Comparing Clinician Impression of Cognition with Standardized Screening in Medicare Annual Wellness Visits

Abstract

Early identification of cognitive changes is crucial for accessing timely interventions, implementing care planning, and optimizing quality of life for patients. Underdiagnosis of cognitive changes, particularly the subtle findings associated with mild cognitive impairment, is a significant issue in primary care. This pilot study compared provider’s impression of a patient’s cognitive status (provider gestalt) with a standardized screening tool (Mini-Cog^©) during the Medicare Annual Wellness Visit (AWV). During patient encounters, medical assistants administered the Mini-Cog in addition to the standard AWV. Clinicians were blinded to the Mini-Cog test result and provided an independent impression of cognitive status. Cohen’s kappa was calculated to determine rates of agreement between providers’ impression and Mini-Cog findings. Statistical analysis with Cohen’s kappa indicated only fair agreement between the provider’s impression of cognitive impairment and the Mini-Cog result (N = 76, k = 0.315). When discordance occurred between provider impression and Mini-Cog results, providers were more likely to document “no concern” for cognitive impairment in the presence of an abnormal Mini-Cog. Specifically, 19.7% of patients demonstrated a positive Mini-Cog when providers identified no cognitive concerns versus 5.3% of patients who demonstrated a negative Mini-Cog when providers documented concerns for cognitive impairment. Our study suggests under-identification of patients with cognitive impairment when provider impression alone is used to guide the diagnosis. Utilization of a standardized screening tool, such as the Mini-Cog, may minimize the risk of missing early signs of cognitive change.