Background/Aims: Diabetes mellitus (DM), hypertension (HTN), and hyperlipidemia (HL) are common geriatric chronic diseases that may increase dementia risk. Screening for cognitive impairment is not currently recommended due to lack of studies demonstrating benefit. HealthPartners piloted using the Mini-Cog to screen patients aged >=65 years and physicians expressed concern about the feasibility of screening all patients. In order to identify sub-groups of patients who may potentially benefit most from cognitive screening, we examined screen fail rates, retrospective healthcare utilization, and the impact of multiple chronic conditions in this cohort. Methods: Data from the 18 months prior to cognitive screening was collected from the electronic medical record and included the Mini-cog score (scored 0-5, fail is less than 4), demographics, presence of diagnosis for 3 chronic diseases, and measures of healthcare utilization. The cohort was divided into 8 sub-groups: DM alone, HTN alone, HL alone, DM+HTN, DM+HL, HTN+HL, DM+HTN+HL, and no chronic conditions. Utilization outcomes were analyzed using Poisson regression accounting for age and sex; the DM only model was not useable due to a small sample size. Results: The mean age of the cohort (n=1124) was 77 years (41% male) and the overall cognitive screen fail rate was 32%. Fail rates were highest in the small DM group (50% fail), followed by patients lacking the 3 chronic conditions (36% fail), and lowest in the DM+HL (21%) and HTN (26%) cohorts. Overall, patients failing the screen had a significantly higher incidence rate of hospitalizations (23%) and emergency room visits (42%). This result was driven by 2 sub-groups of multiple chronic conditions, DM+HTN and DM+HTN+HL, with those failing having significant increases in rates of hospitalizations (535% and 56% respectively) and emergency room visits (252%, 102%). Discussion: Based upon higher retrospective utilization, patients with DM+HTN and DM+HTN+HL may benefit most from cognitive screening. Surprisingly, higher screen fail rates were found in patients without DM, HTN, or HL, which may be partially explained by a lower rate of clinic visits (mean 0.98 visits per month vs 1.33). Prospective longitudinal studies are necessary to better support the role of the screening in specific groups of patients with chronic conditions.