Abstract
Attaining broad consent from individuals who participate in biobanks is the current norm (Steinsbekk et al., 2013). It is difficult for biobanks to achieve fully informed consent because the nature of future research is often unknown or widely variable (Kaye et al., 2011; Kaye et al., 2012). Therefore, until recently, it appeared that broad consent, in which donors consent to a range of experiments before participating in a biobank (Hofmann, 2009), was the only solution. Broad consent assumes that there is minimal risk to an individual donor and therefore that the benefit to research as a whole outweighs the risk an individual incurs by donating samples. Some argue that broad consent is ethical on the basis that consent can be revoked at any time, and in the event protocol changes, an ethics review board will make an informed decision of whether participants must be reconsented. Ethics review boards can decide, for example, that the use of de-identified information is sufficient to allow sample use without reconsent from the participant (Kaye et al., 2011).
The argument for broad consent is sprinkled with a priori assumptions. First, the notion that there is minimal risk to the individual is no longer true in an age where even de-identified information can be traced back to an individual using public databases (Gymrek et al., 2013). Depositing a sample in a biobank does not have the same risks as physically participating in a clinical trial, but the “informational” risk is still of concern and can have unknown future adverse effects. Second, even if the option to revoke consent is available, donors cannot possibly revoke consent if they are not aware that their data are being used or for what they are used. Information sharing between biobanks and participants can somewhat alleviate this problem, but in cases where data or samples are already distributed, of course the individual cannot withdraw from that analysis. Third, ethics review boards are not always fully equipped to assess the safety of an individual's information because of the unknown downstream affects of open, individual genetic information (Hofmann, 2009).
The aforementioned pitfalls of broad consent could justifiably be overlooked in a previous era of limited communication between participants and biobank custodians, an era in which reconsenting for new research purposes would be costly and difficult, and would hinder the advancement of said research (Hofmann, 2009). However, in a highly digitized age with social media, electronic medical records, and advances in health information technology, it is time to reevaluate the process of broad consent and look to the paradigm of dynamic consent.
United Kingdom academics and companies are testing dynamic consent as part of the EnCoRe (Ensuring Consent and Revocation) project (Wee, 2013). Dynamic consent is based on a patient-centric information technology system that allows individuals to set specific preferences for the use of their data. The system also allows participants to modify, change, or revoke their preferences over time. Participants can also track and audit changes made to their privacy settings. In addition, participants are able to choose when and how they would prefer to be contacted for future studies or for the reporting of results (Wee, 2013).
The dynamic consent model is oriented around the participant and involves ongoing communication between the biobank, donor, and researchers. Ongoing communication fosters a culture of respect between participants and researchers and allows for more streamlined recruitment, consent, and reconsent. If donors feel a sense of involvement with the research and are treated as active, equal partners in the process, the recruitment of participants will be strengthened because of increases in confidence and trust in the research and researchers. A patient-oriented system of consent could therefore have the effect of decreasing costs of recruitment into trials (Kaye et al., 2012; Kuehn, 2013).
Broad consent practices usually result in the de-identification of specimens. There are at least three issues with de-identification. First, de-identification, while usually offered as a means to protect participants most fully, actually cuts the individuals off from being full participants. Next, while a single snapshot in time is useful, it is not as scientifically useful as is the longitudinal association of a sample and the emerging phenotypes. For example, a breast cancer tumor sample might be taken at the time of surgery, and de-identifying the sample does not allow the researcher access to critical information, such as whether the cancer recurred, the individual's survival rate, and what other medical or nonmedical issues should be considered. Finally, even the accuracy of the term “de-identified” has become increasingly scrutinized. Several scholars have shown that re-identification is possible with two or more datasets (Gymrek et al., 2013; Malin et al., 2001).
In the dynamic consent model, anonymity is not necessary since ongoing communication allows donors to directly consent for further research and or specific types of research (Kaye et al., 2012; Terry et al., 2013). Because dynamic consent calls for specific sharing preferences to be set before research, the need to revoke consent will lessen since the types of research allowed are clearly delineated (although the option to revoke consent is still present). Dynamic consent also diminishes the work of external ethics review boards because the individual can directly decide to consent to further research.
The patient-oriented tenets of the dynamic consent approach foster an environment of respect and trust in research. In combination with ongoing contact and communication, trial enrollment can increase. Dynamic consent maintains open lines of communication between biobanks, researchers, and donors, thus allowing for trial results to be reported back to the participant, a major weakness of current consent paradigms (Kaye, 2012). Including the preferences of the individual in the dynamic system can mitigate even the ethical dilemmas of reporting incidental findings.
A current limitation of most dynamic consent systems is the potential of isolating certain disadvantaged groups with limited access to technology (Steinsbekk et al., 2013). The use of digital components also minimizes the face-to-face contact between researchers and participants, although if a goal of the research system is to vastly increase the number of participants in biomedical research, then some technology tools for scaling must be implemented. The aforementioned shortcomings, however, can be mediated with further policy interventions and the use of more ubiquitous technologies such as text messaging and e-mail. Overall, a shift from a researcher-oriented perspective to a patient-centered view can undoubtedly benefit researchers, donors, and genomic-based translational research as a whole.