80 Screening Instruments for Mild Cognitive Impairment in the Emergency Department: The Confounding Influence of Health Literacy and Race

Abstract

Study Objectives: To compare the diagnostic accuracy of appropriately brief ED cognitive screening instruments for mild cognitive impairment (MCI) when stratified by ethnicity and health literacy.Methods: This data is an analysis of a randomized, single-center, cross sectional, consecutive sampling trial. Eligible subjects were consenting English-speaking community-dwelling patients over age 65. Using the Standards for Reporting of Diagnostic Accuracy (STARD) criteria, research assistants approached eligible subjects and assessed cognitive dysfunction with the Brief Alzheimer's Screen (BAS), Short Blessed Test (SBT), caregiver-AD8 and MoCA. Health literacy was assessed using the Rapid Assessment of Health Literacy in Medicine (REALM). Mild cognitive impairment was defined as a MoCA below 26. The diagnostic accuracies for mild cognitive impairment of the BAS, SBT, AD8 were assessed stratified by race and health literacy.Results: We enrolled 170 patients with a completed MoCA: 61% African American, 57% female, mean age 74 years old, and 34% with less than a 9th grade reading level. Mild cognitive impairment was detected in 79% of patients who completed the MoCA with 93% mild cognitive impairment prevalence in African Americans and 63% in white Americans. None of the screening instruments can accurately rule-out mild cognitive impairment, but an abnormal cAD8 or SBT may be able to rule-in mild cognitive impairment. The small sample size of mild cognitive impairment-negative African Americans limited the comparative analysis by race for each instrument. All patients with less than a 9th grade reading level were labeled as mild cognitive impairment -positive by the MoCA. Study Objectives: To compare the diagnostic accuracy of appropriately brief ED cognitive screening instruments for mild cognitive impairment (MCI) when stratified by ethnicity and health literacy. Methods: This data is an analysis of a randomized, single-center, cross sectional, consecutive sampling trial. Eligible subjects were consenting English-speaking community-dwelling patients over age 65. Using the Standards for Reporting of Diagnostic Accuracy (STARD) criteria, research assistants approached eligible subjects and assessed cognitive dysfunction with the Brief Alzheimer's Screen (BAS), Short Blessed Test (SBT), caregiver-AD8 and MoCA. Health literacy was assessed using the Rapid Assessment of Health Literacy in Medicine (REALM). Mild cognitive impairment was defined as a MoCA below 26. The diagnostic accuracies for mild cognitive impairment of the BAS, SBT, AD8 were assessed stratified by race and health literacy. Results: We enrolled 170 patients with a completed MoCA: 61% African American, 57% female, mean age 74 years old, and 34% with less than a 9th grade reading level. Mild cognitive impairment was detected in 79% of patients who completed the MoCA with 93% mild cognitive impairment prevalence in African Americans and 63% in white Americans. None of the screening instruments can accurately rule-out mild cognitive impairment, but an abnormal cAD8 or SBT may be able to rule-in mild cognitive impairment. The small sample size of mild cognitive impairment-negative African Americans limited the comparative analysis by race for each instrument. All patients with less than a 9th grade reading level were labeled as mild cognitive impairment -positive by the MoCA.