Abstract

Few studies address the extent to which, during the process of litigation, individuals with moderate-to-severe traumatic brain injury might malinger in their performance on neuropsychological assessment batteries. This study explored whether financial settlement influenced neuropsychological test performance and activities of daily living in litigants (N = 31) who were tested and interviewed both during litigation and 1 year or more after case settlement. Results showed that neuropsychological test scores did not change from assessment during forensic proceedings to assessment after settlement. Although some improvement was evident in activities of daily living, the gains were small and their clinical significance questionable. We found no evidence that individuals with moderate-to-severe TBI, despite clear potential for secondary gain, malingered or delivered sub-optimal effort during neuropsychological evaluation taking place in the context of litigation.

Introduction

About 40% of mild traumatic brain injury (mTBI) patients seeking compensation have been found to embellish deficits (Mittenberg, Patton, Canyock, & Condit, 2002). The extent to which moderate-to-severe TBI litigants tend to embellish deficits as a means of ensuring success in the quest for compensation is not as clear. In contrast to the voluminous literature on long-term outcome in TBI that does not take into account the presence of litigation (see, e.g., Colantonio et al., 2004; Hammond, Hart et al., 2004; Millis et al., 2001; Oddy, Coughlan, Tyerman, & Jenkins, 1985; Olver, Ponsford, & Curran, 1996; Wood, 2008), there is limited literature on long-term outcome in TBI patients who have been through litigation (and particularly on changes in cognitive and functional status from assessments that form part of the litigation process to assessments post-litigation). This study investigated the relationship, in an adult sample of moderate-to-severe TBI individuals, between involvement in litigation and long-term cognitive and functional outcome.

Most research investigating the relationship between litigant status and malingering suggests that, compared with non-litigants, litigants report more post-concussional symptoms and experience higher levels of psychological distress, and that their symptoms persist for longer and are more debilitating in terms of return to work (Fernstein, Ouchterlony, Somerville, & Jardine, 2001; Miller & Donders, 2001; Paniak et al., 2002). Furthermore, there is reportedly a positive association between seeking financial compensation and the likelihood of poor outcome in mTBI (Evans, 1994; Reynolds, Paniak, Toller-Lobe, & Nagy, 2003). Binder and Rohling's (1996) meta-analysis found an inverse relationship between the importance of financial incentives and severity of closed-head injury: Monetary incentive was stronger for patients with mTBI, who presented with “more abnormality and disability” (p. 7), than for those with moderate or severe TBI. They concluded that mTBI individuals who present with severe cognitive deficits for many months after their head injury, and who might be financially compensated due to ongoing impaired function, might be suspected of malingering.

Subsequent studies have confirmed that mTBI patients appear to present with a greater tendency toward malingering than do moderate/severe TBI patients. The same studies have also documented definite malingered neurocognitive dysfunction in the latter group, however. For instance, Bianchini, Greve, and Love (2003) found that moderate/severe TBI patients performed significantly below-chance levels on at least one forced-choice symptom validity test (the Portland Digit Recognition Test or Test of Memory Malingering). Similarly, Boone and Lu (2003) found failed performance on multiple tests of effort (Rey 15-Item Memorization Recognition Score, Rey Dot Counting Test, and Rey Word Recognition Test) and inferred malingering in moderate/severe TBI patients based on (a) non-credible performance on neuropsychological tests and (b) inconsistency between test scores and activities of daily living documented via surveillance tapes.

Other studies investigating malingering in moderate/severe TBI patients have delivered results that contrast with those presented above. For instance, Wood and Rutterford (2006), in a 10-year follow-up study, assessed 31 individuals with severe brain injury who were involved in litigation. They found, based on a comparison of pre- and post-settlement data, that those individuals did not underachieve on the neuropsychological assessment conducted during litigation. McKinlay, Brooks, and Bond (1983) found similar results in a 12-month follow-up study of 21 severe TBI litigants (initially assessed during their period of litigation) and 21 severe TBI non-litigants. Specifically, they found few between-group differences in terms of post-concussional symptoms and cognitive performance, amongst other outcome variables. Taken together, the results from these studies suggest that claimants with severe TBI may not malinger on neuropsychological tests administered during the litigation process.

Even in studies that do find a relationship between the presence of financial incentives and poorer long-term outcome, the mechanisms driving that relationship are not clear. Although Binder and Rohling (1996) make a strong case for malingering as a primary explanatory factor, their argument is not accepted universally. For instance, Suhr, Tranel, Wefel, and Barrash (1997) studied, amongst others, the contributions of malingering and litigation status to memory performance after mTBI. They found that “several memory tests were useful in distinguishing probable malingerers from the other groups” (p. 500, emphasis added). However, they described a complex interaction between non-neurological factors such as malingering, litigation status, affective states, personality traits, and use of medication in the prediction of memory test results; litigation status, by itself, was not a statistically significant predictor. The authors emphasized the need to take the sum of these non-neurological factors into consideration when interpreting poor memory performance in forensic cases. Other researchers agree that social, emotional, and psychological factors (e.g., stress associated with litigation, victimization, feelings of guilt, and emotional adjustment), as well as environmental factors and educational level, may contribute to the relationship between financial incentives and poor cognitive and psychosocial outcome in litigants (Hofman et al., 2001; Reynolds et al., 2003).

In summary, the role of malingering or, more broadly, of non-neurological factors, in functional outcome and neuropsychological test performance after TBI, is a matter of ongoing uncertainty. There are, however, to our knowledge, only two published studies that explore directly whether moderate-to-severe TBI patients who have been part of a litigation process improved after the litigation process was concluded (McKinlay et al., 1983; Wood & Rutterford, 2006). Both of these studies compared litigants with non-litigants. In the current study, however, as in other TBI outcome research (e.g., Millis et al., 2001), the emphasis is on within-group change over the long term as opposed to between-group differences measured at different points over time.

This study, then, investigated cognitive and functional recovery from closed head injury in a group of TBI individuals tested both during litigation (at Time 1, or T1) and 1 year or more after case settlement (at Time 2, or T2). Our primary objectives were to assess changes in neuropsychological test scores, and in functional status/activities of daily living (ADLs; the ability to perform self-care, self-maintenance, and physical activities), between T1 and T2 in order to examine longitudinally what influence litigation may have on clinical presentation in TBI.

The questions addressed here are similar to those addressed by Wood and Rutterford (2006). There are, however, notable differences in the design of our study compared with theirs. First, as noted above, the current study used a within-subjects, rather than between-subjects, design. Second, we administered an identical battery of neuropsychological tests to participants during litigation (T1) and at the post-litigation (T2) assessment. Additionally, we are the first to publish data of this nature from a low- to middle-income country. Although organic brain damage resulting from TBI obviously manifests in a similar manner across the world, socioeconomic circumstances vary from country to country, and access to treatment and rehabilitation vary along with them. South Africa, a country with large disparities in income levels (a Gini coefficient of 65; World Bank, 2009), has a sizeable population of TBI individuals who have no access to any form of rehabilitation (Goosen, Bowley, Degiannis, & Plani, 2003; Levin, 2004). Such non-rehabilitated TBI populations have rarely been the focus of TBI studies, most of which emerge from high-income, developed-world countries.

The study's design allowed us to make the following predictions: (i) If participants delivered less-than-optimal effort at T1 (i.e., when being assessed for litigation purposes), but genuine effort at T2 (i.e., when outside of litigation and being assessed for research purposes only), then they would perform significantly better on the neuropsychological test battery at T2 than at T1, and (ii) because most previous studies (Bercaw, Hanks, Millis, & Gola, 2011; Hammond, Hart, Bushnik, Corrigan, & Sasser, 2004) suggest that symptom presentation tends to stabilize at around 2 years post-injury, self-reported cognition, behavior, and functional status/ADL capability would not be significantly different from T1 to T2, given that the initial assessment took place more than 2 years post-injury.

Methods

Participants

Two hundred and thirty-five potential participants were identified using case files from two neuropsychological practices. Contained in each case file were complete records of neuropsychological assessments conducted for forensic purposes. All of these assessments occurred at least 2 years after the potential participant had sustained a head injury in a motor vehicle accident (MVA).

We screened case files for individuals who (i) were fluent in either English or Afrikaans; (ii) were aged 16 years or older at the time of the MVA; (iii) had no history of TBI or neurological dysfunction before the MVA; and (iv) had undergone their first neuropsychological assessment within 5 years of the MVA. Of the 95 potential participants who satisfied these criteria, 32 could not be traced and 7 declined to participate. Nine participants lived too far (more than 500 km) from the research team's base to be interviewed, three lived in inaccessible areas, and four did not keep their appointments despite repeated scheduling attempts. Finally, 40 potential participants (25 males) were able and willing to give informed consent and were interviewed and assessed successfully. To prevent confounds and interpretational complexity and to ensure reasonable injury-related homogeneity within our sample, we present data here for only the 31 (18 males) of those 40 individuals who had sustained a moderate-to-severe TBI.

Only 11 moderate-to-severe TBI participants who were otherwise eligible (i.e., who were fluent in English or Afrikaans, who were older than 16 years at the time of the MVA, and who had no history of TBI or neurological dysfunction before the MVA) did not participate. Reasons for non-participation or exclusion of those 11 individuals were as follows: 4 individuals could not be contacted; 3 had been assessed for litigation purposes more than 5 years after their MVA; 3 declined to participate; and 1 was excluded because of researcher errors during administration of the neuropsychological battery at T2.

Participants were blind to the purpose of the study. None had experienced any form of cognitive rehabilitation.

Design and Procedure

The study involved a longitudinal repeated-measures design (neuropsychological test data were collected from the same participants at two different times, first in a litigation context and then in a post-litigation research context), and utilized retrospective data. Data were collected using a battery of standard neuropsychological tests and the Head Injury-Family Interview (HI-FI; Kay, Cavallo, Ezrachi, & Vavagiakis, 1995). Information about independent variables and outcome measures were derived from: (a) forensic case files, which included complete records of neuropsychological testing at T1; (b) neuropsychological testing at T2; and (c) separate structured interviews, using the HI-FI, with the participant and a partner or relative of him/her, at T2.

Potential participants were contacted initially by telephone. During that conversation, they were informed that a follow-up study was being conducted for the purposes of gaining insight into long-term sequelae of TBI, and appointments were made for them and their informant to be interviewed and tested at a location convenient to them. Most (n = 22) interviews and testing sessions took place in an office at Groote Schuur Hospital, Cape Town; the rest (n = 9) took place at the participant's home. Each session lasted approximately 3 h. After filling out a questionnaire enquiring about basic sociodemographic information, the participant was administered the neuropsychological test battery and the HI-FI while the informant completed the questionnaires designed for him/her.

The Research Ethics Committee of the University of Cape Town's Department of Psychology granted ethical approval for the study. Participants were not compensated for their involvement in the study.

Measures

Forensic Case Files

These files consisted of information (e.g., ambulance notes, hospital notes, and medical expert reports) related to the MVA. The hospital notes typically included, but were not limited to, Glasgow Coma Scale (GCS; Teasdale & Jennett, 1974) scores, the nature of the trauma, results of CT or MRI scans, medication administered, type of injury, and treatment of injury. The medical expert reports typically included, but were not limited to, reports from neurologists, neurosurgeons, occupational therapists, and psychologists. The neuropsychological report provided information about first recorded GCS (usually taken from either the scene of the accident or upon arrival at hospital) and length of post-traumatic amnesia (PTA).

Sociodemographic Questionnaire

At T2, the participant was asked to provide information regarding age at the time of accident, months since injury, years of education, and employment status at T1 and T2.

Neuropsychological Test Battery

We selected the set of neuropsychological tests that was administered most frequently at T1 for administration at T2. The test battery, which comprised the tests listed below and which was administered in the participant's home language (10 in English, 21 in Afrikaans), covered the domains of memory, executive functioning, attention, language, visuo-spatial functioning, and motor functioning. Theron and colleagues (2001) reported that the FCT recognition trial was an effective test of malingering in a student sample. Hugo and colleagues (2001) used the test in an attempt to detect malingering in a sample of neuropsychiatric patients (N = 38). They found that 40% (n = 15) of their participants met the FCT recognition trial criteria for malingering, in contrast to 25% (n = 9) who met the criterion on a threshold scale (Rogers, 1997) more likely to reflect the true rate of malingering in the sample. There are no published reports of the use of this test in TBI samples, although it is widely used in clinical practice and in forensic assessment in South Africa. These interviews were adapted to suit the repeated-measures design of the study. Specifically, for each item on the first interview, the participant was asked whether or not the symptom was experienced at T1. If the response was yes, he/she was asked to rate the degree (severity) to which the symptom affected his/her daily life (on a scale of 1 to 7) at T1. The participant was then asked whether the symptom severity had changed since that time, and, if yes, how it would rate presently (on the same 1–7 scale). If the symptom had resolved, it was scored 0. The second interview was administered to the informant in much the same fashion.

  • The WAIS-III Block Design, Digit Span, and Digit Symbol-Coding subtests (Wechsler, 1997). We used data from the Digit Span task to derive a Reliable Digit Span (RDS; Greiffenstein, Baker, & Gola, 1994) score for each participant, and used that score as an embedded measure of effort.

  • The Purdue Pegboard test (Lafayette Instrument Company, 1999).

  • The Rey-Osterrieth Complex Figure test (ROCF; Meyers & Meyers, 1995). Only the Copy and Immediate Recall trials were administered. One potential point of difficulty in this study is that the ROCF was not administered consistently at T1. In some instances, the Rey-Osterrieth Complex Figure Test: Form B (Taylor alternate version) was administered. In contrast, the standard version of the ROCF was administered to all participants at T2. When the ROCF and the alternate Taylor figure are compared with one another, reliability coefficients are in the moderate range (Strauss, Sherman, & Spreen, 2006). Some comparison between the tests is possible but using them as parallel forms is not optimal for research purposes. Hence, the ROCF results presented here should be interpreted with caution.

  • The Rey Auditory Verbal Learning Test (RAVLT; Lezak, Howieson, & Loring, 2004).

  • The Trail Making Test (TMT; Reitan, 1955).

  • The Controlled Oral Word Association Test (COWAT; Benton & Hamsher, 1976). For English first-language participants, we used the letter set FAS to test phonemic fluency. For Afrikaans first-language participants, we used the letter set MAS. To test semantic fluency, we used the category “animals” for both Afrikaans- and English-speaking participants.

  • The Forced-Choice Test (FCT; Hugo et al., 2001). This instrument is designed to detect simulated memory impairment. It consists of a 21-item learning and recall list, where each word is a commonly used one-or two-syllable object noun, and a 21-item paired-associates recognition task. The latter consists of seven rhyming word pairs (e.g., brain–train), seven semantically similar word pairs (e.g., hand–finger), and seven semantically unrelated and non-rhyming word pairs (e.g., stars–bread). The examiner reads the 21-item word list to the examinee, and then asks him/her to recall as many of the words as possible. After the completion of that free recall task, the examiner presents the recognition task; the examinee is asked to guess which one of the two words in each pair was in the original list.

  • Psychosocial outcome. The Head Injury-Family Interview, Version 2.0 (HI-FI; Kay et al., 1995), which we administered at T2 only, consists of four distinct interviews. For the purposes of this study, we utilized only two of those four: (a) Interview for the Person with the Head Injury Problem Check List and (b) Significant Other Interview Problem Check List, which includes as a subsection a questionnaire on activities of daily living.

The Activities of Daily Living (ADL) questionnaire (Kay et al., 1995) required the informant to rate the patient's ability to perform common daily activities at T1 and T2, separately. The questionnaire consists of 19 items, each rated on a 5-point scale ranging from 0 (“unable to do at all, even without assistance”) to 4 (“does independently, without prompting”).

Data Management and Statistical Analysis

Regarding data from the neuropsychological test battery, we used raw scores in our analyses because of a lack of locally appropriate norms (Foxcroft, Paterson, Le Roux, & Herbst, 2004; Paterson & Uys, 2005; Shuttleworth-Jordan, 1996); the only exceptions were the WAIS Digit Symbol-Coding and Block Design subtests, for which local norms are available. Regarding data from the informant and participant-completed questionnaires, we derived three composite scores (Affective/Behavior, Cognitive, and Physical Dependency) from the HI-FI Problem Checklist scores by adding the item scores and dividing the sum by the number of items rated. Severity scores were used to compare symptom presence, as a severity score in itself indicates the presence of the symptom. Regarding the ADL questionnaire, we derived two composite scores by adding T1 and T2 scores, respectively, and dividing each sum by the number of items answered at each assessment session.

Once these manipulations were complete, we tested our first prediction across three steps. First, we used a series of paired-sample t-tests to assess differences between performances at T1 and T2 on all neuropsychological test outcome variables. Due to the small sample size and large number of pairwise comparisons, we applied the appropriate Bonferroni correction. Second, we used the Reliable Change Index (RCI; Millis et al., 2001) to investigate whether changes from T1 to T2 in neuropsychological test scores were clinically meaningful. We calculated the RCI with the Reliable Change Criterion Calculator, using ±1.96 standard deviation to establish a 95% change score confidence limit (Evans, 1998). Third, we used two separate hierarchical regression models to determine whether there were linear or curvilinear associations between measures of effort at T1 and change in neuropsychological test performance from T1 to T2. The first of these models included the standardized FCT Recognition T1 scores and the squared values of those scores as predictor variables; the second included standardized RDS at T1 scores and the squared values of those scores as predictor variables. In each case, the criterion variable was the difference between neuropsychological test performance at T2 and that at T1 (with performance in each case captured by a composite derived from the average z-score across all test outcome variables).

To test our second prediction, we used a series of paired-sample t-tests to assess T1 to T2 differences on all of the HI-FI outcome measures (viz., separate self- and informant-reported composite scores for symptom presence and symptom severity in the domains of behavior, cognition, and physical dependency at T1 and T2, as well as for an informant-reported composite score for ADL capability at T1 and T2). To explore changes in ADL capability between T1 and T2, we conducted item-by-item analyses of informant ADL reports. Again, we applied the appropriate Bonferroni correction.

We conducted all analyses using STATISTICA 7.0 and SPSS versions 18 and 20. The threshold for statistical significance was set at p = .05 for all decisions, unless otherwise indicated. For paired-sample t-tests, we used Hedge's g, which is congruent with Cohen's d but adjusted for small sample bias (Cohen, 1988), to estimate the effect size associated with each comparison.

Results

Sample Characteristics

Tables 1 and 2 present complete demographic and clinical characteristics of the sample. As can be seen, there was a wide range of participant ages at injury and at each of the testing occasions. On average, however, the interval between T1 and T2 was <4 years. Regarding education, 15 (48.39%) of the participants had completed high school (12 years of education), and 8 (25.81%) had completed 15 or more years of education.

Table 1.

Demographic characteristics of the participants (N = 31)

Variable M (SDRange 
Age (years) 37.03 (9.75) 24–60 
Age at injury (years) 30.29 (10.28) 16–55 
Years since injury (at T2a6.91 (1.99) 3.67–10.49 
Age at T1 (years) 33.29 (10.12) 19–57 
Years between T1 and T2 3.74 (1.82) 1.24–7.93 
Years since settlement (at T2) 2.45 (1.21) 1.00–6.58 
Years of completed education 11.06 (3.47) 4–16 
Glasgow Coma Scale scoreb 6.10 (2.44) 2–12 
Variable M (SDRange 
Age (years) 37.03 (9.75) 24–60 
Age at injury (years) 30.29 (10.28) 16–55 
Years since injury (at T2a6.91 (1.99) 3.67–10.49 
Age at T1 (years) 33.29 (10.12) 19–57 
Years between T1 and T2 3.74 (1.82) 1.24–7.93 
Years since settlement (at T2) 2.45 (1.21) 1.00–6.58 
Years of completed education 11.06 (3.47) 4–16 
Glasgow Coma Scale scoreb 6.10 (2.44) 2–12 

aT1 refers to the first neuropsychological assessment following the MVA (i.e., that in the context of the forensic assessment); T2 refers to the post-settlement assessment.

bRefers to GCS score upon admission to hospital immediately following the MVA; data available for n = 30 participants.

Table 2.

Employment status pre-accident, at T1, and at T2

Status Pre-accident T1 T2 
Employed 96.81% (30) 67.74% (21) 74.19% (23) 
Unemployed 3.23% (1) 32.26% (10) 25.80% (8) 
Financially gainfully employed 83.87% (26) 51.61% (16) 54.84% (17) 
Type of employment 
 Professional/office work 35.48% (11) 22.58% (7) 25.81% (8) 
 Secondary sector 48.39% (15) 9.68% (3) 16.13% (5) 
 Homemaker 
 Student 12.90% (4) 6.45% (2) 
 Retired 3.23% (1) 6.45% (1) 
 Informal assistance 9.68% (3) 22.58% (7) 
 Manual labor 16.13% (5) 6.45% (2) 
Status Pre-accident T1 T2 
Employed 96.81% (30) 67.74% (21) 74.19% (23) 
Unemployed 3.23% (1) 32.26% (10) 25.80% (8) 
Financially gainfully employed 83.87% (26) 51.61% (16) 54.84% (17) 
Type of employment 
 Professional/office work 35.48% (11) 22.58% (7) 25.81% (8) 
 Secondary sector 48.39% (15) 9.68% (3) 16.13% (5) 
 Homemaker 
 Student 12.90% (4) 6.45% (2) 
 Retired 3.23% (1) 6.45% (1) 
 Informal assistance 9.68% (3) 22.58% (7) 
 Manual labor 16.13% (5) 6.45% (2) 

Note: “Secondary sector” refers to employment in the service industry (e.g., working in a factory or as a security guard). “Informal assistance” refers to unsalaried casual assistance to others.

Regarding TBI-related variables, all of the participants for whom GCS data were available (n = 30) had a score of ≤12 (i.e., their TBI was graded in the moderate-to-severe range, given conventional classification criteria), with a mean score of <7. Durations of PTA were consistent with this grading: 1 participant had a duration of 1–24 h [described as “moderate” injury, following the taxonomy outlined by Lezak et al. (2004, p. 160)]; 6 had a duration of 1–7 days (severe injury); 13 had a duration of 1–4 weeks (very severe injury); and 11 had a duration of >4 weeks (extremely severe injury). Twenty-six participants (83.87%) had a positive CT scan, with 19 (61.29%) showing diffuse axonal injury.

Regarding changes in employment status over the course of the study, there was a small and non-significant increase of 6.45% (two more participants were employed at T2 than at T1; only one of these was financially gainfully employed). Of note here is the fact that, at both T1 and T2, participants were employed at a lower economic level than pre-morbidly.

Preliminary analyses suggested that there were no statistically significant between-group differences on any of the dependent variables with regard to language of test administration (English versus Afrikaans), or with regard to location of testing (hospital versus home). Hence, the data presented below are for the entire sample of 31 participants, not subdivided by either of those grouping variables.

Prediction 1: Changes in Neuropsychological Test Performance from T1 to T2

As Table 3 shows, the series of paired sample t-tests confirmed that there were statistically significant changes at the conventional p < .05 level on the following tests only: phonemic fluency (an improvement in mean scores from T1 to T2), Purdue Pegboard (a decline in average performance for non-dominant hand only), and AVLT Trials 1–5 total score (a decline in average performance). Taking the Bonferroni correction for multiple pairwise comparisons into consideration, however, none of the comparisons remained statistically significant. Table 3 also shows that the estimated effect sizes associated with all of these comparisons ranged, using conventional criteria (Cohen, 1988), from small to medium.

Table 3.

Comparison of neuropsychological test performance at T1 and T2

Outcome variable T1
 
T2
 
df t p Hedge's g 
n M SD n M SD 
WAIS-III           
 DS 31 11.81 3.17 31 12.00 3.44 30 −0.49 .63 0.06 
 BD 29 20.76 10.42 29 20.83 11.29 28 −0.78 .94 0.01 
 DS-Cd 29 33.72 18.37 30 34.23 17.50 28 −0.11 .92 0.03 
Phonemic fluency 28 27.67 13.81 30 30.77 15.06 27 −2.55 .02* 0.21 
Semantic fluency 31 21.48 10.48 31 22.90 11.51 30 −1.35 .19 0.13 
AVLT 
 Trial 5 29 9.76 3.28 31 9.45 3.76 28 1.10 .25 0.09 
 Trials 1–5 total 30 40.73 13.40 31 37.06 14.24 29 2.57 .02* 0.26 
 Delayed recall 30 6.97 4.63 31 6.52 4.40 29 1.08 .29 0.10 
Purdue pegboard 
 Dominant 27 13.11 4.72 26 10.81 4.54 22 1.35 .19 0.49 
 Non-dominant 30 11.13 4.98 26 9.81 4.66 25 2.92 .007* 0.27 
ROCF 
 Copy 21 28.29 12.08 29 27.88 11.34 20 0.37 .71 0.03 
 Recall 21 14.00 10.21 28 14.07 10.32 19 −1.15 .27 0.01 
TMT 
 Part A 30 48.73 30.53 30 50.07 25.52 29 −0.36 .72 0.06 
 Part B 30 78.27 55.51 30 90.47 60.46 29 −1.50 .14 0.21 
Tests of effort 
 FCT           
  Recall 27 4.33 1.71 30 4.30 1.73 26 −1.12 .91 0.02 
  Recognition 27 16.59 2.47 30 16.10 2.40 26 1.12 .25 0.20 
 RDS 31 8.16 2.05 31 8.10 1.62 30 0.21 .84 0.03 
Outcome variable T1
 
T2
 
df t p Hedge's g 
n M SD n M SD 
WAIS-III           
 DS 31 11.81 3.17 31 12.00 3.44 30 −0.49 .63 0.06 
 BD 29 20.76 10.42 29 20.83 11.29 28 −0.78 .94 0.01 
 DS-Cd 29 33.72 18.37 30 34.23 17.50 28 −0.11 .92 0.03 
Phonemic fluency 28 27.67 13.81 30 30.77 15.06 27 −2.55 .02* 0.21 
Semantic fluency 31 21.48 10.48 31 22.90 11.51 30 −1.35 .19 0.13 
AVLT 
 Trial 5 29 9.76 3.28 31 9.45 3.76 28 1.10 .25 0.09 
 Trials 1–5 total 30 40.73 13.40 31 37.06 14.24 29 2.57 .02* 0.26 
 Delayed recall 30 6.97 4.63 31 6.52 4.40 29 1.08 .29 0.10 
Purdue pegboard 
 Dominant 27 13.11 4.72 26 10.81 4.54 22 1.35 .19 0.49 
 Non-dominant 30 11.13 4.98 26 9.81 4.66 25 2.92 .007* 0.27 
ROCF 
 Copy 21 28.29 12.08 29 27.88 11.34 20 0.37 .71 0.03 
 Recall 21 14.00 10.21 28 14.07 10.32 19 −1.15 .27 0.01 
TMT 
 Part A 30 48.73 30.53 30 50.07 25.52 29 −0.36 .72 0.06 
 Part B 30 78.27 55.51 30 90.47 60.46 29 −1.50 .14 0.21 
Tests of effort 
 FCT           
  Recall 27 4.33 1.71 30 4.30 1.73 26 −1.12 .91 0.02 
  Recognition 27 16.59 2.47 30 16.10 2.40 26 1.12 .25 0.20 
 RDS 31 8.16 2.05 31 8.10 1.62 30 0.21 .84 0.03 

Note: Sample size column (n) represents the number of participants who completed the relevant test at T1 and T2.

WAIS-III = Wechsler Adult Intelligence Scale-Third Edition; DS = Digit Span; BD = Block Design; DS-Cd = Digit Symbol-Coding; AVLT = Auditory Verbal Learning Test; ROCF = Rey-Osterrieth Complex Figure Test; TMT = Trail Making Test; FCT = Forced Choice Test; RDS = Reliable Digit Span.

*p < .05.

**Significant after Bonferroni correction at p < .003.

We conducted RCI analyses for those neuropsychological test variables that were statistically significantly different (without the Bonferroni correction) from T1 to T2, or whose change from T1 to T2 was associated with an effect size >0.30. Table 4 shows that the only tests on which any of the participants showed reliable improvement were phonemic fluency (one participant) and Purdue Pegboard (dominant hand; two participants). One participant showed a reliable decline in AVLT Trials 1–5 total score, three participants showed a reliable decline in performance on the Purdue Pegboard (dominant hand) and two participants showed a reliable decline in performance on the Purdue Pegboard (non-dominant hand).

Table 4.

Reliable change index intervals from T1 to T2 for neuropsychological tests

Outcome variable N Prediction interval (±) Percent improved Percent deteriorated 
Phonemic fluency 28 18.36 3.60 
AVLT 
 Trials 1–5 total 30 20.34 3.33 
Purdue Pegboard 
 Dominant 26 5.07 7.69 11.54 
 Non-dominant 26 5.32 7.70 
Outcome variable N Prediction interval (±) Percent improved Percent deteriorated 
Phonemic fluency 28 18.36 3.60 
AVLT 
 Trials 1–5 total 30 20.34 3.33 
Purdue Pegboard 
 Dominant 26 5.07 7.69 11.54 
 Non-dominant 26 5.32 7.70 

Note: Sample size column (n) represents the number of participants who completed the relevant test at both T1 and T2.

AVLT = Auditory Verbal Learning Test.

Regarding performance on the effort tests, as Table 3 shows, there were no statistically significant differences in either FCT or RDS scores between T1 and T2. Furthermore, there was no significant difference in the proportion of participants who “failed” both tests of effort (i.e., obtained scores of <16 on the FCT and <7 on the RDS) at T1 (4 of 31) compared with that at T2 (3 of 31), χ2(1) = 1.23, p = .27, Cramer's V = .20.

To investigate whether there were relationships between T1 effort test scores and changes in neuropsychological test performance from T1 to T2, we conducted two separate hierarchical regression analyses. The first model suggested that there was no linear (B = 0.04, SE B = 0.04, β = 0.19, R2 = .04), or curvilinear (B = −0.01, SE B = 0.03, β = −0.06, R2 = .04), relationship between FCT Recognition scores at T1 and difference in test performance from T1 to T2. The second showed similar results: There was no linear (B = −0.01, SE B = 0.02, β = −0.04, R2 = .01), or curvilinear (B = −0.03, SE B = 0.05, β = 0.15, R2 = .02), relationship between RDS scores at T1 and difference in test performance from T1 to T2.

Prediction 2: Changes in Cognitive, Behavioral, and ADL Functioning from T1 to T2

Table 5 shows the results of the series of paired sample t-tests comparing the T1 and T2 HI-FI outcome measures (viz., separate participant- and informant-reported composite scores for symptom presence and symptom severity in the domains of behavior, cognition, and physical dependency at T1 and T2, as well as for an informant-reported composite score for ADL capability at T1 and T2).

Table 5.

Self- and informant-reported symptom presence at T1 and T2 (N = 31)

 T1a T2 df t p Hedge's g 
Patient 
 Symptom presence       
  Behavior 0.55 (0.17) 0.42 (0.24) 30 6.55 <.001** 0.62 
  Cognition 0.67 (0.29) 0.67 (0.30) 30 1.00 .33 0.00 
  PD 0.5 (0.22) 0.5 (0.20) 30 0.00 1.00 0.00 
 Symptom severity       
  Behavior 1.81 (1.40) 1.48 (1.03) 30 1.64 .11 0.27 
  Cognition 2.94 (1.84) 2.47 (1.61) 30 2.61 .01* 0.27 
  PD 2.21 (1.41) 1.59 (1.07) 30 3.74 .001** 0.49 
Informant 
 Symptom presence       
  Behavior 0.67 (0.25) 0.64 (0.25) 30 2.18 .04* 0.12 
  Cognition 0.75 (0.33) 0.75 (0.31) 30 0.00 1.00 0.00 
  PD 0.59 (0.27) 0.60 (0.27) 30 −0.57 .57 0.04 
 Symptom severity       
  Behavior 2.84 (1.79) 2.49 (1.76) 30 1.24 .23 0.19 
  Cognition 3.43 (1.90) 3.15 (2.01) 30 0.81 .42 0.14 
  PD 2.35 (1.54) 2.35 (1.54) 30 – – – 
 ADL 2.56 (1.28) 3.01 (0.99) 30 −2.87 .008* 0.39 
 T1a T2 df t p Hedge's g 
Patient 
 Symptom presence       
  Behavior 0.55 (0.17) 0.42 (0.24) 30 6.55 <.001** 0.62 
  Cognition 0.67 (0.29) 0.67 (0.30) 30 1.00 .33 0.00 
  PD 0.5 (0.22) 0.5 (0.20) 30 0.00 1.00 0.00 
 Symptom severity       
  Behavior 1.81 (1.40) 1.48 (1.03) 30 1.64 .11 0.27 
  Cognition 2.94 (1.84) 2.47 (1.61) 30 2.61 .01* 0.27 
  PD 2.21 (1.41) 1.59 (1.07) 30 3.74 .001** 0.49 
Informant 
 Symptom presence       
  Behavior 0.67 (0.25) 0.64 (0.25) 30 2.18 .04* 0.12 
  Cognition 0.75 (0.33) 0.75 (0.31) 30 0.00 1.00 0.00 
  PD 0.59 (0.27) 0.60 (0.27) 30 −0.57 .57 0.04 
 Symptom severity       
  Behavior 2.84 (1.79) 2.49 (1.76) 30 1.24 .23 0.19 
  Cognition 3.43 (1.90) 3.15 (2.01) 30 0.81 .42 0.14 
  PD 2.35 (1.54) 2.35 (1.54) 30 – – – 
 ADL 2.56 (1.28) 3.01 (0.99) 30 −2.87 .008* 0.39 

Notes: PD = physical dependency; ADL = activities of daily living.

aIn this column and in the column headed T2, means are presented with standard deviations in parentheses.

*p < .05.

**Significant after Bonferroni correction at p < .004.

Initial analyses suggested that (a) regarding symptom presence, there were statistically significant declines from T1 to T2, reported by both the patient and the informant, in the domain of behavior, (b) regarding symptom severity, there were statistically significant declines from T1 to T2 reported by the patient (but not by the informant) in the domains of cognition and physical dependency, and (c) regarding ADL capability, the informant reported significant declines from T1 to T2. When Bonferroni corrections were applied, however, only two T1 to T2 declines remained statistically significant: self-reported presence of behavioral symptoms and self-reported severity of physical dependency. The effect size associated with the former was in the range described conventionally as large, whereas the effect size associated with the latter was in the medium range.

To explore changes in ADL capability between T1 and T2 further, we conducted item-by-item analyses of informant ADL reports (see Table 6). After application of the Bonferroni correction, “shopping for food” was the only ADL capability that showed statistically significantly change (improvement, with a large effect size) from T1 to T2. Before the Bonferroni correction was applied, there was also a tendency toward general improvement with regard to basic self-care (e.g., preparing and cleaning up after meals, feeding self, brushing teeth, washing hair, going to the toilet, and choosing own clothes) as well as more complex instrumental activities (e.g., paying bills and helping with household chores).

Table 6.

Report of informant on ADL capabilities of patients at T1 and T2

Activity T1
 
T2
 
t p Hedge's g 
n M (SDn M (SD
Shopping for food 30 1.57 (1.36) 30 3.17 (1.18) −6.03 <.001** 1.24 
Preparing meals 30 2.47 (1.57) 30 3.27 (1.11) −3.19 .003* 0.58 
Feeding self 30 3.23 (1.43) 30 3.90 (0.31) −2.61 .01* 0.64 
Cleaning up after meals 30 2.67 (1.52) 30 3.27 (1.11) −2.58 .02* 0.44 
Choosing own clothes 30 2.77 (1.52) 30 3.23 (1.04) −2.04 .05* 0.35 
Dressing self 30 3.17 (1.42) 30 3.47 (1.14) −1.43 .16 0.23 
Washing own clothes 30 2.73 (1.59) 30 2.80 (1.56) −0.29 .77 0.04 
Showering/bathing 30 3.07 (1.46) 30 3.53 (1.01) −1.99 .06 0.36 
Brushing teeth 30 3.37 (1.16) 30 3.73 (0.69) −2.48 .02* 0.37 
Washing hair 30 3.10 (1.35) 30 3.57 (0.86) −2.38 .02* 0.41 
Going to the toilet 30 3.47 (1.20) 30 3.90 (0.40) −2.21 .04* 0.47 
Keeping track of finances 30 1.93 (1.48) 30 2.27 (1.46) −1.51 .14 0.23 
Paying own bills 30 1.90 (1.54) 30 2.47 (1.61) −2.21 .04* 0.36 
Managing own finances 30 1.70 (1.56) 30 1.90 (1.49) −1.65 .11 0.13 
Making necessary purchases for self 30 2.40 (1.61) 30 2.83 (1.42) −1.99 .06 0.28 
Cleaning own room 30 2.63 (1.45) 30 3.13 (1.17) −2.19 .04* 0.37 
Helping with household chores 30 2.73 (1.34) 30 3.30 (0.99) −2.54 .02* 0.48 
Doing yard work and repairs 27 2.44 (1.69) 27 2.96 (1.53) −1.86 .08 0.32 
Could be trusted to take care of self 30 2.37 (1.67) 30 2.83 (1.44) −2.14 .04* 0.29 
Could be trusted to live in own dwelling 30 2.20 (1.75) 30 2.50 (1.66) −1.61 .12 0.17 
Activity T1
 
T2
 
t p Hedge's g 
n M (SDn M (SD
Shopping for food 30 1.57 (1.36) 30 3.17 (1.18) −6.03 <.001** 1.24 
Preparing meals 30 2.47 (1.57) 30 3.27 (1.11) −3.19 .003* 0.58 
Feeding self 30 3.23 (1.43) 30 3.90 (0.31) −2.61 .01* 0.64 
Cleaning up after meals 30 2.67 (1.52) 30 3.27 (1.11) −2.58 .02* 0.44 
Choosing own clothes 30 2.77 (1.52) 30 3.23 (1.04) −2.04 .05* 0.35 
Dressing self 30 3.17 (1.42) 30 3.47 (1.14) −1.43 .16 0.23 
Washing own clothes 30 2.73 (1.59) 30 2.80 (1.56) −0.29 .77 0.04 
Showering/bathing 30 3.07 (1.46) 30 3.53 (1.01) −1.99 .06 0.36 
Brushing teeth 30 3.37 (1.16) 30 3.73 (0.69) −2.48 .02* 0.37 
Washing hair 30 3.10 (1.35) 30 3.57 (0.86) −2.38 .02* 0.41 
Going to the toilet 30 3.47 (1.20) 30 3.90 (0.40) −2.21 .04* 0.47 
Keeping track of finances 30 1.93 (1.48) 30 2.27 (1.46) −1.51 .14 0.23 
Paying own bills 30 1.90 (1.54) 30 2.47 (1.61) −2.21 .04* 0.36 
Managing own finances 30 1.70 (1.56) 30 1.90 (1.49) −1.65 .11 0.13 
Making necessary purchases for self 30 2.40 (1.61) 30 2.83 (1.42) −1.99 .06 0.28 
Cleaning own room 30 2.63 (1.45) 30 3.13 (1.17) −2.19 .04* 0.37 
Helping with household chores 30 2.73 (1.34) 30 3.30 (0.99) −2.54 .02* 0.48 
Doing yard work and repairs 27 2.44 (1.69) 27 2.96 (1.53) −1.86 .08 0.32 
Could be trusted to take care of self 30 2.37 (1.67) 30 2.83 (1.44) −2.14 .04* 0.29 
Could be trusted to live in own dwelling 30 2.20 (1.75) 30 2.50 (1.66) −1.61 .12 0.17 

Note: Sample size column (n) represents the number of informants who completed the relevant items.

*p < .05.

**Significant after Bonferroni correction at p < .0025.

Additional Analyses

We explored the possibility that the inclusion of the 11 otherwise-eligible moderate-to-severe TBI participants (described in the Participants section) might have led to a different pattern of results than those reported above (i.e., we investigated whether the study might have suffered from a selection bias). There were, however, no statistically significant differences between the included participants on which we report (N = 31) and the excluded participants (N = 11) on key demographic and test variables, p > .05 for age, years of education, proportion of individuals tested for litigation purposes (i.e., at T1) in English and Afrikaans, proportion of employed versus unemployed individuals at T1, Forced Choice Test recognition score at T1 (included: M = 16.59, SD = 2.47, seven participants “failed”; excluded: M = 16.73, SD = 2.41, three participants “failed”), and Reliable Digit Span score at T1 (included: M = 8.16, SD = 2.05, nine participants “failed”; excluded: M = 8.00, SD = 2.68, three participants “failed”). Hence, it appears that there were no critical sociodemographic differences between included and excluded participants, and that there is little likelihood that only those who malingered at T1 refused reassessment at T2.

Discussion

To explore whether financial settlement influenced neuropsychological test performance, functional status, and activities of daily living capability, we investigated cognitive and behavioral recovery from closed head injury in a group of moderate-to-severe TBI individuals tested both during litigation (at T1) and 1 year or more after case settlement (at T2). We found no evidence that these individuals, despite clear potential for secondary gain, malingered or delivered sub-optimal effort during neuropsychological evaluation taking place in the context of litigation. Similarly, there were few changes in functional status from T1 to T2. Although some improvement was evident, from T1 to T2, in activities of daily living, the gains were small and their clinical significance questionable.

Sample Characteristics

The sample of 31 participants described here was, compared with the South African population, relatively well educated (a mean of approximately 11 years of education). In 2008, only 77% of the South African adult population was functionally literate, with an average of about 6 years of completed formal education (Department of Basic Education, 2010). This characteristic of the current sample may be a consequence of higher levels of education being associated with greater socioeconomic stability, which in turn allows researchers to trace individuals more easily during the recruitment process.

Regarding changes in employment over the course of the study, our observed rates of employment, at both T1 and T2, in non-competitive, unpaid, or sheltered positions are similar to those reported by Hoofien, Gilboa, Vakil, and Donovick (2001) for a non-litigant TBI sample. We also found that there was no significant difference in the number of employed individuals from T1 to T2. These data contrast sharply with those presented by Olver and colleagues (1996), who reported that, in a non-litigant TBI sample, 32% of those working at 2 years post-injury were unemployed at 5 years post-injury. Olver and colleagues attributed their findings, at least partially, to factors not related directly to the TBI. For instance, they noted that failure to sustain employment might have been due to the economic recession in Australia at the time of the study and a subsequent inability of retrenched workers to find new jobs. Furthermore, they stated that, under these conditions of recession, employers might have been less tolerant of patients' lower productivity or interpersonal difficulties. These factors, taken together with patients' likely impaired cognitive ability to learn new skills when attempting to retrain for new jobs, are presumed to account for the precipitous drop in employment numbers reported in that study.

One further point with regard to the employment data in our study is that more professionals/office workers than secondary sector workers were able to return to work. Our data suggest that the reasons for this disparity did not relate to differences in injury severity or to greater levels of cognitive impairment in secondary sector workers. Rather, we speculate that the disparity might be due to (a) job retention ability (an injured secondary sector worker is much more likely to be replaced quickly than a skilled professional/office worker would be; the latter is more likely to be protected by human resources policy, and might be shifted to a different, less demanding position, rather than being laid off completely) and (b) physical demands of the position (secondary sector workers are more likely than professionals/office workers to have physical demands placed on them by their jobs, and so orthopedic injuries, for instance, would be much more likely to have a negative effect on their ability to return to work).

Changes in Neuropsychological Test Performance from T1 to T2

Consistent with previous findings (e.g., McKinlay et al., 1983; Wood & Rutterford, 2006), our data showed that, in a sample of moderate-to-severe TBI litigants, neuropsychological test performance did not change significantly from assessment during litigation to assessment post-litigation. Comparison of test means at T1 and T2 showed very small changes, with RCI analyses indicating that there was no clinically significant change over time on even those tests where statistically significant change was evident. Analyses of data from the effort tests, and of their association with neuropsychological test performance over time, confirmed that (a) only a small proportion of the participants “failed” efforts tests at T1 and (b) there was no significant relationship between effort test performance at T1 and change in neuropsychological test performance from T1 to T2.

This finding of little to no long-term neuropsychological change in a litigant TBI sample is similar to that reported for non-litigating TBI samples. For instance, Millis and colleagues (2001) found, in a sample of complicated mild-to-severe TBI participants who had received acute care and in-patient rehabilitation, statistically significant improvements in performance on tests of attention, working memory, episodic memory, and visuo-spatial functioning; however, from a clinical perspective, the changes were small when considering the mean and median differences, and RCI analyses indicated no clinical significance. Overall, then, the similarity between long-term neuropsychological outcome in litigant and non-litigant samples argues against the presence in the former of malingering in the forensic context.

Changes in Self- and Informant-Reported Behavior, Cognition, and ADL Capability from T1 to T2

The a priori prediction that cognitive, behavioral, and ADL functioning, as well as level of physical dependency, would remain stable from T1 to T2 was only partially confirmed. Data from the HI-FI indicated, for instance, that both litigants and informants reported a decline in behavioral symptom presence from T1 to T2. Consistent with those reports, informants noted that patients' ADL capability improved from T1 to T2 in specific domains (e.g., shopping, self-care, and hygiene). Whether these changes are clinically significant (i.e., whether the brain-injured individual actually experienced an improved quality of life at T2) is a matter for conjecture. It should be noted, however, that informants did not report a significant decline in the severity of behavioral problems in the participants from T1 to T2, and that all of participants who needed supervision with regard to ADLs at T1 continued to require such supervision at T2.

With regard to individual symptoms, those most frequently endorsed, at both T1 and T2, tended to be in the cognitive domain (e.g., memory difficulties, slowness, word-finding difficulty, distractibility, fatigue, and planning difficulties). The most significant affective/behavioral problems tended to be headaches, fatigue, and irritability (with the informant, but not the litigant, rating lack of control of emotions and restlessness highly), and physical dependency. Although there are no previously published reports of long-term functional outcome in moderate-to-severe TBI litigants, the current findings are consistent with non-litigant data reported by Olver and colleagues (1996), who showed that, 5 years post-TBI, fatigue was the most commonly reported symptom, followed by memory difficulties, irritability, slowed thinking, and planning difficulties. Similarly, Dikmen, Machamer, Powell, and Temkin (2003) and Deguise and colleagues (2008) showed that, although limitations were present in many functional domains, moderate-to-severe TBI participants reported their major struggles involved the cognitive requirements of everyday tasks.

Regarding symptom severity, the litigants, but not the informants, reported less severe symptoms in the spheres of cognition and physical dependency at T2 than at T1. This lack of congruence between patient and informant reports may indicate either real improvement in these domains of functioning from T1 to T2, or may reflect a lack of insight by the injured individual into his/her capabilities. Given that lack of insight in TBI has been widely documented in the literature (Prigatano, Altman, & O'Brien, 1990; Stuss, Picton, & Alexander, 2001), the latter explanation appears likely.

Regarding informant-reported capability with regard to activities of daily living, the data suggested that shopping for food was the area in which most robust improvement was apparent, although there was a tendency toward general improvement in primary self-care, particularly with regard to nutrition and personal hygiene. Although ADL is a complex construct that is not measured uniformly in the literature and that is therefore difficult to compare across studies, the current data are consistent with the findings from outcome studies in non-litigant samples (again, there are no published papers reporting on long-term ADL outcome in TBI litigants). For instance, Olver and colleagues (1996) reported that a significant number of patients in their study improved, from year 1 to year 5 post-TBI, their ability to accomplish activities of daily living independently. Although they did not follow participants longitudinally, Colantonio and colleagues (2004) reported that, at a mean of 14 years post-injury, 88% of their participants could bathe, dress, eat, transfer, and use the toilet and telephone independently.

A point of interest here is the fact that, in the current study, there were reports of improved ADL functioning but no concurrent comprehensive improvement in neuropsychological functioning. One might speculate that improvement in ADL in the absence of neuropsychological improvement might be due to increasing self-awareness, developing a daily routine, or acquiring better coping skills with time.

The current data do not imply, however, that all ADLs could be performed without supervision at T2: As noted earlier, informants reported that all of the participants who needed ADL supervision at T1 continued to require such supervision at T2. In this regard, the current data are consistent with previous studies reporting little global change in independence from year 1 to year 5 post-injury, and similar rates of disability at year 1 and years 5–7 post-injury (Hammond, Hart, et al., 2004; Whitnall, McMillan, Murray, & Teasdale, 2006).

In litigation cases, capability with regard to financial management is, of course, an important outcome consideration for at least two reasons. First, the court has to decide whether the TBI individual is fit to manage his/her own finances; second, there are frequently later requests for removing a curator bonis (e.g., if an individual with TBI was found unfit to manage their money and a curator bonis was appointed initially, that individual may, at a later date, request to have that order lifted in an attempt to regain control of financial management). We found that there was little informant-reported change from T1 to T2 in the ability of patients to manage their own finances (i.e., there were continuing and numerous areas of difficulty with regard to this ability). This finding is consistent with those of Dikmen and colleagues (2003) and Mazaux and colleagues (1997). The latter reported, specifically, that patients showed the greatest impairment in financial management and administrative tasks; similarly, Colantonio and colleagues (2004) reported that financial management was a limitation over and above other ADLs.

Limitations

The data we present here must be interpreted with appropriate caution given the following methodological limitations of our study. First, it is possible that certain information about demographic and health factors at T1 is less reliable than other information gathered about the participants' status at T1. Specifically, the Interview for the Person with the Head Injury Problem Checklist and the Significant Other Interview Problem Checklist, as well as informant-reports on ADL at T1, were all administered retrospectively. Hence, the data derived from these T2 measures might be biased by (a) errors inherent in retrospective recall, (b) lack of reliability of memory functioning (especially in patients with residual neuropsychological deficits), and (c) the way in which these questionnaires were used in the current study (i.e., asking participants and informants to recall information about function at T1 and to then rate the same domains at T2). The latter, in particular, might have inflated the correspondence between ratings at the two measurement points. Furthermore, self-report instruments such as the HI-FI do not contain built-in validity measures. This weakness in their design might present a problem when the instrument is used as the sole measure of behavioral functioning or ADL capability, as it was here. Overall, then, we might be observing a confounded picture of the patient's true T1 status.

Second, our sample was limited to persons who were fluent English or Afrikaans speakers. This inclusion criterion led to the exclusion of a large section of the South African population (there are 11 official languages in the country, and, for instance, 24% of the population of the Western Cape has isiXhosa as a home language; Statistics South Africa, 2001). Similarly, the sample size we used was relatively small; so, the data we report here require replication.

Third, some participants' and informants' relatively low levels and quality of education led to numerous difficulties with the administration of the research questionnaires. We attempted to overcome these difficulties by, for example, using visual aids to explain how to use a Likert-type scale, and by using graphical rather than numerical scales. It is unknown to what extent these modifications might have influenced the validity of the collected data.

Fourth, the composition of our neuropsychological test battery was dictated not by the demands of the research but by the tests used in the initial forensic evaluation. Certainly, our conclusions must be tempered by, for instance, the fact that the same version of the ROCF was not administered consistently at T1, and by the fact that, from a psychometric perspective, the FCT is not the ideal test of effort.

Finally, studies of this kind are frequently susceptible to self-selection or volunteer biases. For instance, the design of the study cannot rule out the possibility that the sample featured individuals who somaticized or adopted a sick role at both occasions of testing.

Summary and Conclusions

The present findings are consistent with those presented by Wood and Rutterford (2006) and McKinlay and colleagues (1983) in showing that claimants with moderate-to-severe TBI do not appear to fake low scores, or to give sub-optimal effort, during neuropsychological assessment in the context of litigation. The methodology of the current study was different from those previous studies, however, in that it was a within-subjects design that featured administration of an identical neuropsychological test battery at litigation and post-litigation assessment. Additionally, to our knowledge, this is the first study to provide data on long-term neuropsychological outcome of unrehabilitated TBI individuals in a low- to middle-income country.

Regarding changes in functional and adaptive behavior from T1 to T2, the current data suggest that, overall, TBI individuals continued to suffer from significant difficulties in all spheres of their lives (including, importantly, financial management skills) even after case settlement. There were, however, informant reports of some gains in ability to complete certain activities of daily living successfully. Furthermore, both patient and informant reports suggested a decrease in the presence, but not severity, of behavioral symptoms from T1 to T2. However, participant self-report of less symptom severity regarding cognition and physical dependency was not supported by informant reports. The lack of agreement between participant and informant reports in this regard may stem from unreliable information provided by TBI individuals who might have lacked insight.

Despite the limitations mentioned above, the current study provides useful information for neuropsychologists working with brain-injured victims of MVAs and similar events, and for public health policymakers concerned with the potential for malingering by such injured parties. Furthermore, the good correspondence between the current data and those presented in studies from resource-wealthy and high-income countries is encouraging for forensic neuropsychologists in low- and middle-income countries concerned about the applicability of those studies to their practice.

Funding

This work was supported by the Deutscher Akademischer Austausch Dienst and the National Research Foundation of South Africa.

Conflict of Interest

None declared.

Acknowledgements

We thank Frances Hemp and Mignon Coetzee for providing access to patient files, and we thank Pedro Wolf and Michelle Henry for assistance with data analyses. We also acknowledge the comments of three anonymous reviewers on the original version of this manuscript.

References

Benton
A. L.
Hamsher
K.
Multilingual aphasia examination
 , 
1976
Iowa City, IA
University of Iowa
Bercaw
E. L.
Hanks
R. A.
Millis
S. R.
Gola
T. J.
Changes in neuropsychological performance after traumatic brain injury from inpatient rehabilitation to 1-year follow-up in predicting 2-year functional outcomes
The Clinical Neuropsychologist
 , 
2011
, vol. 
25
 (pg. 
72
-
89
)
Bianchini
K. J.
Greve
K. W.
Love
J. M.
Definite malingered neurocognitive dysfunction in moderate/severe traumatic brain injury
The Clinical Neuropsychologist
 , 
2003
, vol. 
17
 (pg. 
574
-
580
)
Binder
L. M.
Rohling
M. L.
A meta-analytic review of the effects of financial incentives on recovery after closed-head injury
American Journal of Psychiatry
 , 
1996
, vol. 
153
 (pg. 
7
-
10
)
Boone
K. B.
Lu
P.
Noncredible cognitive performance in the context of severe brain injury
The Clinical Neuropsychologist
 , 
2003
, vol. 
17
 (pg. 
244
-
254
)
Cohen
J.
Statistical power analysis for the behavioral sciences
 , 
1988
2nd ed.
Hillsdale, NJ
Erlbaum
Colantonio
A.
Ratcliff
G.
Chase
S.
Kelsey
S.
Escobar
M.
Vernich
L.
Long-term outcomes after moderate to severe traumatic brain injury
Disability & Rehabilitation
 , 
2004
, vol. 
26
 (pg. 
253
-
261
)
Deguise
E.
LeBlanc
J.
Feyz
M.
Meyer
K.
Duplantie
J.
Thomas
H.
, et al.  . 
Long-term outcome after severe traumatic brain injury: The McGill Interdisciplinary Prospective Study
Journal of Head Trauma Rehabilitation
 , 
2008
, vol. 
23
 (pg. 
294
-
303
)
Department of Basic Education
2009 country report: South Africa
 , 
2010
Pretoria, South Africa
Author
Dikmen
S. S.
Machamer
J. E.
Powell
J. M.
Temkin
N. R.
Outcome 3 to 5 years after moderate to severe traumatic brain injury
Archives of Physical Medicine and Rehabilitation
 , 
2003
, vol. 
84
 (pg. 
1449
-
1457
)
Evans
C.
1998
 
Reliable change criterion calculator. Retrieved December 7, 2007, from http://www.psyctc.org/stats/rcsc1.htm
Evans
R. W.
The effects of litigation on treatment outcome with personal injury patients
American Journal of Forensic Psychology
 , 
1994
, vol. 
4
 (pg. 
19
-
34
)
Fernstein
A.
Ouchterlony
D.
Somerville
J.
Jardine
A.
The effects of litigation on symptom expression: A prospective study following mild traumatic brain injury
Medicine, Science and Law
 , 
2001
, vol. 
41
 (pg. 
116
-
121
)
Foxcroft
C. D.
Paterson
H.
Le Roux
N.
Herbst
D.
Psychological assessment in South Africa: A needs analysis: The test use patterns and needs of psychological assessment practitioners
 , 
2004
Pretoria, South Africa
Human Sciences Research Council
Goosen
J.
Bowley
D. M.
Degiannis
E.
Plani
F.
Trauma care systems in South Africa
Injury: International Journal of Care of the Injured
 , 
2003
, vol. 
34
 (pg. 
704
-
708
)
Greiffenstein
M. F.
Baker
W. J.
Gola
T.
Validation of malingered amnesic measures with a large clinical sample
Psychological Assessment
 , 
1994
, vol. 
6
 (pg. 
218
-
224
)
Hammond
F. M.
Grattan
K. D.
Sasser
H.
Corrigan
J. D.
Rosenthal
M.
Bushnik
T.
, et al.  . 
Five years after traumatic brain injury: A study of individual outcomes and predictors of change in function
Neuro Rehabilitation
 , 
2004
, vol. 
19
 (pg. 
25
-
35
)
Hammond
F. M.
Hart
T.
Bushnik
T.
Corrigan
J. D.
Sasser
H.
Change and predictors of change in communication, cognition, and social function between 1 and 5 years after traumatic brain injury
Journal of Head Trauma Rehabilitation
 , 
2004
, vol. 
19
 (pg. 
314
-
328
)
Hofman
A. M.
Stapert
S. Z.
van Kroonenburgh
M. J. P. G.
Jolles
J.
de Kruijk
J.
Wilmink
J. T.
MR Imaging, single-photon emission CT, and neurocognitive performance after mild traumatic brain injury
American Journal of Neuroradiology
 , 
2001
, vol. 
22
 (pg. 
441
-
449
)
Hoofien
D.
Gilboa
A.
Vakil
E.
Donovick
P. J.
Traumatic brain injury (TBI) 10–20 years later: A comprehensive outcome study of psychiatric symptomatology, cognitive abilities and psychosocial functioning
Brain Injury
 , 
2001
, vol. 
15
 (pg. 
189
-
209
)
Hugo
F. J.
Thornton
H. B.
Emsley
R. A.
Theron
F. H.
de Villiers
N.
Hemp
F.
The prevalence of malingering in persons seeking compensation on the basis of neuropsychiatric symptoms
South African Journal of Psychiatry
 , 
2001
, vol. 
7
 (pg. 
110
-
114
)
Kay
T.
Cavallo
M. M.
Ezrachi
O.
Vavagiakis
P.
The Head Injury Family Interview: A clinical research tool
Journal of Head Trauma Rehabilitation
 , 
1995
, vol. 
10
 (pg. 
12
-
31
)
Lafayette Instrument Company
The Purdue Pegboard Test: Examiner's manual
 , 
1999
Lafayette, IN
Author
Levin
K.
Paediatric traumatic brain injury in South Africa: some thoughts and considerations
Disability & Rehabilitation
 , 
2004
, vol. 
26
 (pg. 
306
-
314
)
Lezak
M. D.
Howieson
D. B.
Loring
D.W.
Neuropsychological assessment
 , 
2004
4th ed.
New York, NY
Oxford University Press
Mazaux
J. M.
Masson
F.
Levin
H. S.
Alaoui
P.
Maurette
P.
Barat
M.
Long-term neuropsychological outcome and loss of social autonomy after traumatic brain injury
Archives of Physical Medicine and Rehabilitation
 , 
1997
, vol. 
78
 (pg. 
1316
-
1320
)
McKinlay
W. M.
Brooks
D. N.
Bond
M. R.
Post-concussional symptoms, financial compensation and outcome of severe blunt head injury
Journal of Neurology, Neurosurgery & Psychiatry
 , 
1983
, vol. 
46
 (pg. 
1054
-
1091
)
Meyers
J. E.
Meyers
K. R.
Rey Complex Figure Test under four different administration procedures
The Clinical Neuropsychologist
 , 
1995
, vol. 
8
 (pg. 
153
-
166
)
Miller
L. J.
Donders
J.
Subjective symptomatology after traumatic head injury
Brain Injury
 , 
2001
, vol. 
15
 (pg. 
297
-
304
)
Millis
S. R.
Rosenthal
M.
Novak
T. A.
Sherer
M.
Nick
T. G.
Kreutzer
J. S.
, et al.  . 
Long-term neuropsychological outcome after traumatic brain injury
Journal of Head Trauma Rehabilitation
 , 
2001
, vol. 
16
 (pg. 
343
-
355
)
Mittenberg
W.
Patton
C.
Canyock
E. M.
Condit
D. C.
Base rates of malingering and symptom exaggeration
Journal of Clinical and Experimental Neuropsychology
 , 
2002
, vol. 
24
 (pg. 
1094
-
1102
)
Oddy
M.
Coughlan
A.
Tyerman
A.
Jenkins
D.
Social adjustment after closed head injury: A further follow-up seven years after injury
Journal of Neurology, Neurosurgery & Psychiatry
 , 
1985
, vol. 
48
 (pg. 
564
-
568
)
Olver
J. H.
Ponsford
J.
Curran
C.
Outcome following traumatic brain injury: A comparison between 2 and 5 years after injury
Brain Injury
 , 
1996
, vol. 
10
 (pg. 
841
-
848
)
Paniak
C.
Reynolds
S.
Toller-Lobe
G.
Melnyk
A.
Nagy
J.
Schmidt
D.
A longitudinal study of the relationship between financial compensation and symptoms after treated mild traumatic brain injury
Journal of Clinical and Experimental Neuropsychology
 , 
2002
, vol. 
242
 (pg. 
187
-
193
)
Paterson
H.
Uys
J.
Critical issues in psychological test use in the South African workplace
South African Journal of Industrial Psychology
 , 
2005
, vol. 
31
 (pg. 
12
-
22
)
Prigatano
G. P.
Altman
I.
O'Brien
K.
Behavioral limitations that traumatic brain injured patients tend to underestimate
The Clinical Neuropsychologist
 , 
1990
, vol. 
4
 (pg. 
1
-
14
)
Reitan
R. M.
The relation of the Trail Making Test to organic brain damage
Journal of Consulting Psychology
 , 
1955
, vol. 
19
 (pg. 
393
-
394
)
Reynolds
S.
Paniak
C.
Toller-Lobe
G.
Nagy
J.
A longitudinal study of compensation-seeking and return to work in a treated mild traumatic brain injury sample
Journal of Head Trauma Rehabilitation
 , 
2003
, vol. 
18
 (pg. 
139
-
147
)
Rogers
R.
Clinical assessment of malingering and deception
 , 
1997
2nd ed.
London, England
Guilford
Shuttleworth-Jordan
A. B.
On not reinventing the wheel: A clinical perspective on culturally relevant test usage in South Africa
South African Journal of Psychology
 , 
1996
, vol. 
26
 (pg. 
96
-
102
)
Statistics South Africa
2001
 
Census 2001. Retrieved January 5, 2012, from http://www.statssa.gov.za/census01/html/CInbrief/CIB2001.pdf
Strauss
E.
Sherman
E. M. S.
Spreen
O.
A compendium of neuropsychological tests: Administration, norms, and commentary
 , 
2006
3rd ed.
New York, NY
Oxford University Press
Stuss
D.
Picton
T.
Alexander
M.
Salloway
S.
Malloy
P.
Duffy
J.
Consciousness, self-awareness and the frontal lobes
The frontal lobes and neuropsychiatric illness
 , 
2001
Washington, DC
American Psychiatric Press
(pg. 
101
-
112
)
Suhr
J.
Tranel
D.
Wefel
J.
Barrash
J.
Memory performance after head injury: Contributions of malingering, litigation status, psychological factors, and medication use
Journal of Clinical and Experimental Neuropsychology
 , 
1997
, vol. 
19
 (pg. 
500
-
514
)
Teasdale
G.
Jennett
B.
Assessment of coma and impaired consciousness. A practical scale
The Lancet
 , 
1974
, vol. 
304
 (pg. 
81
-
84
)
Theron
F. H.
Spangenberg
J. J.
Hugo
F. J.
Emsley
R. A.
Hemp
F.
Maritz
J. S.
Validation of psychometric scales for malingering in a student sample
South African Journal of Psychiatry
 , 
2001
, vol. 
7
 (pg. 
96
-
102
)
Wechsler
D.
Wechsler Adult Intelligence Scale – Third Revision (WAIS-III)
 , 
1997
San Antonio, TX
The Psychological Corporation
Whitnall
L.
McMillan
T. M.
Murray
G. D.
Teasdale
G. M.
Disability in young people and adults after head injury: 5–7 year follow up of a prospective cohort study
Journal of Neurology, Neurosurgery & Psychiatry
 , 
2006
, vol. 
77
 (pg. 
640
-
645
)
Wood
R. L.
Long-term outcome of serious traumatic brain injury
European Journal of Anaesthesiology
 , 
2008
, vol. 
25
 
Suppl. 42
(pg. 
115
-
122
)
Wood
R. L.
Rutterford
N. A.
The effect of litigation on long-term cognitive and psychosocial outcome after severe brain injury
Archives of Clinical Neuropsychology
 , 
2006
, vol. 
21
 (pg. 
239
-
246
)
World Bank
2009
 
World development indicators database. Retrieved September 28, 2010, from http://siteresources.worldbank.org/DATASTATISTICS/Resources/GNIPC.pdf

Author notes

Present address of Hetta Gouse is now at the Department of Psychiatry, University of Cape Town.