Characterization of the Medical Symptom Validity Test in evaluation of clinically referred memory disorders clinic patients

Archives of Clinical Neuropsychology 22 (2007) 753–761 Characterization of the Medical Symptom Validity Test in evaluation of clinically referred memory disorders clinic patients Laura L.S. Howe a,∗ , Ashton M. Anderson a , David A.S. Kaufman a , Bonnie C. Sachs a , David W. Loring b,a a Department of Clinical and Health Psychology, University of Florida, Gainesville, FL, United States b Department of Neurology, University of Florida, Gainesville, FL, United States Accepted 6 June 2007 Abstract We prospectively evaluated performance of 63 referrals to a memory disorders clinic who received the Medical Symptom Validity Test (MSVT) as part of their standard neuropsychological evaluation. The patients were grouped based on independent medical diagnoses and presence or absence of a potential financial incentive to under-perform. Twenty-seven patients (42.9%) scored below cutoffs on the MSVT symptom validity indices. Two individuals in the potential financial incentive group showed clear signs of invalid responding (18.2%). Twenty-two of the remaining 25 patients who failed the symptom validity indices corresponded to the dementia profile. Three individuals did not correspond to the dementia profile but are thought to have performed validly representing a 4.8% false positive rate. When considering all MSVT indices, the base rate of invalid responding in the potential financial incentive to under-perform group increased to 27.3%. Combining all groups our base rate of invalid responding was 4.8%. Specific performances are presented. © 2007 National Academy of Neuropsychology. Published by Elsevier Ltd. All rights reserved. Keywords: Symptom validity; Dementia; Medical symptom validity test; Neuropsychological assessment; Forensic neuropsychology 1. Introduction Symptom validity tests (SVTs) are commonly used in medico-legal contexts to assess response bias since a potential incentive to perform poorly exists. Malingering base rates in forensic neuropsychological evaluations have been estimated to be 30–40% (Larrabee, 2003; Mittenberg, Patton, Canyock, & Condit, 2002). The use of SVTs for routine clinical neuropsychological evaluations, however, is less universally agreed upon because a clear incentive to perform badly is not readily apparent in most cases and, unlike assessments performed in a medico-legal context, time and other assessment resources may be limited (Loring, Lee, & Meador, 2005). Thus, clinical test selection is often based upon expected cost–benefit relationships within the context of the specific clinical circumstances and referral questions. Some recent studies have shown unexpectedly high rates of SVT failure levels in clinical populations without clearly identifiable incentives to perform poorly. Drane et al. (2006) reported 51% of patients with nonepileptic seizures (NES) failed a SVT compared to 8% of patients with epilepsy. One possible explanation for the high rates of SVT failure in selected clinical series is that patients may be either applying for or receiving disability payments, giving rise to a fear ∗ Corresponding author at: 205 SE 16th Ave., Apt. 3G, Gainesville, FL 32601, United States. Tel.: +1 631 365 3157. E-mail address: (L.L.S. Howe). 0887-6177/$ – see front matter © 2007 National Academy of Neuropsychology. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.acn.2007.06.003 754 L.L.S. Howe et al. / Archives of Clinical Neuropsychology 22 (2007) 753–761 of benefit denial or withdrawal. Gervais, Russell, Green, Ferrari, and Pieschl (2001) reported discrepant SVT failure rates based on disability seeking status. Of fibromyalgia patients either applying for or currently receiving disability benefits, 35% failed a SVT. In contrast, only 4% of the fibromyalgia group not seeking or receiving disability benefits failed. Interestingly, 44% of patients seeking disability failed the SVT while only 23% who were already currently on disability scored below cutoffs. Thus, in addition to patients currently applying for disability, there appears to be some perceived incentive for patients already on disability to alter their behavior on neuropsychological measures in order to maintain benefit continuation. In a sample of epilepsy surgery candidates, Loring et al. (2005) observed SVT performances that were in either the questionable or invalid range in 24/120 patients, although information on disability status was not available to be analyzed. More recently, these same authors found that although intentionally distorted response profiles (i.e., less than chance responding) were rare in a mixed group of clinically referred patients, dementia patients often obtained scores that were below various empirically derived cutoffs (Loring, Larrabee, Lee, & Meador, 2007). In order to better understand the implications of SVT failures, it is important to know the base rates and characterization of SVT performance in clinical populations without clearly established incentive to perform poorly. Failure to consider base rate information can lead to misdiagnosis and errors in clinical decision-making (Larrabee, 2005). In particular, practitioners may incorrectly question the validity of some neuropsychological results based on a perceived failed SVT performance when the SVT performance is actually valid and corresponds to a known characteristic profile of performance on that particular SVT (e.g., a dementia profile). An individual giving valid test responses based on true abilities but identified as giving invalid data on a SVT is a false positive error. For SVTs with multiple indices and characterized response profiles, if a patient scores below cutoffs but the SVT profile corresponds to a known responding profile in neurological patients with good task engagement, the individual is not counted as a false positive because the validity of the neuropsychological data is not questioned. Practitioners unfamiliar with particular responding profiles on SVTs, however, may label cases as invalid when they are not. This underscores the importance of characterizing SVT profiles/performance in patient populations and the need for practitioners’ to be familiar with the profiles. Symptom validity testing provides one form of verification of our clinical decision-making. The HIV Neurobehavioral Research Center (HNRC) Group (2003) noted a need for base rate data on SVTs in clinical populations that may be likely to seek disability benefits (Woods et al., 2003). Many studies investigating SVTs in clinical and community dwelling populations have found performance above cutoff scores on the SVTs examined. For example, when investigating subjects with HIV-associated neurocognitive impairment, Woods et al. (2003) found only a 2% base rate of failure on the Hiscock Digit Memory Test (Hiscock & Hiscock, 1989). Similarly, all inpatients diagnosed with major depression (Rees, Tombaugh, & Boulay, 2001) and all community dwelling older adults with mild-to-moderate depr (...truncated)