The use of the MSVT in children and adolescents with epilepsy

Performance validity tests in people with epilepsy: A review of the literature

OBJECTIVE

Professional bodies recommend the use of performance validity tests (PVTs) to aid the interpretation of scores obtained in neuropsychological assessments, but base rates of failure differ according to neurological diagnosis and the associated impairments. This review summarises the PVT literature in people with epilepsy with the aim of establishing base rates of PVT failure and the factors associated with PVT performance in this population.

METHODS

Ovid and PubMed databases were searched for studies reporting PVT test performance in people with epilepsy.

RESULTS

The search strategy identified 29 studies reporting the performance of people with epilepsy on 23 PVTs. The most commonly reported PVTs were memory-based tasks including the Test of Memory Malingering (n=15) and the Word Memory Test (n=9). Highly variable failure rates were reported on these tasks using cut-offs established in other neurological groups. Factors associated with apparent failure on PVTs in people with epilepsy included older age, lower IQ, attention deficits, impaired verbal and working memory and epilepsy-related factors including a left sided seizure focus, the presence of interictal EEG abnormalities and recent seizure activity.

CONCLUSIONS

Epilepsy related factors are associated with apparent failures on some PVTs. Caution should be employed when interpreting scores on memory based PVTs in this population.

Pass/failure on the memory validity profile: one size does not fit all

This study sought to determine the influence of various demographic variables on pass/failure on a pediatric performance validity test, the Memory Validity Profile (MVP) in a mixed clinical sample (n = 393; 62% male, 62% White, median age 11 years). Children who failed the MVP (n = 72, 18%) according to a uniform cutoff of ≤30/32 correct were younger and were more likely to have a special education history than those who passed it (n = 321, 82%). There were no statistically significant group differences on other variables such as sex, race, parental education, history of treatment for ADHD or other psychiatric disorder. Possible false positive findings were relatively most common in children under the age of 10 years and in children who received special education services under the Physical/Other Health Impairment or Speech & Language Impairment qualifications. We conclude that a uniform cutoff for pass/failure on the MVP across any and all ages and diagnostic categories is ill-advised. Instead, we recommend the development of various adjusted cutoffs for this test that maintain 90% specificity at both ends of the age spectrum as well as with different medical or developmental conditions.

New Child and Adolescent Memory Profile Embedded Performance Validity Test

OBJECTIVE

It is essential to interpret performance validity tests (PVTs) that are well-established and have strong psychometrics. This study evaluated the Child and Adolescent Memory Profile (ChAMP) Validity Indicator (VI) using a pediatric sample with traumatic brain injury (TBI).

METHOD

A cross-sectional sample of N = 110 youth (mean age = 15.1 years, standard deviation [SD] = 2.4 range = 8-18) on average 32.7 weeks (SD = 40.9) post TBI (71.8% mild/concussion; 3.6% complicated mild; 24.6% moderate-to-severe) were administered the ChAMP and two stand-alone PVTs. Criterion for valid performance was scores above cutoffs on both PVTs; criterion for invalid performance was scores below cutoffs on both PVTs. Classification statistics were used to evaluate the existing ChAMP VI and establish a new VI cutoff score if needed.

RESULTS

There were no significant differences in demographics or time since injury between those deemed valid (n = 96) or invalid (n = 14), but all ChAMP scores were significantly lower in those deemed invalid. The original ChAMP VI cutoff score was highly specific (no false positives) but also highly insensitive (sensitivity [SN] = .07, specificity [SP] = 1.0). Based on area under the curve (AUC) analysis (0.94), a new cutoff score was established using the sum of scaled scores (VI-SS). A ChAMP VI-SS score of 32 or lower achieved strong SN (86%) and SP (92%). Using a 15% base rate, positive predictive value was 64% and negative predictive value was 97%.

CONCLUSIONS

The originally proposed ChAMP VI has insufficient SN in pediatric TBI. However, this study yields a promising new ChAMP VI-SS, with classification metrics that exceed any other current embedded PVT in pediatrics.

Nonmemory Composite Embedded Performance Validity Formulas in Patients with Multiple Sclerosis

OBJECTIVE

Research regarding performance validity tests (PVTs) in patients with multiple sclerosis (MS) is scant, with recommended batteries for neuropsychological evaluations in this population lacking suggestions to include PVTs. Moreover, limited work has examined embedded PVTs in this population. As previous investigations indicated that nonmemory-based embedded PVTs provide clinical utility in other populations, this study sought to determine if a logistic regression-derived PVT formula can be identified from selected nonmemory variables in a sample of patients with MS.

METHOD

A total of 184 patients (M age = 48.45; 76.6% female) with MS were referred for neuropsychological assessment at a large, Midwestern academic medical center. Patients were placed into "credible" (n = 146) or "noncredible" (n = 38) groups according to performance on standalone PVT. Missing data were imputed with HOTDECK.

RESULTS

Classification statistics for a variety of embedded PVTs were examined, with none appearing psychometrically appropriate in isolation (areas under the curve [AUCs] = .48-.64). Four exponentiated equations were created via logistic regression. Six, five, and three predictor equations yielded acceptable discriminability (AUC = .71-.74) with modest sensitivity (.34-.39) while maintaining good specificity (≥.90). The two predictor equation appeared unacceptable (AUC = .67).

CONCLUSIONS

Results suggest that multivariate combinations of embedded PVTs may provide some clinical utility while minimizing test burden in determining performance validity in patients with MS. Nonetheless, the authors recommend routine inclusion of several PVTs and utilization of comprehensive clinical judgment to maximize signal detection of noncredible performance and avoid incorrect conclusions. Clinical implications, limitations, and avenues for future research are discussed.