Zoccolotti P.
      • Matano A.
      • Deloche G.
      • et al.
      Patterns of attentional impairment following closed head injury: a collaborative European study.
      Cortex. 2000; 36: 93-107

       Cognitive Functioning in Relation to 1-Year Functional Outcome

      We found a higher frequency of disability 1 year postinjury in patients with cognitive impairment diagnosed 3 months postinjury; we thus confirmed that cognitive impairments in the early phase can be considered a risk factor for future disability.
      • Sherer M.
      • Sander A.M.
      • Nick T.G.
      • High Jr, W.M.
      • Malec J.F.
      • Rosenthal M.
      Early cognitive status and productivity outcome after traumatic brain injury: findings from the TBI model systems.
      Arch Phys Med Rehabil. 2002; 83: 183-192
      On the other hand, a significant proportion, 43%, of those with a normal early assessment experienced later complaints as defined by a GOSE score of 7 or less, and we consider this an interesting finding.

       Study Limitations

      The whole sample was not derived consecutively during one study period, as 15% of the patients were injured after the main inclusion period. Secondly, the neuropsychological test performances in our patients with severe TBI should not be generalized to all patients with severe TBI, but can only be applied to those who are able to cooperate with testing. All the patients with severe TBI in this study experienced a favorable outcome, with a GOSE score greater than 4 at 12 months' follow-up. The patients who were unable to cooperate with testing, however, represented a subgroup with very severe injury and poor outcome, and who even at 12 months postinjury were unable to complete a standardized neuropsychological assessment.
      Patients with TBI had significantly lower IQ scores than the control participants. This may demonstrate an adverse effect of TBI, as previously described.
      • Kersel D.A.
      • Marsh N.V.
      • Havill J.H.
      • Sleigh J.W.
      Neuropsychological functioning during the year following severe traumatic brain injury.
      Brain Inj. 2001; 15: 283-296
      However, the high IQ scores among women with TBI argue against this as the sole explanation, and most likely there were some underlying preinjury differences between patients and controls. Recruitment of a control group has been challenged in TBI research,
      • Dikmen S.S.
      • Corrigan J.D.
      • Levin H.S.
      • Machamer J.
      • Stiers W.
      • Weisskopf M.G.
      Cognitive outcome following traumatic brain injury.
      J Head Trauma Rehabil. 2009; 24: 430-438
      , 
      • Schretlen D.J.
      • Shapiro A.M.
      A quantitative review of the effects of traumatic brain injury on cognitive functioning.
      Int Rev Psychiatry. 2003; 15: 341-349
      but there is no general agreement on the method of choice. We consider a possible preinjury difference between men and women patients with TBI an interesting finding, as this could confound possible sex differences regarding effects of TBI.
      We could have included measures of symptom validity, because many individuals with TBI were injured in traffic accidents and will eventually be involved in litigation. Poor effort, however, would presumably have indicated a falsely high rather than low sensitivity of the tests.
      Finally, no systematic masking of the outcome assessor was applied, although test scores were not at hand during the GOSE scoring, and the categorization into impaired or normal was not yet performed.

      Conclusions

      Patients with moderate and severe levels of TBI were cognitively impaired compared with a healthy control group, and cognitive impairment at 3 months was associated with later disability. We demonstrated a fair specificity of impairment below 1.5 SD, as this was seldom found in the control group. On the other hand, the sensitivity to TBI of a conventional assessment was lower than expected. From a clinical perspective, we consider this an important and novel finding. A seemingly normal assessment was common in patients with recent TBI confirmed by MRI findings, despite the presence of concurrent disability in the majority of these patients. Patients' subjective report of complaints affecting daily life at 1-year follow-up was also considerable.
      Supplier
      aSPSS, Inc, 233 S Wacker Dr, 11th Fl, Chicago, IL 60606.

      Acknowledgments

      We thank neuropsychologist Brit Bjørklimo, CPsych, for assistance with collection and preparation of the neuropsychological data and reading of the manuscript; neuroradiologists Kjell Arne Kvistad, MD, PhD, and Mari Folvik, MD, for evaluation of the MRI scans; and Beate Holmqvist Karlsen for research assistance.

      Appendix 1. Explanation of How Patients were Selected for Neuropsychological Assessment

      Population
      • •
        Of 172 patients admitted, 169 consented to follow-up.
      • •
        Of these, 115 persons were 15–65 years of age.
      Reasons for exclusion
      • •
        16 persons died.
      • •
        17 persons had preinjury morbidity.
      • •
        1 person developed a chronic subdural hemorrhage requiring surgery at 3 months.
      • •
        2 persons were not fluent in Norwegian language.
      • •
        12 persons could not cooperate; of these:
      •  •
        7 had been in vegetative state 4 weeks postinjury.
      •  •
        3 had been minimally conscious at 4 weeks postinjury.
      •  •
        2 had been in PTA at 4 weeks postinjury.
      •  •
        All had poor outcome at 12 months.
      •  •
        Only 1 could be tested by 12 months.
      • •
        14 persons rejected testing at the study hospital; of these:
      •  •
        11 had a moderate TBI; 3 had severe TBI.
      •  •
        11 had a good recovery at 12 months' follow-up.
      •  •
        3 had moderate disability at 12 months' follow-up.
      Lost to follow-up
      1 person was lost to follow-up.
      52 persons included from the main cohort
      Convenience sample
      9 persons in the main database, tested at 3 months in 2008 and 2009
      Final study sample of 61 individuals with TBI

      References

        • Dikmen S.S.
        • Corrigan J.D.
        • Levin H.S.
        • Machamer J.
        • Stiers W.
        • Weisskopf M.G.
        Cognitive outcome following traumatic brain injury.
        J Head Trauma Rehabil. 2009; 24: 430-438
        View in Article
        • Mathias J.L.
        • Wheaton P.
        Changes in attention and information-processing speed following severe traumatic brain injury: a meta-analytic review.
        Neuropsychology. 2007; 21: 212-223
        View in Article
        • Willmott C.
        • Ponsford J.
        • Hocking C.
        • Schonberger M.
        Factors contributing to attentional impairments after traumatic brain injury.
        Neuropsychology. 2009; 23: 424-432
        View in Article
        • Vakil E.
        The effect of moderate to severe traumatic brain injury (TBI) on different aspects of memory: a selective review.
        J Clin Exp Neuropsychol. 2005; 27: 977-1021
        View in Article
        • Dikmen S.S.
        • Machamer J.E.
        • Winn H.R.
        • Temkin N.R.
        Neuropsychological outcome at 1-year post head-injury.
        Neuropsychology. 1995; 9: 80-90
        View in Article
        • Novack T.A.
        • Bush B.A.
        • Meythaler J.M.
        • Canupp K.
        Outcome after traumatic brain injury: pathway analysis of contributions from premorbid, injury severity, and recovery variables.
        Arch Phys Med Rehabil. 2001; 82: 300-305
        View in Article
        • Rassovsky Y.
        • Satz P.
        • Alfano M.S.
        • et al.
        Functional outcome in TBI I: neuropsychological, emotional, and behavioral mediators.
        J Clin Exp Neuropsychol. 2006; 28: 567-580
        View in Article
        • Ponsford J.
        • Draper K.
        • Schonberger M.
        Functional outcome 10 years after traumatic brain injury: its relationship with demographic, injury severity, and cognitive and emotional status.
        J Int Neuropsychol Soc. 2008; 14: 233-242
        View in Article
        • Dikmen S.
        • McLean Jr, A.
        • Temkin N.R.
        • Wyler A.R.
        Neuropsychologic outcome at one-month postinjury.
        Arch Phys Med Rehabil. 1986; 67: 507-513
        View in Article
        • Boake C.
        • Millis S.R.
        • High Jr, W.M.
        • et al.
        Using early neuropsychologic testing to predict long-term productivity outcome from traumatic brain injury.
        Arch Phys Med Rehabil. 2001; 82: 761-768
        View in Article
        • Kreutzer J.S.
        • Gordon W.A.
        • Rosenthal M.
        • Marwitz J.M.
        Neuropsychological characteristics of patients with brain injury: preliminary findings from a multicenter investigation.
        J Head Trauma Rehabil. 1993; 8: 47-59
        View in Article
        • Schretlen D.J.
        • Shapiro A.M.
        A quantitative review of the effects of traumatic brain injury on cognitive functioning.
        Int Rev Psychiatry. 2003; 15: 341-349
        View in Article
        • Strauss E.
        • Sherman E.M.S.
        • Spreen O.
        A compendium of neuropsychological tests.
        3rd ed. Oxford University Pr, New York2006
        View in Article
        • Iverson G.L.
        • Brooks B.L.
        • Holdnack J.A.
        Misdiagnosis of cognitive impairment in forensic neuropsychology.
        in: Heilbronner R.L. Neuropsychology in the courtroom: expert analysis of reports and testimony. 1st ed. The Guilford Pr, New York2008: 243-266
        View in Article
        • Schretlen D.J.
        • Testa S.M.
        • Winicki J.M.
        • Pearlson G.D.
        • Gordon B.
        Frequency and bases of abnormal performance by healthy adults on neuropsychological testing.
        J Int Neuropsychol Soc. 2008; 14: 436-445
        View in Article
        • Skandsen T.
        • Kvistad K.A.
        • Solheim O.
        • Strand I.H.
        • Folvik M.
        • Vik A.
        Prevalence and impact of diffuse axonal injury in patients with moderate and severe head injury: a cohort study of early magnetic resonance imaging findings and 1-year outcome.
        J Neurosurg. 2010; 113: 556-563
        View in Article
        • Stein S.C.
        • Spettell C.
        The Head Injury Severity Scale (HISS): a practical classification of closed-head injury.
        Brain Inj. 1995; 9: 437-444
        View in Article
        • Lezak M.D.
        • Howieson D.B.
        • Loring D.W.
        Neuropsychological assessment.
        4th ed. Oxford University Pr, New York2004
        View in Article
        • Wechsler D.
        Wechsler Abbreviated Scale of Intelligence Scale.
        The Psychological Corporation, San Antonio1999
        View in Article
        • Matthews C.G.
        • Klove K.
        Instruction manual for the Adult Neuropsychology Test Battery.
        University of Wisconsin Medical School, Madison1964
        View in Article
        • Delis D.C.
        • Kaplan E.
        • Kramer J.
        Delis Kaplan Executive Function System.
        The Psychological Corporation, San Antonio2001
        View in Article
        • Smith A.
        Symbol Digit Modality Test (SDMT) manual.
        (Ninth printing) Western Psyhological Services, Los Angeles2002
        View in Article
        • Conners C.K.
        • Staff M.
        Conners' Continuous Performance Test II (CPT II) computer program for Windows, technical guide and software manual.
        Multi-Health Systems, Toronto2000
        View in Article
        • Trahan D.E.
        • Larrabee G.J.
        Continous Visual Memory Test.
        Psychological Assessment Resources, Odessa1988
        View in Article
        • Taylor L.B.
        Localisation of cerebral lesions by psychological testing.
        Clin Neurosurg. 1969; 16: 269-287
        View in Article
        • Meyers J.E.
        • Meyers K.R.
        Rey Complex Figure Test and Recognition Trial: professional manual.
        Psychological Assessment Resources, Odessa1995
        View in Article
        • Delis D.C.
        • Kramer J.H.
        • Kaplan E.
        • Ober B.A.
        California Verbal Learning Test second edition (CVLT-II).
        (Adult version, manual) The Psychological Corporation, San Antonio2000
        View in Article
        • Wechsler D.
        Wechsler Adult Intelligence Scale.
        3rd ed. The Psychological Corporation, San Antonio1997
        View in Article
        • Heaton R.K.
        Wisconsin Card Sorting Test, computer version-2 research edition.
        Psychological Assessment Resources, Odessa1993
        View in Article
        • Wilson J.T.
        • Pettigrew L.E.
        • Teasdale G.M.
        Structured interviews for the Glasgow Outcome Scale and the extended Glasgow Outcome Scale: guidelines for their use.
        J Neurotrauma. 1998; 15: 573-585
        View in Article
        • Grissom R.J.
        • Kim J.J.
        Effect sizes for research.
        Lawrence Erlbaum, Mahwah2005
        View in Article
        • Cohen J.
        Statistical power analysis for the behavioral sciences.
        2nd ed. Erlbaum, Hillsdale1988
        View in Article
        • Lydersen S.
        • Fagerland M.W.
        • Laake P.
        Recommended tests for association in 2 × 2 tables.
        Stat Med. 2009; 28: 1159-1175
        View in Article
        • Schafer J.L.
        • Graham J.W.
        Missing data: our view of the state of the art.
        Psychol Methods. 2002; 7: 147-177
        View in Article
        • Novack T.A.
        • Alderson A.L.
        • Bush B.A.
        • Meythaler J.M.
        • Canupp K.
        Cognitive and functional recovery at 6 and 12 months post-TBI.
        Brain Inj. 2000; 14: 987-996
        View in Article
        • Corrigan J.D.
        • Selassie A.W.
        • Lineberry L.A.
        • et al.
        Comparison of the Traumatic Brain Injury (TBI) Model Systems national dataset to a population-based cohort of TBI hospitalizations.
        Arch Phys Med Rehabil. 2007; 88: 418-426
        View in Article
        • Zakzanis K.K.
        Statistics to tell the truth, the whole truth, and nothing but the truth: formulae, illustrative numerical examples, and heuristic interpretation of effect size analyses for neuropsychological researchers.
        Arch Clin Neuropsychol. 2001; 16: 653-667
        View in Article
        • Reitan R.M.
        • Wolfson D.
        Can neuropsychological testing produce unequivocal evidence of brain damage?.
        Appl Neuropsychol. 2008; 15: 33-38
        View in Article
        • Belanger H.G.
        • Curtiss G.
        • Demery J.A.
        • Lebowitz B.K.
        • Vanderploeg R.D.
        Factors moderating neuropsychological outcomes following mild traumatic brain injury: a meta-analysis.
        J Int Neuropsychol Soc. 2005; 11: 215-227
        View in Article
        • Chaytor N.
        • Schmitter-Edgecombe M.
        The ecological validity of neuropsychological tests: a review of the literature on everyday cognitive skills.
        Neuropsychol Rev. 2003; 13: 181-197
        View in Article
        • Ruff R.M.
        • Iverson G.L.
        • Barth J.T.
        • Bush S.S.
        • Broshek D.K.
        Recommendations for diagnosing a mild traumatic brain injury: a National Academy of Neuropsychology education paper.
        Arch Clin Neuropsychol. 2009; 24: 3-10
        View in Article
        • Carroll L.J.
        • Cassidy J.D.
        • Holm L.
        • Kraus J.
        • Coronado V.G.
        Methodological issues and research recommendations for mild traumatic brain injury: the WHO Collaborating Centre Task Force on Mild Traumatic Brain Injury.
        J Rehabil Med. 2004; : 113-125
        View in Article
        • Maas A.I.
        • Stocchetti N.
        • Bullock R.
        Moderate and severe traumatic brain injury in adults.
        Lancet Neurol. 2008; 7: 728-741
        View in Article
        • Felmingham K.L.
        • Baguley I.J.
        • Green A.M.
        Effects of diffuse axonal injury on speed of information processing following severe traumatic brain injury.
        Neuropsychology. 2004; 18: 564-571
        View in Article
        • Fork M.
        • Bartels C.
        • Ebert A.D.
        • Grubich C.
        • Synowitz H.
        • Wallesch C.W.
        Neuropsychological sequelae of diffuse traumatic brain injury.
        Brain Inj. 2005; 19: 101-108
        View in Article
        • Scheid R.
        • Walther K.
        • Guthke T.
        • Preul C.
        • von Cramon D.Y.
        Cognitive sequelae of diffuse axonal injury.
        Arch Neurol. 2006; 63: 418-424
        View in Article
        • Hill B.D.
        • Elliott E.M.
        • Shelton J.T.
        • Pella R.D.
        • O'Jile J.R.
        • Gouvier W.D.
        Can we improve the clinical assessment of working memory?.
        J Clin Exp Neuropsychol. 2010; 32: 315-323
        View in Article
        • Strong C.A.
        • Donders J.
        Validity of the Continuous Visual Memory Test (CVMT) after traumatic brain injury.
        J Clin Exp Neuropsychol. 2008; 30: 885-891
        View in Article
        • Riccio C.A.
        • Reynolds C.R.
        • Lowe P.
        • Moore J.J.
        The continuous performance test: a window on the neural substrates for attention?.
        Arch Clin Neuropsychol. 2002; 17: 235-272
        View in Article
        • Zoccolotti P.
        • Matano A.
        • Deloche G.
        • et al.
        Patterns of attentional impairment following closed head injury: a collaborative European study.
        Cortex. 2000; 36: 93-107
        View in Article
        • Sherer M.
        • Sander A.M.
        • Nick T.G.
        • High Jr, W.M.
        • Malec J.F.
        • Rosenthal M.
        Early cognitive status and productivity outcome after traumatic brain injury: findings from the TBI model systems.
        Arch Phys Med Rehabil. 2002; 83: 183-192
        View in Article
        • Kersel D.A.
        • Marsh N.V.
        • Havill J.H.
        • Sleigh J.W.
        Neuropsychological functioning during the year following severe traumatic brain injury.
        Brain Inj. 2001; 15: 283-296
        View in Article

      Article Info

      Footnotes

      Supported by a research grant from Liaison Committee between the Central Norway Regional Health Authority and the Norwegian University of Science and Technology .

      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.

      Identification

      DOI: https://doi.org/10.1016/j.apmr.2010.08.021

      Copyright

      © 2010 American Congress of Rehabilitation Medicine. Published by Elsevier Inc. All rights reserved.

      ScienceDirect

      Access this article on ScienceDirect
      We use cookies to help provide and enhance our service and tailor content. To update your cookie settings, please visit the for this site.
      Copyright © 2022 Elsevier Inc. except certain content provided by third parties. The content on this site is intended for healthcare professionals.


      RELX