Neuropsychological function among cardiac surgery patients: the feasibility of virtual assessment

The purpose of the present study is to evaluate the use of an abbreviated version of the Creyos Research battery and assess the feasibility of using this remote method to measure neuropsychological performance among cardiac surgery patients before and after their operation. We conducted a pilot and feasibility observational study of patients ≥ 18 years old who were scheduled to undergo open heart surgery at two tertiary care hospitals in London and Hamilton, Canada between January 2022 and November 2024. Feasibility outcomes included estimations of digital testing usability, recruitment, and retention rates. Remote cognitive assessment was conducted via the Creyos Research battery, a digital cognitive assessment platform consisting of 12 tasks that collectively assess multiple domains of cognition including reasoning ability, verbal processing skills and memory. Patients also completed demographic, psychological and clinical health questionnaires pre- and post-operatively. Remote administration of cognitive testing was feasible; 94% (95% CI, 85%, 98%) of participants reported no barriers to digital assessment. Among 32 consented patients, 26 completed the cognitive tests before surgery and 15 participated 1-month after their operation (58% retention rate; 95% CI, 37%, 77%), indicating that participant attrition related to loss to follow-up was a limitation. Relative to healthy norms, cardiac patients demonstrated impaired performance on the Creyos battery both before and after surgery, specifically on tasks that relied on verbal processing skills. We found that performance calculated by the abbreviated Creyos-6 battery was strongly correlated with performance calculated using the entire 12-task battery for all three cognitive domains (correlations ranging from r = 0.84 to r = 0.92). Use of web-based patient-administered cognitive testing is feasible in a cardiac patient cohort; however, a small sample size and challenges with participant retention limit generalizability. These findings can inform future research that aims to implement the assessment in large-scale interventional trials to identify domain-specific cognitive impairment and to longitudinally monitor cognitive recovery.