CTSI Blogs

The Research Ethics Blog by Bernard Lo, MD

A Biobank for Genomics Research: Do we need Patient Consent?

To facilitate clinical research, Vanderbilt University has constructed a biobank that contains DNA specimens derived from leftover blood remaining after clinical tests have been carried out, together with information extracted from the electronic medical record.  The blood would be discarded if not used for research.  All information is de-identified by the biobank before it is provided to researchers. See article.

The Vanderbilt IRB has determined that because researchers receive only de-identified information, specific projects carried out with biobank data and materials are not human subjects research, and therefore no consent is required.  Researchers who use the biobank must sign a data use agreement promising not to try to re-identify patients.  As an additional safeguard, patients may opt-out of having their blood and data placed in the biobank; about 5% of patients have done so.  Provisions to opt-out of the biobank are contained in a separate section of the Consent to Treatment / Agreement to Pay form that patients sign every year.  The program has been publicized through posters at the medical center, information pamphlets given to patients, and newspaper advertisements. 

Questions:

  • Do you believe that the arrangements for this biobank are ethically acceptable?  In other words, should the federal regulations permit such arrangements? 
  • In particular, should research that involves sequencing of an entire genome be permitted on leftover blood without active, explicit consent from participants? 

Comments

I think the arrangements are perfectly ethically acceptable since the confidentiality of the subjects is well respected by the fact that the information from patients medical records as well as the left over specimens are deidentified and also the patients have an option of opting out which after adquate information which is displayed publically hence adhered to respect of persons principle. However, i feel that the consent form for the study should not really be tanged to the treatment/agreement to pay forms because it compromises the patients ability to freely accept to participate in study without having a feeling that his/her treatment is being compromised. Research for coding entire genome I think should involve consent because a persons genome is almost an identifier which can be used to re-identify the person and hence break confidentiality which is key in the respect of persons ethical principle. This is is on the background that for the authorities this would be a good tool to try and crack down on law breakers.

I found it interesting and laudable that the researchers conducted a community survey to get a sense of their patients' opinions about using leftover specimens for DNA research. A significant minority (5%) of patients were opposed to having their DNA banked, and I feel that it is important to respect these patients' autonomy. If a plan for use of specimens in research is already in place at the time of collection, it is feasible and appropriate to inform patients and allow them to refuse to participate. In low-risk projects such as this, an opt-out approach seems reasonable for cost-containment and accessibility. But I disagree with the Vanderbilt IRB's assertion that patients do not need to be informed about this project or have a right to choose whether to participate. In addition, having a mechanism for consent protects the project from lawsuits, which could prevent the project from attaining its objectives after huge amounts of resources have been invested.

De-identified specimens are human subjects and consent should be required. Simply restraining researchers from re-identifying tissue/specimen donors or blocking the sequencing of the DNA will be insufficient in being ethical. With recent advances in stem cell technology and replication science, replicated organs or even cloned individuals can be produced from these de-identified tissue/specimens. The ramifications are unfathomable.

As for my views on DNA sequencing...the idea of it may appear frightening, but prohibiting it does nothing, and yet imposes inconveniences for scientists. To match DNA, current practice more often uses single nucleotide polymorphisms (SNP’s) as opposed to sequencing. Sequencing of the DNA is actually rarely done scientifically and when used, only to provide pertinent information regarding the specimen. Of course, sequencing methods are still improving and will be done more often as it unravels maybe too much information about the specimen. Understandably, this amount of information stirs fear. However, the fear associated with DNA sequencing should be dealt with by education regarding DNA technology, dealing with this issue with the explicit rule prohibiting DNA sequencing poses an obstacle for those who are scientists and patronizes those who are not.

Vanderbilt University’s use of leftover blood from clinical specimens to create a biobank is analogous to the use of neonatal blood spots for research. In the latter case, Texas law deemed that an “opt-out” passive consent process was adequate. Vanderbilt has taken a similar approach, adding an option to opt-out of the biobank in the Consent to Treatment/Agreement to Pay form that patients sign every year. In both circumstances, it was argued the use of these clinical samples for research met ethical standards for no consent because they were deemed to be adequately de-identified and there was no physical risk to the subject. However there are two aspects of both the neonatal blood spot research and the biobank project that raise concerns: 1) Is the subject truly be “de-identified” if his entire genome is sequenced? An entire genome is as much an identifier as a birth date, social security number or fingerprint, as it is unique to the individual in question. 2) Some participants may be agreeable to having their genetic material used for certain research projects but not others. Vanderbilt has taken some steps to take both of these concerns into account. For instance searches of the de-identified data base that “return very small data sets (currently

I agree with Stephanie's concern that through genome sequencing it is possible to uniquely identify individuals. This is the technology used to rule out paternity. From an IRB perspective, what if future researchers use the genome to identify certain races and ethnicities, as was done in controversial substance abuse research involving indigenous Americans. How about if the genome data is linked to the integrated data repository as is done in the Partners Group in Boston? It may be a mistake to assume that patients won't be reidentified by removing HIPAA identifiers. What if a future researcher finds a disease association for a gene that is enriched in individuals with Norwegian, Korean and Somali ancestry?

The utility of stored serum for future testing is invaluable. We obviously can not predict which tests will be run in the future, as new study questions arise. However, it is important to summarize the types of testing that may be performed, ie testing for infectious diseases, testing for markers of disease, performing genetic studies, etc, in order to provide an overview to patients of the types of study questions that their serum may help to address. This case is reminiscent of the case we discussed in class in which stored cells were used to create embryonic stem cells with the ultimate goal for women post radiation to potentially procreate. Donating patients had no a priori knowledge that a study of this nature would be conducted. Although their tissue was de-identified, we enter a realm in which patients’ fundamental religious or spiritual beliefs may oppose “non-natural” methods of child bearing. This issue is pertinent to the Vanderbilt scenario, as genomic sequencing may violate the beliefs of the serum donors. Vanderbilt has taken appropriate measures in consenting patients for the use of their serum in future studies, although they provide no information to patients on the genres of studies that may be conducted. As highlighted above, providing a framework to patients of the types of studies that may be conducted is critical, as many studies now enter a realm of very personal religious and ethical questions. Sequencing of patients’ genomes is a very heated topic in the world of medical ethics, especially in the heart of Tennessee where the general public’s opinion of genomic research is quite conservative in nature. One way to address this issue would be for the institution’s ethical committee to delineate the types of studies that meet criteria for ethical controversy, and address these in their IRB. This process would ultimately require specific consent for use of any patient tissue in studies meeting these pre-defined criteria. This process would undoubtedly lead to a higher than 5% opt out rate, though would allow these studies to be conducted in the most ethical light. Posted by Monika Sarkar for Responsible Conduct of Research course

The applicability of stored serum for future testing is essential for helping us understand what we currently do not know. It is impossible for us to determine which test methods will be utilized and relevant for future studies, but notifying future patients of the different clinical applications their serum can accomplish can help patients be more open to storing serum. The stored serum can be tested for infectious diseases, have the serum act as biological markers of disease, etc. Another relevant purpose of storing serum, which was mentioned in a clinical case during lecture, was to create embryonic stem cells for women post-radiation, so that they can give birth to children. Even though previous donors were de-identified, the religious and political beliefs in the area of Vanderbilt are more conservative from the rest of the nation, since they view certain uses of science in a negative light. As stated in the scenario, Vanderbilt has acted on improving consent of donors by informing the donors that their serum will be used in future studies, but without actually indicating their specific purposes. One way to deal with this ethically is to inform the patients of what their serum might be applied to in a certain category and allow them to decide if donating their serum is something they would want to do. The opt-out rate would be higher, but the approach would definitely be more ethical and it gives the patient the option to choose to participate in the study or not.

The Vanderbilt University IRB was correct to conclude that the biobank materials are not equivalent to human subjects in their de-identified state. By requiring researchers to refrain from any attempts to re-establish the link between samples and patients, the biobank successfully demonstrates a high level of commitment to the protection of public health information. Furthermore, the biobank offers patients the opportunity to opt out of the program. Overall, the biobank's framework is an ethically acceptable one. However, these arrangements are not perfect. First of all, the provision to opt-out of the program is in a separate section of the consent for treatment. It makes sense that the average patient might miss this entirely, especially if biased research assistants subconsciously (or consciously) fail to point it out. Furthermore, for those patients who see the form and choose to allow the use of their materials, it is very unlikely that they understand the full impact of a discovery made using their samples. They will likely appreciate the fact that their samples are de-identified until further genetic studies reveal something that could serve as a predictor for future health outcomes. In this case the patients may ultimately feel that the researchers are withholding valuable information, rather than rejoicing at the discovery that they contributed to. In this way, the provision to prevent re-identification of patients may actually be harmful. Since the option to opt-out does not seem obvious and since it is rather unlikely that the research assistants collecting consents take the time to spin out a scenario such as the one described here, it does not seem that the federal government SHOULD permit the biobank's proposed arrangement without first demanding a more complete and careful consenting process. Lastly, the biobank SHOULD seek consent if the leftover blood is used to sequence an entire genome. Under HIPAA, if fingerprints count as personal identifiers then a fully sequenced genome must also count as a personal identifier. This dilemma is analogous to the case of the newborn screen blood spots that contain potential for full genomic identification. If parents in Texas expressed significant concern about this possibility, then it stands to reason that someone somewhere would not approve of the biobank's methods if it were to fail to obtain full consent for all possible full genome sequencing. may elicit feelings of patients are able to opt out of the biobank program, increases with potential compromise to the research. Without such a requirement, there is the possibility that the researchers' findings could be linked to an individual's risks or exposures. essentially eliminates access to information on predictor variables or risk factors that might have pointed them towards their discoveries more readily. :

After reading the article and considering the information provided about the ethical regulations in this project, to me it seems that the arrangements for conducting projects derived from this biobank are ethically acceptable as the OHRP state this kind of procedure has to be managed. Although the article stated that the study met all of the IRB requirements in terms of the guidance for human subject protection I think my main concern is that the patients are well informed (Respect for Persons principle) when considering the “opting-out” alternative. I am concerned about the fact that during the process of gathering information of patients through surveys about their perceptions of creating the DNA databank, it showed that only a minority of people could recollect the information provided (in posters at the medical center). It was after someone read information to them about it, that they were “comfortable with the idea of a DNA databank concept” (p. 363). I ask myself to what extent people can really understand the potential consequences (positive & negative) that providing DNA information could pose on them in the future, when this information is presented as a section inside of another form intended to Consent to Treatment and Agreement to Pay. I worry that it could make patients feel compelling to participate if they are confused or perceive there is a connection with receiving treatment and being willing to participate of the biobank protocol. Therefore, I think the consent form for receiving treatment should be separate from the one to participate in the biobank. There should be explicit and exhaustive information of potential benefits and risks of providing information or material (blood samples in this particular case) that could be used on secondary research (which is not the one you’re signing up for). I think that each study that uses data that can potentially lead to trace and identify a person should be requested to ask for an informed consent.

I do believe that the arrangement made for this biobank are ethically acceptable. Given the fact that the samples are completely de-identified before they are given to the researchers and the researchers are not allowed to attempt to identify any patients from the samples, this does protect that confidentiality of the patient who provided the sample. Secondly, the patients are given the opportunity to opt out of the program should they find it unethical in any way. Finally, there is no question that the samples which would otherwise be wasted can contribute to many future research projects ad is a great resource. It also appears that the university has done some advertising via posters and information pamphlets to patients to make them aware of the potential to utilize and unused sample for any research purposes. However, the issue has been raised by others comments about the use of this blood for whole genome sequencing which does then allow a personal identification of patients and their potential diseases or diseases that they might be at risk for. As the patients do not know what research their sample may be used for it is more difficult for them to make an informed decision to opt out or allow their samples for any potential research program. I think that some amends to consenting process should be added so that patients can understand that their samples may be used for genetic testing which may reveal personal information about them and potential diseases they may be at risk for. Given this situation they may be better informed to decide whether they want to stay in the program or opt out and also whether they would like to be contacted should any information pertinent to their health be revealed from the research. This comment was posted by Sanoj Punnen in epi201

Many possible sequalae exist with the current insufficiency in being ethical. If seeing a younger twin of you (or even an army of them) walking around on the streets 30 years from now is not shocking enough or sound too science-fiction-like, the idea of replicated organ from a de-identified source is chaotic. With the replication technology, a blood sample is no longer simply blood. Hair, skin, fingers, ears or even gametes may be produced from these samples. When these reproduced tissue are intentionally or non-intentionally connected to crime by either the recipient of the tissue transplant or lab technicians and researchers, the donor of the tissue will be unable to convince others of their lack of involvement because crime-solving are still so dependent on these “scientific” methods. The donor will not even know about the specimen being replicated. Allowing experimentation of de-identified tissue especially in the field of replication-related science requires either restriction or the education and modification of current idea of using DNA as a method to identify individuals, and in essence, the current practice of law. Despite my views to allow DNA sequencing (see below), I believe limitations on replication and stem cell research (and perhaps some other research fields such as proteomics or stress-response but think I’d be digging too deep to go into this) should be implemented. Amount and type of research allowed is arbitrary. The surest way to respect the donors and the specimen would be to require notification of all donors prior to initiation of a new project on any of the specimen used. But this leads to the issue of whether de-identification is necessary, since no consent could be obtained after specimens are de-identified. I think it would require a few more decades of scientific research and ethical debate to realize the whole notion of de-identification is non-sense, but in the meantime, limiting amount and type of research done or obtaining informed consent on all projects to be performed should be the least that we could do.

I do not think the Vanderbilt IRB is acting appropriately in allowing the researchers to use de-identified left over blood of patients who have had an opportunity to opt out. I have some reservations. In order for these samples to be useful, investigators will need to know some of the baseline characteristics of the patients themselves (age, sex, medications, etc). Having random blood samples without any knowledge of the patient to me would seem fruitless (what types of questions could you answer?). Thus, this brings into question the true definition of the term de-identified. These patients are presumably from the Nashville area and may possibly be identified by the patient characteristics. This is a bit extreme though, and I suppose the scientific benefit of such studies may outweigh the small risk of being identified. With regard to future technology and what this would entail...I am fearful. Technology is going to advance, particularly in the field of cloning, and I do fear that patient's samples could be used for purposes upon which the initial patient would not be supportive of. How can we even begin to understand where advances in biotechnology will take us and thus how can we assure patients if they don't opt out that their samples will not be used for unforseen purposes? I believe some international standards need to be set now to address these issues before disastrous future consequences ensue. This was posted by Vijay U. Rao of the Ethical Conduct of Research class.

In general, I think the arrangements for the biobank were ethical and conscientious in that first, a "synthetic derivative" of the EMR was derived to assure that the phenotypic data to be used would be properly de-identified; second, an opt-out revision of the informed consent process was created to give patients the choice of not participating. I think that the investigators were guided by not just be ethical practice, but by existing research done in the area of biobanking consent. In fact, as recent as 2009, Kaufman et al published in the American Journal of Human Genetics a great piece that summarizes the "Public Opinion about the Importance of Privacy in Biobank Research" - from a survey of almost 5,000 US adults, it seems that while most are concerned about privacy, the majority would allow academic researchers to study their data, and provide consent if an oversight panel had approved the research. Given these findings, I think the Vanderbilt investigators balanced the need to carry on a scientific mission, potentially for greater good, with the imperative to secure privacy and offer reasonable consent for involved patients. However, I do agree with another post that the context for consent also matters; if the biobank research is to be carried out in a potentially more conservative geographical region of the US, then existing data about moderate public opinions about biobanks might not hold. It would be helpful in this case if the Vanderbilt team conducted a localized attitudes survey before assuming that biobank/whole genome sequencing is a relatively non-controversial matter in its home state. In addition, while the opt out is a reasonable consent feature, how the actual content of the "opt-out" is framed by the informed consent may not be ideal. I do agree with previous postings that aggregating the opt out for both participation in the biobank and in the research to have genome based treatment should be separated.

Researchers at Vanderbilt University have taken reasonable precautions to ensure respect for patients and manage the use of private heath information. An active consent arrangement would provide additional regard to patients. Moreover, a comprehensive approach to educate the public, such as the one taken at the Marshfield Clinic Biobank, could further strengthen the ethical acceptability and viability of such a powerful genomics research tool. Marshfield Clinic educates the public through community events to increases awareness and conducts focus groups, which directly address public concerns thus improving common perceptions of Biobanks. Such programs encourage participation, which both creates justice through greater distribution of subjects as well as maximizes beneficence through more meaningful generalizability of research results across diverse populations. While both the opt-out and opt-in are acceptable arrangements, the greatest benefits to society, patients and researchers alike, occur when participation is eager and widespread.

As genome analysis tools become more sophisticated, the potential for identifying the actual patients from whom blood was drawn (whether purposefully or not) is increasing. Down the line, this could have serious implications for those patients (insurance, insurance for family members, ...). Eventually, this kind of thing could be tantamount to invasion of privacy. That being said, the likelihood of this happening for a given person/patient is very low, and most people would allow the use of their blood to forward science and accept the risks that are entailed. That's just the altruistic side of our nature. If a patient is going to take that risk, however, I think it is better that an opt-in (rather than opt-out) paradigm be used. A simple statement on a portion of the consent form describing ongoing analyses (and possible consequences of those analyses) of their blood sample over an indefinite period of time would suffice. Speaking for myself, I would have no problem allowing my blood to be banked and used for this type of ongoing research as long as I was told about this possibility before my blood was drawn. This was posted by Prashant Bhave of the Ethical Conduct of Research class.

This article brings up an interesting ethical dilemma. I feel the nature of this controversy revolves around the theme "Should de-identified blood samples be equivalent to the medical record in terms of obtaining informed consent?" Certainly most projects that use medical records to gather patient identification get an expedited IRB approval and do not require consent. To me, de-identified blood samples should not be included in this categorization. Stored blood sampling is common in many clinical studies and informed consent is always obtained. Vanderbilt has an easy solution. There should be an informed consent early in the hospitalization were someone from the laboratory before drawing their blood consents them for the permission to bank the rest of their sample, to be used in clinical research. This way patients themselves would be able to decide the benefits/risks, and in this scenario I believe the overwhelming majority of patients would give their consent.

One of the central questions in research ethics is whether the potential benefit – to the individual and/or society – outweighs the risks. The benefits of this project to the individual are nil. The premise of its safety is that the human tissue utilized in research cannot be connected to the person from whom it originates. The benefits to society are unknown, but presumed to be huge, although they will only accrue slowly over time. We should approach this question, then, assuming that although this is a project of great social value, because it does not benefit the subjects at all, it should essentially have no risk. First is the question of opt-in versus opt-out. The authors provide a good description of the scientific benefits of an opt-out approach: it liberates the researchers from a narrow database focused on specific, rare diseases and allows one that is more broadly representative and can address genetic associations with common diseases. More importantly, perhaps, is that the opt-out approach provides more individual protection, because it obligates extremely complete de-identification. In these aspects, I think this database is analogous to one of discarded pathologic specimens, more so than, say, retrospective analyses of the medical record, for which informed consent is typically not required. There is little to no risk that the genetic information would ever become re-identified and, even if it were, it is not clear to me that the information could be used in a way harmful to subjects (at least not given our current understanding of the genome). Another important aspect is whether patients are truly aware that they are opting-out. This is the weakest point in the protections established by the authors of the database. I think a procedure should be put in place to alert every patient to the implications of the opt-out clause each time he or she signs the form.

The Vanderbuilt University biobank has established a database of patient material that holds sensitive information. The ethical dilemma is rooted in defining "patient de-identified research" to be equivalent to "non-human subject research". This means that the traditional sense of identification is independent of genetic information. Yet, given our scientific progress that allows genetic identification to be even more specific than the traditional methods of identification, our definition of identity is fundamentally changing. For instance, if a genetic bar code is instituted as truly identifying data in the future, then the principle of patient protection in the de-identifying process has been undermined. Due to the uncertainty of the extent of genetic information and its use in the future, it is imperative that patients be informed of the studies their information will be applied to. The ability to properly inform patients of all repercussions is impossible in a blanket consent or opt out. Ideally, the nature of compiling comprehensive amounts of information as a biobank for later use would benefit from educating patients with access to opting out on a study by study basis or by continually being able to monitor how their information is being used. This would give researchers and patients a greater ability to communicate the use of sensitive material and relevant consequences/benefits the patient should be aware of. The evolution of genetic information and the relationship human beings have with it is ever changing. Therefore, in order to continue ethically acceptable arrangements, the consent process should be applied on a continual basis to reflect the evolving nature of identity.

The Vanderbilt University biobank brings up several interesting and important questions. First, does this count as human subjects research? According to the Department of Health and Human Services, human subject means “a living individual about whom an investigator (whether professional or student) conducting research obtains (1) Data through intervention or interaction with the individual, or (2) Identifiable private information”. Research is defined as “systematic investigation, including research development, testing and evaluation, designed to develop or contribute to generalizable knowledge.” Though the development of the data bank is a commendable and powerful project, the conclusion that it is not human subjects research appears flawed. As stated in the article: “For the DNA databank, the outcome of de-identification is not binary; rather, it is continuous and recognizes that biological data are so inherently rich in attributes that the re-identification potential is never zero. Accordingly, a data use agreement has been put in place that permits only approved queries and data analyses, and specifically prohibits attempts at re-identification at the risk of institutional sanctions” (Clinical Pharmacology & Therapeutics (2008); 84, 3, 362–369). Though researchers are discouraged from identifying subjects via university sanctions, this may not be enough insurance against the re-identification of subjects and loss of privacy. Given the rate at which genomic research is developing, it is difficult to anticipate the ways in which banked DNA specimens could be used in the future. Because the potential for re-identification of subjects is not zero, by definition this type of research falls under human subjects research because it involves obtaining data with identifiable patient information. Using this argument, patients should be required to give consent for the use of their DNA for such research. The researchers have made a commendable effort to survey the population to discover willingness to participate, but I have some qualms about the current consent process. First, as pointed out by several prior posters, tying consent into papers required for the patients to receive care at the medical center may cause confusion about what refusal would mean for future medical care. Second, I would like to see a survey of individuals after they have signed this consent that assessed their understanding of what type of information would be gleaned. The research gained from this powerful databank clearly has the potential to be groundbreaking, and many individuals may agree to participation. However, given the risk of loss of sensitive personal genetic information, a more active consent process may be more appropriate.

I agree with Sarah. It seems a bit of a stretch to claim that just because data has been de-identified it no longer falls under the purview of human subjects research. This is especially pertinent since, as Sarah pointed out, researchers have to sign a document in which they agree not to try and re-identify the data. For me, at the end of the day, the decision by Vanderbilt not to consider the genomic database human subjects research doesn't pass my "gut-check test". Let's say that someone entered a research lab and stole a reagent or a thumb drive with unidentified data from experiments that had been run (i.e. it is just raw data--there is no information about which samples were used or how the experiments were set up). Besides being frustrated over lost time and money, the question I have to ask myself is would I feel differently about the theft had it been my own unidentified blood sample or genomic DNA sequence? Obviously everyone will feel differently about this; however if the answer is yes, you would feel differently, then perhaps more oversight regarding these human derived samples is warranted. I commend Vanderbilt for the project they are undertaking. At the same time, when human derived samples are involved, I feel that more oversight is warranted. This isn't to say that researchers should have to go through the exact same steps required to obtain DNA samples from patients themselves; however some oversight should probably be instituted instead of the essentially complete hands-off approach Vanderbilt is taking.

I feel that the biobank arrangement is ethically acceptable because the specimens have been de-identified and patients have the opportunity to opt out of participating. The potential harm to these patients is minimal once the specimens have been de-identified. While others have expressed concern about the fact that the consent is part of the Consent to Treatment / Agreement to Pay form, I don't feel that simply putting it on another piece of paper would make it more clear that treatment is separate from consent for research, as long as this is stated explicitly on the consent form. As for genomic sequencing, I think it is a bit of a leap at this point to assume that genomic sequencing will automatically lead to cloning and replication down the line. I feel that each research project done with these samples should be properly evaluated by the appropriate IRB, but theoretical concerns about the ethicality of future potential research projects should not prevent the use and creation of the entire biobank to begin with. Overall, I feel that the potential benefits and risks to the patients who are contributing to the biobank are well equilibrated - the benefits include the benefit to society of advancing scientific research (which is to say, pretty minimal), but the risks are also pretty minimal given that the specimens have already been de-identified. In order to link genomic material with the individual, the individual would have to later on opt in on a separate occasion to being sequenced, and even in that case it is difficult to see how use of the biobank could be harmful.

A clear distinction should be made between anonymous genetic studies and studies that put sensitive information at risk. The IRB should protect from researchers trying to identify subjects, as mentioned, and should also protect from studies that inadvertently reveal identifying information. For example, studies of entire genomic sequence could in theory be used to identify individuals without other identifying information. Additional concerns could be raised if preliminary studies reveal certain samples to be of high interest. Genetic information accumulated from subsequent studies could present a greater risk than data from studies analyzed independently. The IRB should ensure that samples are used only for studies that are proposed and should establish clear guidelines about what type of genetic information is considered a threat to confidentiality.

This is a interesting quandary as since the specimens have been de-identified and a consent signed it appears that precautions to safe guard the ambiguity of the individual supplying the specimen will be . The interesting piece of this is that a “form" that is signed by those doing experiments on the banked, de-identified blood stating that no attempt to re-identify the specimen will made. I’m not really sold on the fact that a “form” will keep those with “high ambitions” from doing the later. I’d hope that this isn’t the case but we have seen this time and time again in research in certain populations mislead to believe one thing whilst something completely different was occurring. Not so long ago; the Tuskegee Syphilis experiments, the pharmaceutical experiments at the Holmesburg Prison complex and the MK-Ultra Project were lauded as “justifiable” research situations when in fact they were the opposite. Also, lets not forget the story of Henrietta Lacks and the HeLa immortal cell line.

Due to the fact that the samples will be de-identified to both individuals AND the community (this latter issue is particularly relevant given the latest conflict between the Havasupai Indian tribe and ASU) and because the participants have the opt-out option, I don't see an ethical dilemma present. In general, if the informed consent was written to include the possibility of further genes being tested (particularly genes not involved with the primary study), this covers the principle of a genome-wide association study, and no further explicit consent should be needed.

I agree with previous posts by Sarah and Kyle that these arrangements are dubious, especially under the pretense that this is "not human subjects research." The opt-out process should be more transparent and should include informed consent about the possibility that entire genome analysis will be performed. Research involving sequencing of an entire genome should not be permitted on leftover blood without active, explicit consent from participants. Opt-out consent is not sufficient especially if it is on a separate form from the usual paperwork administered at admission. True de-identification is impossible in this scenario given the linkage to clinical states and rare genetic conditions are implicitly identifiable, particularly in the range of one hospital. The guidelines could be somewhat different if the biobank were somehow accrued from across the entire nation in a de-identified, anonymous way.

I do not think the arrangements for this biobank are ethically acceptable. It provides an amazing opportunity for genetics research, but I do not think the promise of medical breakthroughs should lead us to compromise our principles of informed consent. First, an opt-out system hardly ensures informed consent. It can be very useful when enrolling large numbers of patients when the majority is expected to agree to participate. However, it is not clear in the description of the case that patients understand the system. I agree with previous comments that the option to opt-out should be made more apparent, for example, by putting it on its own form. Second, even if patients were told the samples would be de-identified, they probably received little information on the risk of re-identification. The fact that researchers must sign a data use agreement promising not to try to re-identify patients is worrisome. Finally, I wonder if Vanderbilt assessed the opt-out system, for example, by studying what percentage of patients were aware of or approved of the biobank and the opt-out system, given the posters, pamphlets, and forms. This case is similar to the story of the Havasupai tribe, in that patients are agreeing to participate when they do not know the full range of research studies to be done. However, in this case, the Vanderbilt patients are more heterogeneous in background and beliefs, and may find it more difficult to protest the system. Posted by Natalie Lui.

This case of a Vanderbilt biobank using left over patient blood and medical records for research, without preauthorized consent raises some interesting concerns. Research, such as genome sequencing has the potential to identify future diseases that patients may be at risk of and creates the ethical dilemma of ones responsibility to notify the patient. Such decisions should involve the patients given they are providing their blood and records to do such research. Therefore, I believe consent should be attained with at least a small discussion of the biobank and the potential for future involvement in research whenever patients give their blood. Any patient being involved in research should have the right to know they are being involved in research and how their confidentiality will be upheld. Without this discussion, we potentially put patients security and privacy at risk without their knowledge.