Medication Profiles of Patients with Cognitive Impairment and High Anticholinergic Burden

Abstract

Background: Drugs with anticholinergic properties are considered potentially inappropriate in patients with cognitive impairment because harms-including delirium, falls, and fractures-may outweigh benefits.

Objective: To highlight opportunities to improve clinical decision making and care for patients with cognitive impairment and multiple chronic conditions, we identified distinct subgroups of patients with mild cognitive impairment (MCI) and dementia who had high cumulative anticholinergic burden and specific patterns of anticholinergic use.

Patients and methods: We conducted a retrospective cohort study in a not-for-profit, integrated delivery system. Participants included community-dwelling adults aged 65 years and older (n = 13,627) with MCI or dementia and at least two other chronic diseases. We calculated the Anticholinergic Cognitive Burden (ACB) score for each participant from pharmacy and electronic health record (EHR) data. Among individuals with a mean 12-month ACB score ≥ 2, we used agglomerative hierarchical clustering to identify groups or clusters of individuals with similar anticholinergic prescription patterns.

Results: Twenty-four percent (3257 participants) had high anticholinergic burden, defined as an ACB score ≥ 2. Clinically meaningful clusters based upon anchoring medications or drug classes included a cluster of cardiovascular medications (n = 1497; 46%); two clusters of antidepressant medications (n = 633; 20%); and a cluster based on use of bladder antimuscarinics (n = 431; 13%). Several clusters comprised multiple central nervous system (CNS)-active drugs.

Conclusions: Cardiovascular and CNS-active medications comprise a substantial portion of anticholinergic burden in people with cognitive impairment and multiple chronic conditions. Antidepressants were highly prevalent. Clinical profiles elucidated by these clusters of anticholinergic medications can inform targeted approaches to care.

Figure 1. Timeline for study

We assessed anticholinergic drug exposure in the 12 months prior to October 31, 2016 in a cohort of individuals who had a claims-based diagnosis for dementia or MCI, as well as 2 or more additional chronic medical conditions out of 185, anytime between May 1, 2014 and May 1, 2016. Medication use was only assessed following the documented diagnosis. This time frame ensured that all cohort members had a minimum of 6 months to experience medication exposure and a 24-month period to capture a claims-based diagnosis.

Figure 2. Overview of cohort selection

^a KPCO = Kaiser Permanente Colorado