Finding dementia in primary care: the results of a clinical demonstration project

To assess the effect of screening on diagnosing cognitive impairment.

Quality improvement initiative.

Seven Veterans Affairs Medical Centers.

Veterans aged 70 or older without a prior diagnosis of cognitive impairment.

Veterans failing a brief cognitive screen (Mini-Cog score <4/5) at a routine primary care visit were offered a further, comprehensive evaluation with an advance practice registered nurse trained in dementia care and integrated into the primary care clinic. Veterans completing the evaluation were reviewed in a consensus conference and assigned a diagnosis of dementia; cognitive impairment, no dementia; or no cognitive impairment. Total numbers of screens, associated scores (0-5), and the consensus diagnoses were tallied. New cognitive impairment diagnoses were also tracked for veterans who passed the screen but requested further evaluation, failed but declined further evaluation, or were not screened. Primary care provider satisfaction with the program also was assessed.

Of 8,342 veterans offered screening, 8,063 (97%) accepted, 2,081 (26%) failed the screen, 580 (28%) agreed to further evaluation, and 540 (93%) were diagnosed with cognitive impairment, including 432 (75%) with dementia. For screen passes requesting further evaluation, 87% (103/118) had cognitive impairment, including 70% (82/118) with dementia. Screen failures declining further evaluation had 17% (259/1,501) incident cognitive impairment diagnosed through standard care, bringing the total newly documented cognitive impairment in all screens to 11% (902/8,063), versus 4% (1,242/28,349) in similar clinics without this program. Eighty-two percent of primary care providers in clinics with this program agreed that it provided a useful service.

Screening combined with offering further evaluation increased new diagnoses of cognitive impairment in older veterans two to three times. Veterans accepted screening well, and providers found the program useful.