Quantitative measures of memory malingering on the Wechsler Memory Scale—Third edition in mild head injury litigants

Archives of Clinical Neuropsychology 18 (2003) 181–197 Pauline M. Langeluddecke a,∗ , Sara K. Lucas b a b Suite 1003, 37 Bligh Street, Sydney 2000, NSW, Australia Department of Medical Psychology, Westmead Hospital, Westmead, NSW, Australia Accepted 10 December 2001 Abstract Wechsler Memory Scale—Third edition (WMS-III) performance in 25 mild traumatic brain injury (TBI) litigants who met the criteria for probable malingered neurocognitive dysfunction (MND) was compared with 50 nonmalingering subjects. The base rate for probable MND in the population studied was 27%. Overall, malingerers showed globally depressed memory function. They returned significantly poorer scores than nonmalingerers on all WMS-III indexes and subtests, and on selected WMS-III index difference scores and intelligence–memory difference scores. Using the minimum score returned in the nonmalingerers as the cut-off for malingering, the delayed auditory recognition memory tasks were highly effective in detecting malingering. Raw scores below 43 on the auditory recognition-delayed (AR-D) subtest or below 18 on word list II-recognition, identified around 80% of the malingerers. In a group of 50 severe TBI litigants, only a very small proportion (i.e., <10%) returned scores below the cut-offs for malingering for the mild TBI subjects. © 2002 National Academy of Neuropsychology. Published by Elsevier Science Ltd. All rights reserved. Keywords: Malingered neurocognitive dysfunction; Traumatic brain injury; Neuropsychological assessment 1. Introduction Various definitions and criteria for malingering have been proposed, those of the Diagnostic and Statistical Manual of Mental Disorders—Fourth edition (DSM-IV) probably being the ∗ Corresponding author. Tel.: +61-2-9238-0200; fax: +61-2-9238-0244. E-mail address: (P.M. Langeluddecke). 0887-6177/02/$ – see front matter © 2002 National Academy of Neuropsychology. PII: S 0 8 8 7 - 6 1 7 7 ( 0 1 ) 0 0 1 9 5 - 0 Quantitative measures of memory malingering on the Wechsler Memory Scale—Third edition in mild head injury litigants 182 P.M. Langeluddecke, S.K. Lucas / Archives of Clinical Neuropsychology 18 (2003) 181–197 most widely used (American Psychiatric Association, 1994). In an attempt to facilitate both clinical practice and systematic research in relation to malingering in neuropsychology, Slick, Sherman, and Iverson (1999) have proposed a set of diagnostic criteria for possible, probable, and definite malingering of cognitive dysfunction based on psychometric, behavioural, and collateral data. “Probable” malingering of neurocognitive dysfunction (MND) requires a substantial incentive to fabricate cognitive dysfunction, the presence of evidence strongly suggesting volitional exaggeration or fabrication of cognitive dysfunction, and the absence of plausible alternative explanations. The prevalence of conscious and intentional faking of cognitive impairment on neuropsychological assessment is unknown. Malingerers rarely identify themselves and foolproof methods of detecting malingering remain to be found. Base rates for malingering of cognitive impairment in litigants remain speculative, but appear to be quite high. A meta-analysis of studies evaluating the impact of financial incentives on persistent symptomatology after closed head injury reported a particularly strong effect size (of 1.47 S.D.) in the mild head injury group (Binder & Rohling, 1996). These researchers concluded that “money matters” and they recommended that malingering be considered in all medico-legal cases. Actual estimates of malingering in the population of head injury litigants are reported to vary considerably, allegedly from 1 to 50% (Reynolds, 1998). The present study assessed the prevalence of probable MND, with particular emphasis on feigned memory impairment, in a mild head injury litigant population. Base rate information is important in clinical practice and is critical in interpreting the sensitivity and specificity of malingering measures (Bianchini, Mathias, & Greve, 2001; Gouvier, Hayes, & Smiroldo, 1998). Memory disorder is a common complaint in relation to traumatic brain injury (TBI) and is thought to be the type of cognitive disability which is most likely to be feigned (Williams, 1998). Various techniques have been developed to screen for dissimulation of memory impairment, a detailed discussion of which is beyond the scope of the present paper (see reviews of Haines & Norris, 1995; Millis & Putnam, 1996; Nies & Sweet, 1994; Slick et al., 1999; Williams, 1998). Some methods utilize qualitative aspects of test performance (e.g., slowness in responding, Beetar & Williams, 1995; items failed, Killgore & DellaPietra, 2000) or excessive inconsistencies in test performance over time (Reitan & Wolfson, 1998). Others involve tests designed to identify feigned or exaggerated memory impairment, the Rey 15-item Visual Memory Test (RVMT; Rey, 1964) probably being the most popular of these. This task presents the subject with a task which is easy but appears difficult, on which the malingerer is identified because of a performance which is poorer than, or different to, that of even the severely brain damaged person. This instrument is sometimes used in forensic assessments (Lees-Haley, Smith, Williams, & Dunn, 1995) but remains poorly researched in litigating samples and has been criticized regarding both its sensitivity and specificity to malingering (Arnett, Hammeke, & Schwartz, 1995; Lee, Loring, & Martin, 1992; Schretlen, Brandt, & Krafft, 1991; Vickery, Berry, Hanlon Inman, Harris, & Orey, 2001). An approach to screening for malingered memory impairment which is gaining increasing popularity involves the use of forced choice responding, generally in a recognition memory format. Early research utilizing this type of technique classified persons who returned a score that was below a chance level as malingerers on the grounds that it is necessary to consciously provide incorrect responses to perform this poorly. The usefulness of this technique is limited P.M. Langeluddecke, S.K. Lucas / Archives of Clinical Neuropsychology 18 (2003) 181–197 183 because very few (i.e., less than 30%) of malingering subjects actually perform at a lower than chance level (Bianchini et al., 2001; Trueblood & Schmidt, 1993; Williams, 1998) and, thus, it detects only “the most blatant malingerers” (Haines & Norris, 1995, p. 138). Attempts have been made to preserve this format but use less stringent cut-off points to identify malingering, an approach which appears well suited to mild head injury samples where the likelihood of gross memory impairment is minimal. Tests on which this method has been utilized and shown considerable promise include the Warrington Recognition Memory Test (WRMT) (e.g., Iverson & Franzen, 1998) and the recognition trials of both the Rey Auditory Verbal Learning Test (Bernard, 1990; Greiffenstein, Baker, & Gola, 1994) and California Verbal Learning Test (...truncated)