Abstract 70: Importance of Both Self-consent and Proxy- consent Enrollments to Avoid Bias in Acute Stroke Randomized Trials
Background: Informed consent options for acute stroke randomized trials include self-consent from patients with retained competency, proxy consent from a legally authorized representative (LAR), and exception from informed consent in emergency circumstances (EFIC). Mechanisms other than self-consent trespass upon the ethical principle of autonomy (respect for persons), and should be used only if self-consent alone would bias the patient population.
Methods: We analyzed acute stroke patients enrolled within 2h of symptom onset in the NIH Phase 3 Field Administration of Stroke Therapy - Magnesium (FAST-MAG) trial, comparing entry characteristics and outcomes with different consent mechanisms.
Results: Among 1700 patients, median time from onset to study entry was 45 min (IQR 35-62). Overall, 60% of patients were competent and provided self-consent, 39% noncompetent and enrolled by proxy consent from an on-scene LAR, and 1% noncompetent enrolled using EFIC/delayed consent. Compared to self-consented patients, proxy-consented patients were: older, 73.8 vs 66.5, p<0.001; more often female, 47% vs 40%, p<0.02; and a higher frequency of several vascular risk factors, including hypertension, 81% vs 76%, p<0.05, diabetes, 26% vs20%, p<0.01, atrial fibrillation, 29% vs 17%, p<0.001, and coronary artery disease, 24% vs 19%, p<0.01. Proxy consent patients had more severe initial stroke deficits, prehospital LAMS, 4.1 vs 3.5, p<0.001. Proxy consent patients with acute cerebral ischemia had more advanced initial ischemic stroke lesions, ASPECTS 8.3 vs 9.3, p<0.001, and received tPA more often, 35% vs 20%, p<0.001. Proxy consent patients with intracerebral hemorrhage patient had larger initial hemorrhage volume (cc), 42.2 vs 22.6, p<0.001. Proxy consent patients had worse final outcomes, with freedom from disability (mRS 0-1) at 90d, 23.6% vs 44.3%, p<0.001.
Conclusions: Proxy-informed consent patients differ systematically from self-consent patients, with older age, more comorbidities, more severe deficits, and worse long term outcomes. To ensure that acute stroke trial cohorts are representative of the general stroke population, it is important that enrollment mechanisms include proxy and/or deferred consent.
Author Disclosures: K. Shkirkova: None. N. Wu: None. S. Starkman: None. N. Sanossain: None. S. Hamilton: None. D. Liebeskind: None. M. Eckstein: None. S. Stratton: None. F. Pratt: None. L. Sharma: None. L. Restrepo: None. M. Valdes-Sueiras: None. M. Kim-Tenser: None. R. Conwit: None. J.L. Saver: None.
- © 2017 by American Heart Association, Inc.