COMPENSATION AND MALINGERING IN TRAUMATIC BRAIN INJURY: A DOSE-RESPONSE RELATIONSHIP?

Transcription

1 The Clinical Neuropsychologist, 20: , 2006 Copyright # Taylor and Francis Group, LLC ISSN: print= online DOI: / COMPENSATION AND MALINGERING IN TRAUMATIC BRAIN INJURY: A DOSE-RESPONSE RELATIONSHIP? Kevin J. Bianchini, Kelly L. Curtis, and Kevin W. Greve University of New Orleans, New Orleans, LA, USA and Jefferson Neurobehavioral Group, Metairie, LA, USA The purpose of this study was to determine if there is a dose-response relationship between potential monetary compensation and failure on psychological indicators of malingering in traumatic brain injury. 332 traumatic brain injury patients were divided into three groups based on incentive to perform poorly on neuropsychological testing: no incentive; limited incentive as provided by State law; high incentive as provided by Federal law. The rate of failure on five well-validated malingering indicators across these groups was examined. Cases handled under Federal workers compensation laws showed considerably higher rates of failure and diagnosable malingering than cases handled under State law. The findings indicate that monetary compensation associated with workers compensation claims is a major motive for exaggeration and malingering of problems attributed to work-related brain injuries. The clinician s index of suspicion regarding exaggeration and malingering of symptoms and deficits should be much higher in the context of Federal workers compensation claims, particularly in patients who have suffered only mild traumatic brain injury. Many Americans will suffer a traumatic brain injury (TBI) some time in their lives that ultimately impacts their ability to function in daily activities of life to a varying extent (Bazarian et al., 1999). Four hundred forty-four in one hundred thousand Americans will be treated for brain injury in an Emergency Room (Jager, 2000). The vast majority, roughly 80%, fall into the mild category and have a Glasgow Coma Scale of 15 (Kraus & Nourjah, 1988). Between 85 and 90% of persons who have sustained a MTBI are able to return to normal functioning after a relatively short amount of time post-injury, usually within six months of the injury (Alexander, 1995; Dikmen, Machamer, Winn, & Temkin, 1995; Dikmen, Temkin, Machamer, Holubkov, Fraser, & Winn 1994; Youngjohn, Burrows, & Erdal, 1995). The remaining 10 to 15% of MTBI patients report symptoms that persist beyond the abovereported recovery period (Alexander, 1995; Binder, 1997). It is well-established that financial incentive is related to outcome in brain injury, in general, and the effect of money is strongest at the mild end of the severity continuum (Binder, 1986; Binder & Rohling, 1996; Paniak et al., 2002; Price & Stevens, 1997; Reynolds, Paniak, Toller-Lobe, & Nagy, 2003; Youngjohn et al., 1995). Address correspondence to: Kevin W. Greve, Ph.D., Department of Psychology, University of New Orleans-Lakefront, New Orleans, LA Tel.: Fax: [email protected] Accepted for publication: March 23,

2 832 KEVIN J. BIANCHINI ET AL. The specific influence of financial compensation on severity of symptoms was assessed in a meta-analysis conducted by Binder and Rohling (1996) who found that the strength of the association between outcome measures including neuropsychological test results and financial incentive (d) was of 0.47 overall (.89 for mild TBI) but was not significant (generally less than.10) for injury severity (Binder & Rohling, 1996; Binder, Rohling, & Larrabee, 1997). It was shown that an incentive by injury severity interaction existed, with MTBI patients with incentive performing worse on neuropsychological measures than individuals with more severe injuries or without incentive. Hill (1965) posited nine criteria for judging whether an association was causal. Among these are the idea that a strong association is more likely to reflect a causal relationship than a weak one. A second criterion is evidence of consistency in that the effect occurs across studies and populations. Binder and Rohling (1996) demonstrated that the incentive effect is both strong and consistent. Moreover, subsequent research has supported their findings (Bernstein, 1999; Boone & Lu, 2003; Cato, Brewster, Ryan, & Giuliano, 2002; Iverson & Binder, 2000; Reynolds et al., 2003; Youngjohn et al., 1995). Carroll et al. (2004) published a comprehensive review of mild TBI, concluding that where symptoms persist, compensation=litigation is a factor (p. 84). Thus, this strong and consistent incentive effect indicates that some aspect of the medicolegal context leads to exaggeration of deficit and raises the possibility that some patients with persistent symptoms are motivated by potential financial compensation to intentionally exaggerate their cognitive deficits (i.e., malinger). Mittenberg and colleagues (Mittenberg, Patton, Canyock, & Condit, 2002), using survey methodology, have estimated that the baserate of malingering in TBI is between 30 and 40%. Larrabee (2005) reviewed a number of studies of performance on effort tests and found a similar rate using a very different methodological approach. Thus, a higher percentage of persons with incentive, compared to persons without incentive, perform poorly on neuropsychological tests and have test findings consistent with malingering. This would suggest that the presence of incentive is a factor motivating poor performance. However, this methodology does not disentangle the influence of financial compensation and the influence of processes inherent in the medico-legal context (e.g., litigation itself, delay in receiving treatment, unemployment, etc.). One way of disentangling the effects of these two inter-related factors is to establish what is effectively a dose-response relationship between the magnitude of potential compensation and failure on tests sensitive to malingering. The dose-response relationship is another of Hill s criteria for causality (Hill, 1965). A fundamental concept in pharmacology and toxicology (Eaton & Klaassen, 2001), the dose-response relationship is also important in the study of the behavioral effects of traumatic brain injury (TBI). In TBI, clear population dose-response relationships have been observed between acute indicators of neurological dysfunction (e.g., length of loss of consciousness, time to respond to commands) and neuropsychological test performance and work return (Dikmen et al., 1994; Dikmen et al., 1995; Rohling, Meyers, & Millis, 2003). Similarly, Green and colleagues (Green, Rohling, Iverson, & Gervais, 2003) have demonstrated a dose-response relationship between severity of brain injury and olfactory dysfunction, but only in those patients demonstrating valid effort. The apparent absence of similar doseresponse relationships in individuals (as opposed to populations) has been used to

3 COMPENSATION AND RESPONSE BIAS 833 question the source of cognitive deficits in some cases of electrical injury (Bianchini, Love, Greve, & Adams, 2005). Thus, if potential external incentive were indeed motivating poor performance in some TBI patients, it would be hypothesized that greater potential incentive should result in a higher frequency of exaggerated test results and possibly in the report of a greater variety of subjective symptoms (e.g., adding emotional and=or physical symptoms to cognitive complaints after TBI). Specifically, Bianchini and colleagues (Bianchini, Greve, & Love, 2003; Bianchini, Etherton, & Greve, 2004; Bianchini, Greve, & Glynn, 2005) have argued that the potential for greater incentive may result in a greater variety of exaggerated symptoms because a broader symptom presentation would produce greater disability which is more valuable in the medicolegal context. So, how does one relate magnitude of potential incentive to malingering test performance? Ideally, one would correlate the value of the claimed loss with test performance; unfortunately, the detailed analysis of settlement demands in the large numbers of cases necessary for such an analysis is impractical. On the other hand, the structure of workers compensation law may provide a more practical method of examining the relationship between incentive and bias in litigated TBI. Specifically, work-related injuries come under the legal control (jurisdiction) of different laws as a function of where and under what circumstances the injury occurred. In Louisiana, the medical management of and potential financial compensation from the employer for a work-related injury are governed by one of three different jurisdictions. The Longshore and Harbor Worker s Compensation Act (LHWCA; 1927) and Jones Act (JA; 2002), both federal statutes, cover workers who suffer personal injury in the course and scope of their employment. The Louisiana Workers Compensation Law covers work injuries sustained by employees not engaged in so-called maritime employment (the latter being covered by the LHWCA unless the worker is a seaman who is covered by the Jones Act). Injured employees covered by this law are typically limited to collecting weekly compensation benefits from their employers (two thirds of lost wages subject to a maximum rate) plus medical expenses and in some circumstances, vocational rehabilitation. Moreover, the system was devised as a compromise whereby the employee gives up his rights to sue for damages in tort (generally referred to as a liability or tort suit) in exchange for the no-fault, but limited recovery provided by the workers compensation law. Under these regulations, employers are given immunity from lawsuits for negligence and strict liability, but remain liable (in addition to their obligation for lost wages and medical expenses) only for intentional torts. The two Federal remedies against employers for work-related injuries provided by the Jones Act and LHWCA provide for higher awards, although by different mechanisms, when compared to the state workers compensation remedy. Individuals qualifying for Jones Act, known as seaman, are entitled to file suits against their employer not only for maintenance and cure (which is similar but not the same as weekly benefits and medical care provided by the workers compensation law), but also for other damages, including pain and suffering and lost income=earning capacity (Jones Act, 2002). Furthermore, this law is one of the most protective laws in the United States and as such, can hold the employer liable for negligence to the seaman, traditionally considered to be a ward of the court, under a significantly

4 834 KEVIN J. BIANCHINI ET AL. reduced burden of proof relative to that imposed on claimants by general liability law, hence the somewhat inaccurate but common description of slight negligence. In other words, there is a lowered threshold for recovery in claims of negligence for injuries sustained in Jones Act cases which allow individuals increased opportunities for potentially greater rewards. Like state workers compensation law, the Longshore Act has protections in favor of the employer (i.e., exclusive remedy) regarding negligence, however, the limits for the weekly compensation rate are higher, thus leading to generally higher overall recovery when compared to the state workers compensation remedy. Thus, differences in these jurisdictions provide an opportunity to examine the dose-response relationship between the magnitude of potential compensation and performance on indicators of malingering in traumatic brain injury. Specifically, incentive was categorized into three levels: no incentive; limited incentive as provided by Louisiana workers compensation law; and, high incentive as provided by Federal law. The effect of compensation was examined in the overall TBI sample and in mild and moderate-severe TBI separately. The indicators of malingering used here have been well-validated in TBI and the selected cutoffs have been demonstrated to be very specific to malingering. This is the first article to examine the unique effects of different levels of incentive, beyond the simple presence or absence of incentive, and performance on a variety of tests and indicators empirically-determined to be sensitive to malingering. METHODS Participants Participants were 332 consecutive traumatic brain injury referrals seen for neuropsychological evaluation in a neuropsychology group practice in the Southeastern United States between March 1994 and May Classification of brain injury severity was based on the criteria for mild TBI of the Mild Traumatic Brain Injury Committee of the Head Injury Interdisciplinary Special Interest Group for the American Congress of Rehabilitation Medicine (1993). Thus, patients were considered to have suffered a mild traumatic brain injury (TBI) if they met the following Table 1 Descriptive statistics for each incentive group Incentive status No incentive Total WC State LS=JA Sample size Mild Mod-Severe Total Age Mean sd Education Mean sd Gender % female

5 COMPENSATION AND RESPONSE BIAS 835 Table 2 Descriptive statistics for injury characteristics by incentive group Incentive status No incentive mean (sd) Total WC mean (sd) State mean (sd) LS=JA mean (sd) Injury to evaluation (months) Mild (41.2) (19.6) (19.0) (21.7) Mod-Severe (24.9) (28.1) (28.3) (28.2) Total (29.6) (23.0) (22.8) (24.1) Glasgow coma scale Mild (0.6) (0.4) (0.5) (0.3) Mod-Severe 7.83 (4.0) (4.2) (4.2) 9.30 (4.3) Total 9.88 (4.7) (3.2) (3.1) (3.4) Focal signs (% positive) Mild Mod-Severe Total Skull fracture (% positive) Mild Mod-Severe Total Neuroimaging (% positive) Mild Mod-Severe Total Note. All skull fractures in mild TBI patients (n ¼ 4) were nondisplaced skull fractures. criteria: an initial Glascow Coma Score of at least 13, loss of consciousness of 30 minutes or less, post-traumatic amnesia not greater than 24 hours. Individuals not meeting all three of these criteria were classified as having sustained a moderate-severe traumatic brain injury. Thus, the mild TBI group contained only persons who clearly had no worse than a mild TBI and might have had no TBI. In total, 178 were classified as mild and 154 were classified as moderate-severe. Table 1 provides sample sizes and the demographic characteristics of these patients while Table 2 presents the characteristics of their injuries. Incentive Cases were divided into groups based on their incentive status and the jurisdiction of their workers compensation claim (No Incentive [n ¼ 94], State [n ¼ 184], Longshore=Jones Act [n ¼ 54]). Due to the limited number of cases associated with both Longshore Act (1927) and Jones Act (2002), the cases in these two jurisdictions were combined into a single group (LS=JA; n ¼ 54). Patients seen in the context of non-work personal injury claims were excluded from this study. Malingering Indicators Clinical examination included administration of a comprehensive neuropsychological battery and review of medical and other injury related information.

6 836 KEVIN J. BIANCHINI ET AL. Table 3 Sample sizes by validity indicator, injury severity, and incentive category Incentive status Injury severity No incentive Total WC State LS=JA Portland digit recognition test (N ¼ 241) Mild Mod-Severe Total Test of memory malingering (N ¼ 116) Mild Mod-Severe Total Reliable digit span (N ¼ 330) Mild Mod-Severe Total MMPI-2 fake bad scale (N ¼ 231) Mild Mod-Severe Total MMPI-2 infrequency (F) scale (N ¼ 231) Mild Mod-Severe Total Because of changes in the battery over time and specific demands of some cases, not all patients completed all tests. The following malingering indicators were utilized to assess exaggeration of cognitive deficits and exaggeration of psychological and physical complaints. See Table 3 for a breakdown of the samples sizes according to incentive category and injury severity. Portland Digit Recognition Test (Binder, 1993; PDRT). The PDRT is a 72 item SVT employing recognition of five-digit number strings. The PDRT uses an active, counting distractor period between stimulus presentation and recognition trial PDRT (Denney, 1996). The test appears to increase in difficulty as distractor periods grow from 5 seconds during the first 18 items to 15 seconds for the second quartile and 30 seconds for the final half of the 36 trials. Despite its appearance, the PDRT actually has little real memory requirement. A positive PDRT finding was defined as the presence of one or more scores below the recommended cut-off s (19 out of 36 of the easier half, 18 out of 36 for the difficult items, 39 out of 72 total items). These cut-offs are associated with 100% specificity in TBI (Bianchini, Mathias, Greve, Houston, & Crouch, 2001; Binder & Kelly, 1996). Test of Memory Malingering (Tombaugh, 1996; TOMM). The TOMM is a two-choice discrimination task using line drawings of common objects. Two learning trials and a retention trial are administered. TOMM performance is minimally affected by numerous factors that impact cognitive test performance including severe brain pathology (Ashendorf, Constantinou, & McCaffrey, 2004; Constantinou &

7 COMPENSATION AND RESPONSE BIAS 837 McCaffrey, 2003; Etherton, Bianchini, Greve, & Ciota, 2005; Rees, Tombaugh, & Boulay, 2001; Teichner & Wagner, 2004; Tombaugh, 1997, 2002). Scores of less than 45 on Trial 2 and=or the Retention trial are considered positive (Tombaugh, 1996). Data for TBI patients (Tombaugh, 1996, 1997; Rees et al., 2001) indicates that specificity is 100% at these cutoffs. Reliable Digit Span (Greiffenstein, Baker, & Gola, 1994; RDS). The RDS score is derived from the Wechsler Adult Intelligence Scale (Wechsler, 1981, 1997a) or Wechsler Memory Scale (Wechsler, 1987, 1997b) Digit Span subtest. For forward and backward span the patient must repeat increasingly longer strings of digits with two strings presented at each span length. In backward span, the participant must repeat the strings in reverse order. Forward and backward span are discontinued after failure on both strings of a given length. The RDS score is determined by summing the longest forward and backward digit strings for which both trials were completed without error. Like the TOMM, RDS has been validated as an effort measure in a range of conditions (Etherton, Bianchini, Ciota, & Greve, 2005; Etherton, Bianchini, Greve, & Heinly, 2005; Greiffenstein et al., 1994; Larrabee, 2003a; Mathias, Greve, Bianchini, Houston, & Crouch, 2002; Meyers & Volbrecht, 1998). An RDS score of less than seven is associated with specificity of 99% or higher in TBI, and was considered positive. Fake Bad Scale (Lees-Haley, English, & Glenn, 1991; FBS). FBS was derived from the Minnesota Multiphasic Personality Inventory, 2nd edition (Butcher, Dahlstrom, Graham, Tellegren, & Kaemmer, 1989) (MMPI-2) and is sensitive to exaggeration of complaints associated with physical injury as opposed to psychopathology (see F scale below). FBS has been studied extensively and there are published Specificity and Sensitivity data on the FBS that have been derived from methodologically sound studies and, in general, scores greater than 29 are associated with 100% specificity (Greiffenstein, Baker, Gola, Donders, & Miller, 2002; Greve, Bianchini, Love, Brennan, & Heinly, 2006; Iverson, Henrichs, Barton, & Allen, 2002; Larrabee, 2003b; Meyers, Millis, & Volkert, 2002; Ross, Millis, Krukowski, Putnam, & Adams, 2004). Therefore, this study considered a score of greater than or equal 30 to be a positive finding. Infrequency (F) Scale. The F scale is one of the original MMPI validity scales and is composed of items endorsed in the scored direction by less than 10% of the non-clinical respondents (Graham, 1990). One underlying assumption of the F scale is that items infrequently endorsed by normal persons would be over-endorsed by individuals attempting to exaggerate or feign their psychiatric problems (Berry, Baer, & Harris, 1999). It is sensitive to exaggeration of psychiatric symptoms. The effectiveness of MMPI validity scales in detecting exaggeration and=or malingering of symptoms in a range of conditions has recently been reviewed (Lees-Haley, Iverson, Lange, Fox, & Allen, 2003; Rogers, Sewell, Martin, & Vitacco, 2003). Butcher et al. (1989) suggested that T-scores of 90 or greater reflects exaggeration. Greve et al. (2006) found that a T-score of greater than 90 produced a false positive error rate of 2% in TBI while correctly identifying about 20% of malingerers. Thus, an F scale T-score of greater than 90 was considered positive for this study.

8 838 KEVIN J. BIANCHINI ET AL. Malingering Diagnosis A diagnosis of malingering requires the careful review and integration of a range of information, of which psychometric indicators of bias or effort, like those described above, are an important part. Slick, Sherman and Iverson (1999) have published comprehensive and well thought out criteria for the diagnosis of Malingered Neurocognitive Dysfunction (MND). In determining the presence of MND, the case must be evaluated on the basis of four criteria: a) presence of substantial external incentive; b) evidence from neuropsychological testing; c) evidence from self-report; and, d) behaviors meeting the necessary b and c criteria are not fully accounted for by psychiatric, neurological, or developmental factors. Using this system, all diagnoses of malingering require the presence of external incentive (Criterion a) plus Criterion b and=or c evidence as noted below. In the context of external incentive, Criterion b behaviors are sufficient for a diagnosis of malingering on their own and a finding of definite response bias alone is sufficient for a diagnosis of Definite MND. A diagnosis of Probable MND can be made with two types of Criterion b evidence or one type of Criterion b evidence and one or more types of Criterion c evidence. Criterion c evidence is not sufficient for a diagnosis in the absence of Criterion b evidence. Possible MND is diagnosed when the criteria for Probable MND have been met but Criterion d factors are present. A diagnosis of MND can be derived from the psychometric data analyzed in this study. For purposes of this study, patients were diagnosed with MND in one of two ways. First, Definite MND was considered present when a patient with incentive had a statistically-significantly below-chance performance on either the PDRT or TOMM, thus meeting Criterion b1. Second, Probable MND was considered present if a patient with incentive was positive on any of the three cognitive malingering indicators (PDRT, TOMM, RDS) thereby meeting Criterion b2 and were simultaneously positive on either of the MMPI-2 validity indicators (F, FBS), thus meeting Criterion c5. In other words, they must have had a hit on B2 and C5 since either alone is insufficient for a diagnosis of malingering. The relationship between incentive status and MND diagnosis and the components of that diagnosis (b1=2, c5) was examined. Analyses The first set of analyses simply examined the frequency of positive findings on each of the malingering indicators, b1=2, c5, and MND diagnosis across each incentive level (No Incentive, State WC, LS=JA WC). Also, because previous studies examined a generic incentive group, the present study also examined an undifferentiated workers compensation group (Total WC; n ¼ 238) for comparison purposes. The percentage values represent the base-rate or prevalence of positive findings in each group. Using the frequency data, a second round of analyses used the odds-ratio statistic to examine the relative probability of being positive, given the presence of incentive. The odds ratio indicates the risk of being positive on an indicator relative to the rate of positive findings in some control group. In this case, the No Incentive group served as the standard. Thus, the odds ratio would describe the risk of a TBI

9 COMPENSATION AND RESPONSE BIAS 839 patient with incentive being positive on a malingering indicator relative to persons without incentive. Because of the nature of the data, the specific control group needs to be carefully described. The No Incentive mild TBI group was small and was never positive on any indicator. Thus, the odds ratio would be undefined because of division by zero. Therefore, the mild TBI patients with incentive were compared to all No Incentive patients. This solves the first two problems by increasing the sample size and adding positive findings so that the odds ratio could actually be calculated. It also makes the estimates more conservative because mild TBI patients with incentive are compared to both no incentive mild TBI and moderate-severe TBI. In contrast, the No Incentive moderate-severe TBI sample was larger and contained some positive findings. Therefore, it did not need to be supplemented. In any case, adding the No Incentive mild TBI patients would have put the moderate-severe TBI patients with incentive at a disadvantage in that they would be compared to persons with little or no deficit. Thus, for purposes of calculating the odds ratio, the mild TBI patients with incentive were compared to the entire No Incentive sample and the moderate-severe TBI patients with incentive were compared only to the moderate-severe TBI patients without incentive. Because none of the No Incentive patients failed the TOMM or were positive on both b1=2 and c5, odds ratios could not be calculated based on the actual data. Therefore, a 5% rate of positive findings was arbitrarily established for those two variables to allow calculation of an estimated odds ratio. RESULTS Malingering Indicators Table 4 provides the percentage of hits on the individual clinical indicators and related odds-ratio statistics. Examination of the odds ratio data in Table 4 indicates that overall, greater incentive was associated with an increase probability of being positive on the malingering indicators under study. The odds ratios for the Total WC group were minimally higher than those for the State WC group while those for the LS=JA group were, on average almost twice as high as the Total WC group. Mild TBI patients were more likely to be positive on indicators than the moderatesevere patients (except for the TOMM) and showed a stronger incentive effect overall. In the moderate-severe group there was no overall incentive effect for RDS, and MMPI-2 F; there was no dose-response effect for the PDRT and TOMM. In the mild TBI sample scale F showed an interesting effect. With limited incentive (State WC), the odds of being positive were no different than in the No Incentive group. However, the probability of being positive nearly doubled in the LS=JA group. This suggests that an increased rate of exaggerated psychiatric symptomology was not seen until greater incentive was present. In summary, greater incentive was associated with a greater likelihood of psychometric findings consistent with malingering, particularly in the mild TBI patients. This incentive effect was seen in psychometric indicators of exaggerated cognitive, physical, and psychiatric deficits and complaints. With the exception of FBS, an incentive effect was generally not seen in the moderate-severe TBI patients.

10 840 KEVIN J. BIANCHINI ET AL. Table 4 Percentage of positive findings on the five validity indicators and their related odds ratios Raw percentage Odds ratio Injury severity No incentive Total WC State LS=JA Total WC State LS= JA Portland digit recognition test Mild Mod-Servere Total Test of memory malingering 1 Mild Mod-Servere Total Reliable digit span Mild Mod-Servere Total MMPI-2 fake bad scale Mild Mod-Servere Total MMPI-2 infrequency (F) scale Mild Mod-Servere Total Odds ratios cannot be calculated when the control group (No Incentive) frequency is 0. Therefore, the odds ratios for TOMM are based on an assumed 5% base-rate in the No Incentive sample. Because the baserate is really 0, the odds ratios are underestimates. Even for FBS, the moderate-severe patients were less likely to be positive than the mild TBI patients. In the mild TBI patients, greater incentive resulted in the addition of exaggerated psychiatric symptoms as reflected in MMPI-2 F. Malingering Diagnosis Table 5 gives the percentage of patients positive on any b2 indicator (PDRT, TOMM, or RDS), c5 indicator (FBS, F), and both b2 and c5 combined (i.e, malingering diagnosis) and associated odds ratios. The same pattern of incentive effects was seen on these composite variables as was seen for the original test scores. Interestingly, almost 60% of LS=JA mild TBI patients failed at least one of the cognitive malingering indicators compared to almost 35% of those whose work injuries occurred on land. While the likelihood of being positive on a b2 indicator was almost three times higher for the State WC mild TBI patients compared to those with No Incentive; the LS=JA mild TBI patients were more than twice as likely to be positive than even the State cases. The prevalence of test performances which would meet the Slick et al. (1999) criteria for a diagnosis of MND ranged from about 18% in the State WC cases up to 33% in the LS=JA mild TBI cases. The likelihood of meeting the test performance criteria for MND in a worker with mild TBI injured in a Federal jurisdiction is

11 COMPENSATION AND RESPONSE BIAS 841 Table 5 Percentage of hits on any B2 indicator, any C5 indicator, and combined B2-C5 criterion as a function of incentive category Raw percentage Odds ratio Injury Severity No Incentive Total WC State LS=JA Total WC State LS=JA At least one B2 indicator positive (n ¼ 332) Mild ( ) 2.47 ( ) 7.31 ( ) Mod-Severe ( ) 2.42 ( ).50 ( ) Total ( ) 2.60 ( ) 3.62 ( ) At least one C5 indicator positive (n ¼ 231) Mild ( ) 3.01 ( ) 5.33 ( ) Mod-Severe ( ) 1.44 ( ) 2.88 ( ) Total ( ) 2.62 ( ) 4.86 ( ) Meets both slick B2 and C5 1 (n ¼ 234) Mild ( ) 7.75 ( ) ( ) Mod-Severe ( ) 4.41 ( ) 4.15 ( ) Total ( ) 7.03 ( ) ( ) 1 Odds ratios cannot be calculated when the control group (No Incentive) frequency is 0. Therefore, the odds ratios for Meets both Slick B2 and C5 are based on an assumed 5% base-rate in the No Incentive sample. Values in parentheses are 95% confidence intervals for the odds ratio. more than twice that of a person injured under State jurisdiction and 18 times greater than a similar injury with no incentive. These numbers are low relative to estimates of the prevalence of malingering derived via other methods (i.e., about 30%). However, in this study, an MND diagnosis was based on a limited set of research criteria, the current cutoffs are conservative, and other information that might indicate the presence of intentional exaggeration was not considered. Thus, the current cutoffs are more likely to produce false negative errors. At the same time, it is possible that some of the positive cases reflect false positive errors, particularly in the moderate-severe TBI group. But, the probability of this should be relatively low because the cut-offs used have high specificity even in moderate-severe TBI. Thus, the rates of a positive MND finding likely reflect the lower bound of the prevalence of MND in work-related TBI. In summary, incentive effects like those observed for the individual variables were also observed for the composite variables created on the basis of the Slick et al. criteria (1999) for MND. The data for MND diagnosis demonstrates what may be a lower bound for the prevalence of malingering in TBI and indicate that the likelihood of a mild TBI patient meeting the criteria for MND on the basis of psychometric data is almost twice as high in LS=JA cases as in ordinary State WC cases. DISCUSSION The study examines symptom validity test and validity index failures and their relationship to incentive and is the first study to our knowledge to examine financial incentive as more than a binary concept=construct (i.e., Incentive versus No Incentive). First, the presence of financial compensation was positively associated with the

12 842 KEVIN J. BIANCHINI ET AL. increased prevalence of findings on the various validity indicators. There was a complex relationship between injury severity and rate of positive findings. Although both mild and moderate-severe groups showed an increased failure rate in the presence of incentive, individuals with mild TBIs were more likely to have hits on malingering indicators than individuals with moderate to severe injuries (with the exception of the TOMM and MMPI-F scale where the moderate-severe individuals had slightly higher percentages of hits than the mild TBI). Odds-ratio statistics provided a clearer picture in which individuals with mild TBI showed a greater likelihood of being positive on malingering indicators when incentive was present than individuals with moderate-severe injuries. This was particularly striking with the MMPI-FBS where mild TBIs were approximately ten times more likely to be positive compared to controls, whereas moderate-severe TBIs were only about two and a half times more likely to fail. In general, both groups showed an increased likelihood of being positive when incentive was present but mild TBIs were substantially more likely to be positive than moderate-severe TBIs. These results replicate and expand upon past research on monetary impact on disability exaggeration (Binder & Rohling, 1996; Paniak et al., 2002; Price & Stevens, 1997; Reynolds et al., 2003; Youngjohn, 1995). Next, beyond the simple incentive=no incentive comparison, the magnitude of potential compensation as represented by the legal jurisdiction under which the workers compensation claim is handled was clearly associated with malingering test failures. It was hypothesized that greater incentive would be associated with high rates of positive findings on malingering indicators, with higher rates of malingering, and with greater diversity of subjective complaints. For every indicator, individuals injured offshore (Longshore Act and Jones Act cases), showed considerably higher rates of positive findings and diagnosable malingering than individuals sustaining work injuries on land. The moderate-severe TBI patients consistently showed a general incentive effect. They did not, however, show a dose-response relationship like that seen in the mild TBI patients, except for FBS. This likely reflects a consequence of the small number of moderate-severe TBI patients in the LS=JA group and does not mean that moderate-severe TBI patients do not respond differentially to higher incentive. It will be important to re-examine this finding in a larger sample of mildmoderate TBI patients. An interesting effect was observed for complaints of psychiatric symptoms. The likelihood of exaggeration of psychiatric complaints was not elevated in State cases relative to No-Incentive controls but persons with Jones Act and Longshore claims were nearly twice as likely to exaggerate their psychological symptoms. One does need to be cautious in generalizing the findings of this study as there are some important limitations. The primary limitation is sample size, particularly that of the No-Incentive control group and the LS=JA group, which often resulted in broad and over-lapping 95% confidence intervals for the odds ratios. However, the results are consistent across indicators and when indicators are used in combination (i.e., b2 and c5). Moreover, because cutoffs and computational methods were very conservative, it is likely that the present results, especially for mild TBI, underestimate the problem with exaggeration and malingering in workers compensation claims. Also, the relationship of malingering findings in Federal versus state jurisdictions may vary as a function of the potential for incentive specified by the workers

13 COMPENSATION AND RESPONSE BIAS 843 compensation laws of different states. Similar studies across states with varying laws would likely further illuminate the relationship between incentive and exaggeration=malingering. Similar research should be attempted to examine the effects of incentive in personal injury cases as well. This study also has important implications for interpretation of clinical and psychometric findings suggestive of malingering. First, while non-specific psychosocial factors doubtless do impact performance on cognitive measures and tests of effort or response bias, these results along with those of Binder and Rohling (1996) and related studies meet at least three of Hill s (1965) criteria for causality. Thus, it is reasonable to conclude that positive findings on malingering tests are motivated by financial incentive and that conclusion will be stronger with greater financial incentive. This is possible because it appears that malingering is more prevalent in Federal jurisdictions (twice the rate of State cases) with base rates potentially as much as twice the 30 to 40% prevalence estimates from previous research (Larrabee, 2005; Mittenberg et al., 2002). Positive predictive power, which states in probabilistic terms the confidence one has that a positive finding reflects the condition of interest (i.e., malingering) varies with base-rate, prevalence, or pre-test odds (Hennekens, 1987). Thus in higher incentive contexts where there is a higher prevalence of malingering, a given score would be stronger evidence of malingering than the same score in a context in which the potential compensation is more limited and therefore the rate of malingering is lower. Finally, the findings of this study related to the magnitude of the incentive raises important questions for future research. Specifically, since malingering by definition involves intentional behavior, the patient s behavior may change= vary=respond to different incentive parameters. In this study magnitude of incentive was examined; there are other important parameters, such as timing. Consider for example the proximity in time of a given examination of symptoms=disabilities to important incentive events, such as trial dates. Does the patient s behavior or incentive to malinger change in concert with such important events? For example, a patient whose attorney is not able to settle a case satisfactorily may increase reports of symptoms to maximize the appearance of disability as the trial date approaches. Similarly, the incentive to malinger a certain set of symptoms may also change depending on the outcome of other clinical circumstances of the case. For example, a worker s compensation patient who is released to return to work from a physical perspective may now have a greater incentive to manifest disabling psychological or cognitive symptoms that would leave him disabled despite his lack of physical disability. This study has already demonstrated that the variety of symptoms manifest increases with greater incentive. Understanding the complexity of malingering at this level would require research designs that take into account different incentive parameters and is an important area for future research. In summary, this study demonstrated an apparent dose-response relationship between level of potential compensation (none, limited, high) and positive findings on a range of psychometric indicators of malingering and demonstrated the addition of new symptoms in the context of the highest level of incentive. These findings support the conclusion that potential monetary compensation rather than nonspecific psychosocial factors associated with workers compensation claims is a major motive for exaggeration of problems attributed to work-related brain injuries.

14 844 KEVIN J. BIANCHINI ET AL. CONCLUSIONS A diagnosis of malingering implies that symptoms=deficits are intentionally fabricated or exaggerated in order to obtain some easily recognizable incentive (Slick et al., 1999; DSM-IV TR, 2000). The fact that patients with financial incentive to appear disabled tend to be more impaired on neuropsychological testing and fail malingering indicators more frequently than those without incentive seems to bear this out. Yet nonspecific psychosocial factors associated with the litigation process still provide a reasonable alternative explanation for the incentive effect observed in previous research. This study is the first to demonstrate that failure on malingering tests and diagnosable malingering increase as a function of potential financial compensation. Combined with previous research, this study reinforces the notion that financial incentive motivates intentional symptom exaggeration. ACKNOWLEDGMENTS The authors would like to thank Adrianne Brennan, Jeff Love, and Bridget Doane for their assistance with data collection. We would also like to thank Joseph Guilbeau for his guidance regarding the details of workers compensation law. Some data reported in this manuscript were originally presented at the 23rd annual meeting of the National Academy of Neuropsychology, Dallas, TX. REFERENCES Alexander, M. P. (1995). Mild traumatic brain injury: Pathophysiology, natural history, and clinical management. Neurology, 45, American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text revised). Washington, DC: American Psychiatric Association. Ashendorf, L., Constantinou, M., & McCaffrey, R. J. (2004). The effect of depression and anxiety on the TOMM in community-dwelling older adults. Archives of Clinical Neuropsychology, 19, Bazarian, J. J., Wong, T., Harris, M., Leahey, N., Mookerjee, S., & Dombovy, M. (1999). Epidemiology and predictors of post-concussive syndrome after minor head injury in an emergency population. Brain Injury, 13, Bernstein, D. M. (1999). Recovery from mild head injury. Brain Injury, 13, Berry, D. T. R., Baer, R., & Harris, M. (1999). Detection of malingering on the MMPI: A meta-analysis. Clinical Psychology Review, 11, Bianchini, K. J., Etherton, J. L., & Greve, K. W. (2004). Diagnosing cognitive malingering in patients with work-related pain: Four cases. Journal of Forensic Neuropsychology, 4, Q2 Bianchini, K. J., Greve, K. W., & Glynn, G. (2005). The Diagnosis of Malingered Pain- Related Disability: Lessons from cognitive malingering research. The Spine Journal, 5, Bianchini, K. J., Greve, K. W., & Love, J. M. (2003). Definite Malingered Neurocognitive Dysfunction in moderate=severe traumatic brain injury. The Clinical Neuropsychologist, 17, Q2 Bianchini, K. J., Love, J. M., Greve, K. W., & Adams, D. (2005). Detection and diagnosis of malingering in electrical injury. Archives of Clinical Neuropsychology, 20,

15 COMPENSATION AND RESPONSE BIAS 845 Bianchini, K. J., Mathias, C. W., Greve, K. W., Houston, R. J., & Crouch, J. A. (2001). Classification accuracy of the Portland Digit Recognition Test in traumatic brain injury. The Clinical Neuropsychologist, 15, Binder, L. M. (1986). Persisting symptoms after mild head injury: A review of the Postconcussion Syndrome. Journal of Clinical and Experimental Neuropsychology, 8, Binder, L. M. (1993). Portland Digit Recognition Test manual (2nd ed.). Portland, OR: Private Publication. Binder, L. M. (1997). A review of mild head trauma: II. Clinical implications. Journal of Clinical and Experimental Neuropsychology, 19, Binder, L. M. & Kelly, M. P. (1996). Portland Digit Recognition Test performance by brain dysfunction patients without financial incentives. Assessment, 3, Binder, L. M. & Rohling, M. L. (1996). Money matters: A meta-analytic review of the effects of financial incentives on recovery after closed-head injury. American Journal of Psychiatry, 153, Binder, L. M., Rohling, M. L., & Larrabee, G. (1997). A review of mild head trauma. Part I: Meta-analytic review of neuropsychological studies. Journal of Clinical and Experimental Neuropsychology, 19, Boone, K. B. & Lu, P. (2003). Noncredible cognitive performance in the context of severe brain injury. The Clinical Neuropsychologist, 17, Butcher, J. N., Dahlstrom, W. G., Graham, J. R., Tellegren, A., & Kaemmer, B. (1989). MMPI- 2: Manual for administration and scoring. Minneapolis, MN: University of Minnesota. Carroll, L. J., Cassidy, J. D., Peloso, P. M., et al. (2004). Prognosis for mild traumatic brain injury: Results of the WHO Collaborating Centre Task Force on Mild Traumatic Brain Injury. Journal of Rehabilitation Medicine, (43 Suppl), Cato, M. A., Brewster, J., Ryan, T., & Giuliano, A. J. (2002). Coaching and the ability to simulate mild traumatic brain injury symptoms. The Clinical Neuropsychologist, 16, Constantinou, M. & McCaffrey, R. J. (2003). Using the TOMM to evaluate children s effort to perform optimally on neuropsychological measures. Child Neuropsychology, 9, Denney, R. L. (1996). Symptom validity testing of remote memory in a criminal forensic setting. Archives of Clinical Neuropsychology, 11, Dikmen, S. S., Machamer, J. E., Winn, H. R., & Temkin, N. R. (1995). Neuropsychological outcome at 1-year post head injury. Neuropsychology, 9, Dikmen, S. S., Temkin, N. R., Machamer, J. E., Holubkov, A. L., Fraser, R. T., & Winn, H. R. (1994). Employment following traumatic head injuries. Archives of Neurology, 51, Eaton, D. L. & Klaassen, C. D. (2001). Principles of toxicology. In C. D. Klaassen (Ed.), Casarett and Doull s Toxicology: The basic science of poisons (pp ). New York, NY: McGraw-Hill. Etherton, J. L., Bianchini, K. J., Ciota, M. A., & Greve, K. W. (2005). Reliable Digit Span is unaffected by laboratory-induced pain: Implications for clinical use. Assessment, 12, Etherton, J. L., Bianchini, K. J., Greve, K. W., & Ciota, M. A. (2005). Test of Memory Malingering performance is unaffected by laboratory-induced pain: Implications for clinical use. Archives of Clinical Neuropsychology, 20, Etherton, J. L., Bianchini, K. J., Greve, K. W., & Heinly, M. T. (2005). Sensitivity and specificity of reliable digit dpan in malingered pain-related disability. Assessment, 12, Graham, J. R. (1990). MMPI-2: Assessing personality and psychopathology. New York, NY: Oxford Press. Green, P., Rohling, M. L., Iverson, G. L., & Gervais, R. O. (2003). Relationships between olfactory discrimination and head injury severity. Brain Injury, 17,

16 846 KEVIN J. BIANCHINI ET AL. Greiffenstein, M. F., Baker, W. J., & Gola, T. (1994). Validation of malingered amnesia measures with a large clinical sample. Psychological Assessment, 6, Greiffenstein, M. F., Baker, W. J., Gola, T., Donders, J., & Miller, L. (2002). The Fake Bad Scale in atypical and severe closed head injury litigants. Journal of Clinical Psychology, 58, Greve, K. W., Bianchini, K. J., Love, J. M., Brennan, A., & Heinly, M. T. (2006). Sensitivity and Specificity of MMPI-2 Validity Scales and Indicators to Malingered Neurocognitive Dysfunction in Traumatic Brain Injury. The Clinical Neuropsychologist, 20, Hennekens, C. H. & Buring, J. E. (1987). Epidemiology in medicine. Boston, MA: Little, Brown and Company. Hill, A. B. (1965). The environment and disease: Association or causation? Proceedings of the Royal Society of Medicine, 58, Iverson, G. L. & Binder, L. M. (2000). Detecting exaggeration and malingering in neuropsychological assessment. Journal of Head Trauma Rehabilitation, 15, Iverson, G. L., Henrichs, T. F., Barton, E. A., & Allen, S. (2002). Specificity of the MMPI-2 Fake Bad Scale as a marker for personal injury malingering. Psychological Reports, 90, Jager, T. E. (2000). Traumatic brain injuries evaluated in U.S. emergency departments, Academy of Emergency Medicine, 7, Jones Act, 46 USC Appendix 688 (2002). Kraus, J. F. & Nourjah, P. (1988). The epidemiology of mild, uncomplicated brain injury. Journal of Trauma, 28, Larrabee, G. J. (2003a). Detection of malingering using atypical performance patterns on standard neuropsychological tests. The Clinical Neuropsychologist, 17, Larrabee, G. J. (2003b). Detection of symptom exaggeration with the MMPI-2 in litigants with malingered neurocognitive dysfunction. The Clinical Neuropsychologist, 17, Larrabee, G. J. (2005). Assessment of malingering. In G. J. Larrabee (Ed.), Forensic neuropsychology: a scientific approach (pp ). New York, NY: Oxford University Press. Lees-Haley, P. R., English, L. T., & Glenn, W. J. (1991). A fake bad scale for the MMPI-2 for personal injury claimants. Psychological Reports, 68, Lees-Haley, P. R., Iverson, G. L., Lange, R. T., Fox, D. D., & Allen, L. M. (2003). Malingering in Forensic Neuropsychology: Daubert and the MMPI-2. Journal of Forensic Neuropsychology, 3, Longshore Harbor Worker s Compensation Act, 33 USC 901 et seq (1927). Mathias, G. W., Greve, K. W., Bianchini, K. B., Houston, R. J., & Crouch, J. A. (2002). Detecting Malingered Neurocognitive Dysfunction using the Reliable Digit Span in traumatic brain injury. Assessment, 9, Meyers, J. E., Millis, S. R., & Volkert, K. (2002). A validity index for the MMPI-2. Archives of Clinical Neuropsychology, 17, Meyers, J. E. & Volbrecht, M. (1998). Validation of reliable digits for detection of malingering. Assessment, 5, Mild Traumatic Brain Injury Committee of the Head Injury Interdisciplinary Special Interest Group of the American Congress of Rehabilitation Medicine. Definition of mild traumatic brain injury. (1993). Journal of Head Trauma Rehabilitation, 8, Mittenberg, W., Patton, C., Canyock, E. M., & Condit, D. C. (2002). Baserates of malingering and symptom exaggeration. Journal of Clinical and Experimental Neuropsychology, 24, Paniak, C., Reynolds, S., Toller-Lobe, G., Melnyk, A., Nagy, J., & Schmidt, D. (2002). A longitudinal study of the relationship between financial compensation and symptoms after treated mild traumatic brain injury. Journal of Clinical and Experimental Neuropsychology, 24,

17 COMPENSATION AND RESPONSE BIAS 847 Price, J. R. & Stevens, K. B. (1997). Psycholegal implications of malingered head trauma. Applied Neuropsychology, 4, Rees, L. M., Tombaugh, T. N., & Boulay, L. (2001). Depression and the Test of Memory Malingering. Archives of Clinical Neuropsychology, 16, Reynolds, S., Paniak, C., Toller-Lobe, G., & Nagy, J. (2003). A longitudinal study of compensation-seeking and return to work in a treated mild traumatic brain injury sample. Journal of Head Trauma Rehabilitation, 18, Rogers, R., Sewell, K. W., Martin, M. A., & Vitacco, M. J. (2003). Detection of feigned mental disorders: A meta-analysis of the MMPI-2 and malingering. Assessment, 10, Rohling, M. L., Meyers, J. E., & Millis, S. R. (2003). Neuropsychological impairment following traumatic brain injury: A dose-response analysis. The Clinical Neuropsychologist, 17, Ross, S. R., Millis, S. R., Krukowski, R. A., Putnam, S. H., & Adams, K. M. (2004). Detecting incomplete effort on the MMPI-2: An examination of the Fake-Bad Scale in mild head injury. Journal of Clinical and Experimental Neuropsychology, 26, Slick, D. J., Sherman, E. M. S., & Iverson, G. L. (1999). Diagnostic criteria for malingered neurocognitive dysfunction: Proposed standards for clinical practice and research. The Clinical Neuropsychologist, 13, Teichner, G. & Wagner, M. T. (2004). The Test of Memory Malingering (TOMM): Normative data from cognitively intact, cognitively impaired, and elderly patients with dementia. Archives of Clinical Neuropsychology, 19, Tombaugh, T. (1996). Test of Memory Malingering manual. New York, NY: MultiHealth Systems. Tombaugh, T. N. (1997). The Test of Memory Malingering (TOMM): Normative data from cognitively intact and cognitively impaired individuals. Psychological Assessment, 9, Tombaugh, T. N. (2002). The Test of Memory Malingering (TOMM) in forensic psychology. Journal of Forensic Neuropsychology, 2, Wechsler, D. A. (1981). Wechler adult intelligence scale-revised. New York, NY: Psychological Corporation. Wechsler, D. A. (1987). Wechsler memory scale-revised manual. New York, NY: Psychological Corporation. Wechsler, D. A. (1997a). Wechsler adult intelligence scale-iii. New York, NY: Psychological Corporation. Wechsler, D. A. (1997b). Wechsler memory scale-iii. New York, NY: Psychological Corporation. Youngjohn, J. R., Burrows, L., & Erda, l K. (1995). Brain damage or compensation neurosis? The controversial post-concussion syndrome. The Clinical Neuropsychologist, 9,