Mental Health Service, Ralph H. Johnson Veterans Affairs Health Care System, Charleston, SC, USA.
Department of Psychiatry and Behavioral Sciences, Medical University of South Carolina, Charleston, SC, USA.
J Clin Exp Neuropsychol. 2024 Feb;46(1):25-35. doi: 10.1080/13803395.2024.2314737. Epub 2024 Feb 14.
It is common to use normative adjustments based on race to maintain accuracy when interpreting cognitive test results during neuropsychological assessment. However, embedded performance validity tests (PVTs) do not adjust for these racial differences and may result in elevated rates of false positives in African American/Black (AA) samples compared to European American/White (EA) samples.
Veterans without Major Neurocognitive Disorder completed an outpatient neuropsychological assessment and were deemed to be performing in a valid manner (e.g., passing both the Test of Memory Malingering Trial 1 (TOMM1) and the Medical Symptom Validity Test (MSVT), ( = 531, EA = 473, AA = 58). Five embedded PVTs were administered to all patients: WAIS-III/IV Processing Speed Index (PSI), Brief Visuospatial Memory Test-Revised: Discrimination Index (BVMT-R), TMT-A (secs), California Verbal Learning Test-II (CVLT-II) Forced Choice, and WAIS-III/IV Digit Span Scaled Score. Individual PVT false positive rates, as well as the rate of failing two or more embedded PVTs, were calculated.
Failure rates of two embedded PVTs (PSI, TMT-A), and the total number of PVTs failed, were higher in the AA sample. The PSI and TMT-A remained significantly impacted by race after accounting for age, education, sex, and presence of Mild Neurocognitive Disorder. There were PVT failure rates greater than 10% (and considered false positives) in both groups (AA: PSI, TMT-A, and BVMT-R, 12-24%; EA: BVMT-R, 17%). Failing 2 or more PVTs (AA = 9%, EA = 4%) was impacted by education and Mild Neurocognitive Disorder but not by race.
Individual (timed) PVTs showed higher false positive rates in the AA sample even after accounting for demographic factors and diagnosis of Mild Neurocognitive Disorder. Requiring failure on 2 or more embedded PVTs reduced false positive rates to acceptable levels across both groups (10% or less) and was not significantly influenced by race.
在神经心理评估中，为了准确解释认知测试结果，通常会使用基于种族的规范调整。然而，嵌入式绩效验证测试（PVT）并没有调整这些种族差异，这可能导致与欧洲裔美国人/白人（EA）样本相比，非裔美国人/黑人（AA）样本中的假阳性率升高。
没有主要神经认知障碍的退伍军人完成了门诊神经心理评估，并被认为表现出有效的方式（例如，通过记忆欺骗测试 1（TOMM1）和医疗症状有效性测试（MSVT），（ = 531，EA = 473，AA = 58）。所有患者均进行了 5 种嵌入式 PVT 测试：WAIS-III/IV 处理速度指数（PSI）、简要视觉空间记忆测试修订版：辨别指数（BVMT-R）、TMT-A（秒）、加利福尼亚语言学习测试-II（CVLT-II）强制选择和 WAIS-III/IV 数字跨度量表分数。计算了单个 PVT 假阳性率以及两个或多个嵌入式 PVT 失败的比率。
AA 样本中，两个嵌入式 PVT（PSI、TMT-A）失败率和总 PVT 失败率更高。在考虑年龄、教育程度、性别和轻度认知障碍存在的情况下，PSI 和 TMT-A 仍然受到种族的显著影响。两个组的 PVT 失败率都超过了 10%（被认为是假阳性）（AA：PSI、TMT-A 和 BVMT-R，12-24%；EA：BVMT-R，17%）。两个或多个 PVT 失败（AA = 9%，EA = 4%）受教育程度和轻度认知障碍的影响，但不受种族的影响。
即使考虑了人口统计学因素和轻度认知障碍的诊断，单个（计时）PVT 在 AA 样本中的假阳性率也更高。要求 2 个或更多嵌入式 PVT 失败将两个组的假阳性率降低到可接受水平（10%或更低），并且不受种族的显著影响。
