Performance of the participant self-rating version of the Quick Dementia Rating System in a racially diverse cohort of non-demented older adults

Background Early detection of clinical Alzheimer's disease and related dementias is a public health priority. The Clinical Dementia Rating is used for staging but is time intensive. The Quick Dementia Rating System (QDRS) is a brief, self-administered alternative, although studies have focused on informant rather than patient ratings. Objective This study evaluated the clinical utility of the QDRS Participant Self-Rating version for classifying cognitive status of White and Black/African Americans. Methods Participants (79 White, 95 Black/African American) enrolled in the Goizueta Alzheimer's Disease Research Center with Clinical Dementia Rating (CDR) Global scores of 0.0 or 0.5 completed the QDRS. Concordance rates and associations with the Montreal Cognitive Assessment (MoCA) were examined. Results Agreement between QDRS and CDR Global scores was 69% (κ = 0.36), and comparable between racial groups. QDRS Sum of Boxes demonstrated moderate correlations with CDR Sum of Boxes for the full sample and each group. Participants classified as impaired on the QDRS had lower MoCA Total and Memory Index scores, with similar effect sizes between racial groups. ROC analyses demonstrated good discrimination of CDR 0.0 versus 0.5 for the overall sample (AUC = 0.789, SE = 0.040, 95% CI = 0.711–0.866) and for White (AUC = 0.792, SE = 0.055, 95% CI = 0.683–0.900) and Black/African American participants (AUC = 0.777, SE = 0.058, 95% CI = 0.663–0.891). Positive predictive value of the QDRS Global score was 50%, and negative predictive value was 85%. Conclusions The QDRS Participant Self-Rating version is a clinically useful prescreening tool to rule out cognitive impairment in both White and Black/African American persons with early cognitive impairment.