Cognitive change in donepezil treated patients with vascular or mixed dementia

Background: vascular dementia (VaD) and mixed Alzheimer's disease (AD/VaD) are common. How best to monitor treatment is not clear. Our objective was to compare responsiveness and construct validity of change scores, following donepezil treatment, of the standardized Mini-Mental State Examination (sMMSE) and other measures potentially usable in primary care. Methods: A six-month, outcome measurement study. The Disability Assessment for Dementia (DAD), ClOX-1 and 2, Phonetic Fluency, a short Neuropsychiatric Inventory, (the NPI-Q), Clinical Global Impression (CGI) and the SymptomGuide™ (SG) were measured. Construct validity was tested by correlating change scores, and responsiveness by calculating standardized response means (SRMs). Results: Of 148 treated patients, 116 completed. The mean sMMSE increased by 0.7 (95% Confidence Interval (CI) = -0.005, 1.41; p=0.06; SRM= 0.15). There was no statistically significant difference in the DAD. The NPI-Q (-1.4; 95% CI = -2.08, -0.72; p<0.01; SRM=0.24), ClOX-1 (0.9; 95% CI = 0.19, 1.61; p<0.01; SRM=0.21), ClOX-2 (0.9; 95% CI = 0.17, 1.63; p=0.03; SRM=0.26), Phonetic Fluency (0.9; 95% CI = 0.19, 1.61; p=0.02; SRM=0.21) and SG (0.35; 95% CI = 0.20,0.51; p<0.01; SRM=0.28) each detected significant improvement. The CGI suggested improvement in 74 completers (64%) - mostly "minimal" (44/116, 38%) - while 21/116 (18%) were worse. Change scores at 24 weeks were at best modestly correlated with each other (range -0.22 to 0.30). Discussion: Different measures showed different responsiveness, in a setting in which the mean treatment effect seems to have been small, but clinically detectable. Patient-centered and executive function measures might be useful in vascular and mixed dementia.