Abstract T P311: Prospective Assessment of Medical Decision-Making Capacity in Stroke Patients using a Rapid Standardized Questionnaire: Can My Patient Consent for Treatment or Clinical Trial?

Abstract

Background and Objective: Stroke deficits frequently alter patient medical decision-making capacity ( MDC ) resulting in lost trial recruitment and reducing validity of qualitative outcome measures. Since no standardized tool exists for MDC evaluation in stroke, we tested a validated standardized questionnaire used for medical patients, the Aid to Capacity Evaluation ( ACE ), vs. independent clinician assessment in mild-to-moderate severity stroke patients. We hypothesized that the ACE would show similar agreement with clinicians and therefore be appropriate for rapid bedside screening for MDC. Methods: Ischemic or hemorrhagic stroke patients underwent 3 independent capacity assessments by a medical student (ACE), psychiatrist ( PS ) and neuropsychologist ( NP ). Inter-rater reliability was assessed using intraclass correlation ( ICC ) and Cohen’s kappa. Assuming the clinician as the gold-standard, we tested sensitivity and specificity vs. ACE. Results: All planned 30 patients (90% ischemic; mean age 67.8; 60% male; median NIHSS = 6) were prospectively enrolled between 7/13- 8/13. The median time from stroke onset to first capacity assessment was 3.3 days. 11 (37%) had aphasia and/or neglect and 38% had left hemispheric stroke (see table). ACE agreed with PS and NP in 59% (kappa 0.293; 95% CI 0.08-0.51) and 76% (kappa 0.494; 95% CI 0.19-0.80) of cases, respectively. Despite low sensitivity and NPV, specificity and PPV of ACE vs. clinicians ranged 88-100%; only classifying 1 patient capable when clinicians scored incapable. ICC among all raters was 0.474 (95% CI 0.25-0.68). Conclusions: There was fair overall agreement between a standardized questionnaire and expert clinicians. The ACE was highly specific in identifying mild-to-moderate severity stroke patients who lacked MDC. The ACE might be a useful screening tool to determine lack of capacity in stroke patients, but low sensitivity for identifying presence of capacity warrants caution and further study.