How Informed Is Informed Consent?

Abstract 2556

The informed consent process is a critical component of human subject protection in biomedical research, with the goals of informing participants of the purpose of the study, as well as the likely risks, benefits and alternatives. The signed informed consent document (ICD) is a required legal disclosure which documents that the informed consent process has taken place and provides research subjects with comprehensive information about their role in a study. Despite efforts to optimize the ICD, only limited data are available regarding the utilization of these documents by participants in biomedical research. We measured the time taken by participants, in a genetic study of hematologic traits, to review a 2833 word online ICD prior to indicating consent. ICDs were generated utilizing standard templates provided by the University of Michigan IRB with the addition of a hyperlink at word 2254 of the ICD that read “If you are reading this form, please click on this sentence”. Identification of the hyperlink, as a proxy for thorough reading of the ICD, was recorded. The study was approved by the University of Michigan IRB (IRBMED# 2005-0080.) A total of 1209 subjects were recruited from the University of Michigan, Ann Arbor student population from 2/12/08 to 1/30/09. Age ranged from 14 to 35 years with a mean of 21. Standard reading speeds predicted a range of 567 to 850 seconds to read the full ICD text. The distribution of consent times was heterogeneous, heavily weighted toward times that would suggest little to no reading of the ICD. Twenty-eight percent of participants consented within ten seconds. Only 78 participants (6.4%) took longer than the minimum estimated reading time (566 seconds) to indicate consent. The hyperlink was identified by 2.2% of participants with a median reading time of 621 seconds, significantly longer than those who did not (52 seconds, p<3×10⁻¹¹). Additionally, significant differences in consent time were noted for sex, age, and downloading of the ICD PDF file. Our results demonstrate that the majority of participants in this study (93.6%) provided consent without spending sufficient time to thoroughly read and comprehend the ICD. The 6.4% of participants with consent times greater than the minimum predicted ICD reading times is likely a significant overestimate of ICD comprehension since the consent interval recorded could also include time spent on other distracting activities. In conclusion, these observations imply that the consent by subjects to participate in this and many other low-risk studies is unlikely to have been as truly informed as originally intended by the investigators and the IRB. This lack of truly informed consent is also likely to extend beyond research studies, to include ICDs used for treatment in the routine clinical care setting. These data suggest that current ICDs, particularly for low-risk studies, may no longer serve the intended informed consent purpose and that ICD length and complexity should be reassessed.