Context: ICU patients are often rendered incapable of making decisions as a result of their illness. The accuracy with which patients' surrogates consent to research on their behalf is not known. Objective: To determine if surrogate decision makers provide accurate consent for intensive care research. Design: Cross-sectional, paired, face-to-face interviews. Setting: A large, managed-care, cardiac surgery service. Patients and participants: One hundred elective cardiac surgery patients and their self-appointed surrogates were enrolled. Intervention: Patients agreed or declined to provide informed consent to two hypothetical research trials. One trial represented minimal risk to those enrolled; the other trial represented greater-than-minimal risk. Surrogates attempted to predict the patients' responses. Main outcome measures: The accuracy of surrogate consent was analyzed in a fashion analogous to the evaluation of a diagnostic test. Predictors of accuracy were evaluated using multiple logistic regression. Results: Overall surrogate positive predictive value for the low-risk study was 84.0% and for the high-risk study was 79.7% (p = 0.72, McNemar test). Predictors of accurate consent were not consistent across the two studies. Conclusions: Surrogate decision makers for critical-care research resulted in false-positive consent rates of 16 to 20.3%. Further assessment and evaluation of the practice of surrogate consent for intensive care research is, therefore, recommended.
