Abstracts

Rationale: The clock-drawing test (CDT) is a commonly used tool when screening for dementia, it is easily administered and can offer the examiner a wealth of information. While it has been extensively validated in the evaluation for dementia, less is known about the use and scoring of the test in patients with epilepsy. Cognitive ability declines with increasing seizure frequency, particularly in older patients. The aim of this study is to examine whether or not certain errors in the CDT occur in higher incidence in patients with dementia, independent of a diagnosis of epilepsy.

Methods: A Mini-Cog was done on elderly patients presenting for routine visits in the Rush Epilepsy clinic for the elderly. The CDT was assessed and scored according to the error types previously described in the literature for dementia screening. The scorer was blinded to the clinical characteristics at the time of scoring, including diagnosis of dementia and epilepsy type. Patients with confirmed epilepsy were included, they were classified based on type of epilepsy and if they had other common comorbidities associated with epilepsy in older individuals, such as dementia, stroke and brain tumors. A chi-square and Fisher’s Exact Test was performed to evaluate for any significant differences in the rate of each error type based on the type of epilepsy. Errors were classified into three types: those involving the clock hands, numbers, and spatial orientation.

Results: A total of 97 patients (53 female) evaluated for epilepsy between January 2018 and May 2022 were included. Errors involving the clock hands were the most common error type among all patients, specifically hand length, followed by hands not being centered. Older age correlated with higher error rates, r=0.258, p=0.01. Patients with dementia (n=24) had a significantly higher incidence of all error types, and were much more likely to make more than three errors (93% vs. 26%, p< 0.0001). Patients with stroke (n=18) were more likely to make spatial errors (69% vs. 49%, p< 0.0001) followed by number errors (52% vs. 25%, p< 0.0001). In patients with epilepsy only (n=51), they were much less likely to make errors involving the numbers (7.4% vs. 54%, p< 0.0001) and somewhat less likely to make errors involving spatial orientation (42% vs. 65%, p< 0.0001), but were just as likely to make errors involving the clock hands (82% vs. 86%). Patients with brain tumors performed similarly, however they were the least likely to make hand errors (61% vs. 87%, p< 0.0001).