The development of ACE-III performance validity tests for dementia screening.

Abstract

Objective: Performance validity tests (PVTs) are an essential component of a neuropsychological evaluation. There is a lack of brief validated PVTs for use alongside cognitive screening assessments for dementia. The current study sought to develop and evaluate a series of brief PVTs, which could be embedded into the Addenbrooke's Cognitive Examination (ACE-III) and used in a dementia clinic.

Method: This study used a diagnostic accuracy design to compare the performance of 31 individuals with Alzheimer's disease dementia and 26 healthy older adults who simulated having Alzheimer's disease. Participants were asked to complete ACE-III, the Test of Memory Malingering (TOMM), and five new brief PVTs developed based on ACE-III items, including the Object Recognition Test, Word Recognition Test, Dot Counting, ACE-III Digit Span, and the Coin in the Box test.

Results: The TOMM Trial 2 achieved excellent classification accuracy (AUC = 0.92). However, only 65% of people with moderate dementia were able to complete this test, and the standard cut off (45) had low specificity (0.69). Of the ACE-III PVTs, the Combined Object Recognition Test and ACE-III Digit Span test were both accurate in distinguishing between genuine and simulated cognitive impairment and achieved high levels of diagnostic accuracy (AUC = 0.89 and 0.85 respectively). The remaining PVTs achieved modest classification ability.

Conclusions: The findings of this research demonstrate that the Object Recognition and ACE-III Digit Span tests are candidate PVTs to be incorporated into the ACE-III for cognitive screening assessments. However, research using a known-group design and confirmatory data analysis procedures are required before these tests can be recommended for clinical practice. Similarly, the administration of PVTs by non-neuropsychologists during cognitive screening will require careful consideration.