Dementia has high public health significance due to its prevalence, adverse impact on patients and caregivers, high economic cost to society and the rapidly expanding population of older Americans. As the largest provider of geriatric care in the U.S., dementia is particularly relevant to the VHA. However, only a small proportion of patients with dementia are diagnosed early when treatments are most effective and the opportunity for careful life planning remains, in part because of the absence of feasible, well-validated case-finding instruments.
Our primary aim was to evaluate the performance of four cognitive screening instruments for dementia and for the combined category of dementia or cognitive impairment-not dementia (CIND). We also determined the prevalence of cognitive impairment in veterans attending primary care clinics age 65 and older and followed a subset of CIND patients longitudinally to determine the rate of progression to dementia.
We recruited a random sample of primary care patients age 65 years old, without an ICD9 code for dementia, who completed four brief cognitive screeners that were compared to an independent diagnostic standard for dementia or CIND. The sensitivity, specificity, and diagnostic odds ratios were calculated for dementia and for CIND or dementia. Care patterns for patients with dementia and CIND were described based on detailed chart abstractions. A subset of patients was reassessed with the diagnostic standard evaluation.
Of 826 eligible Veterans, 630 enrolled and completed all assessments. Participant characteristics were: mean age 74.8 years old (SD 36.6), mean education 13.4 years (SD 3.4), 73.2% white race and 26.2% black race, and with high rates of chronic medical conditions. Dementia was diagnosed in 3.3% and CIND in 39.2% of the sample. Using ICD9 codes for dementia in combination with our observed rate of unrecognized dementia, we estimate the overall prevalence of dementia to be 10% in the population of veterans, age 65 years old, enrolled in the three participating primary care clinics. Of the four cognitive screeners, three (the 3MS, Mini-Cog , and a novel 2-item functional screen), performed moderately well for dementia: sensitivity (73%-86%), specificity (73%-83%), diagnostic odds ratio (8.5 to 22.4). None of the screeners performed well for identifying patients with dementia or CIND. Of the 131 patients with CIND who were reassessed at a median of 2.5 years later, 16 (12.2%) were diagnosed with dementia.
The prevalence of cognitive impairment in older veterans attending primary care clinics is high. Several cognitive screeners have adequate performance characteristics to identify dementia but not CIND. The proportion who progress from CIND to dementia is substantial. These results suggest a need for trials of interventions to delay or prevent progression of cognitive impairment. The data on performance characteristics can be used along with evidence on treatment effects to develop evidence informed policies on dementia screening in VHA.
External Links for this Project
- Holsinger T, Plassman BL, Stechuchak KM, Burke JR, Coffman CJ, Williams JW. Screening for cognitive impairment: comparing the performance of four instruments in primary care. Journal of the American Geriatrics Society. 2012 Jun 1; 60(6):1027-36. [view]
- Holsinger T, Plassman BL, Stechuchak KM, Burke JR, Coffman CJ, Williams JW. Stability of Diagnoses of Cognitive Impairment, Not Dementia in a Veterans Affairs Primary Care Population. Journal of the American Geriatrics Society. 2015 Jun 1; 63(6):1105-11. [view]
- Plassman BL, Williams JW, Burke JR, Holsinger T, Benjamin S. Systematic review: factors associated with risk for and possible prevention of cognitive decline in later life. Annals of internal medicine. 2010 Aug 3; 153(3):182-93. [view]
- Williams JW. Innovations in Integrated Mental Health and Primary Care. Paper presented at: VA ORD State of the Art (SOTA) Conference; 2008 Mar 1; Tysons Counter, VA. [view]