Detection of MCI in the clinic: evaluation of the sensitivity and specificity of a computerised test battery, the Hopkins Verbal Learning Test and the MMSE

Abstract

Introduction: the sensitive detection of mild cognitive impairment (MCI) in older adults is an important problem that requires objective assessment. We evaluated whether the computerised cognitive test battery, CogState, was as sensitive to MCI as two well-validated 'paper-and-pencil' tests, the Hopkins Verbal Learning Test (HVLT) and the Mini-Mental Status Examination (MMSE).

Methods: these tests were administered with a subjective memory questionnaire and an 'Activities of Daily Living' scale to 21 individuals with MCI and 98 cognitively healthy controls matched for sex, education and IQ levels. The sensitivity and specificity of the tests and their discrimination between groups were determined.

Results: the HVLT had a maximum discrimination between controls and MCI cases of 90%, compared with 86% for CogState and 65% for the MMSE. Only CogState showed correlations with subjective memory complaints (SMC) and activities of daily living for the whole cohort when controlled for age, sex and years of education. Logistic regression analyses showed that diagnosis (control:MCI) was predicted by HVLT and a CogState ratio score. Age was a significant predictor of HVLT performance, while age and SMC predicted CogState performance. The computerised test battery was well tolerated by older adults, but presentation speed was a limiting factor for some participants.

Conclusions: overall, we conclude that the HVLT has better sensitivity for the detection of MCI in older adults than the CogState, but that CogState may enable the identification of cognitive deficits above and beyond impairments in memory.