Digit Span Forward as a Performance Validity Test in Dementia Evaluations: Specificity in Mild Cognitive Impairment, Mild Dementia, and Moderate Dementia

Objective Traditional performance validity tests (PVTs) often yield high false positive rates in dementia evaluations. The current study examined the frequency of extremely low scores (≤ 2 percentile) on WAIS-IV Digit Span Forward (DSF) in older adults with Mild Cognitive Impairment (MCI) or dementia to evaluate its possible utility as a PVT in these populations. Method Archival data from outpatient neuropsychological evaluations were analyzed. Individuals who were not diagnosed with a neurocognitive disorder, had missing data, or were believed to be invalidly performing were excluded. Participants (n = 195; mean age = 72.8; mean education = 13.2 years) were classified according to their evaluation diagnosis of MCI (n = 72; mean RBANS Total Score = 86.8) or dementia. Dementia patients were further divided by MoCA score into groups of mild dementia (n = 90; MoCA≥15; mean RBANS Total Score = 71.0) or moderate dementia (n = 33; MoCA < 15; mean RBANS Total Score = 55.9). Frequencies of scaled scores were analyzed to calculate specificity values for each group. Results A WAIS-IV DSF scaled score of ≤4 (≤ 2 percentile) resulted in specificity values of 0.99 and 0.94 in MCI and mild dementia, respectively. Conversely, in moderate dementia, ≥0.90 specificity was achieved only when using a more conservative cutoff of ≤2. Conclusions Low DSF scaled scores occurred infrequently in MCI and mild dementia, indicating strong specificity and potential utility as a PVT in these populations. However, in moderate dementia, low DSF scores were more common, requiring use of a more stringent cutoff. Future research should examine DSF sensitivity to invalid performance, as well as DSF specificity according to specific etiologies of MCI and dementia.