Let Patients Agree to be Subjects
If information collected by researchers could be used to identify individuals, then they become not just patients, but also research subjects. In my view, researchers should have to obtain these subjects’ informed consent in nearly all cases. Consent is an educational process for both potential subjects and researchers. It requires researchers to explain and justify their projects to patients and the public in plain language. This is not always easy to do, but it is important. Health research is meant to benefit future patients, and the consent process helps equip researchers and potential subjects to be partners in the research enterprise. Much of the public is unaware that their medical records might be used for purposes other than their treatment, and is likewise unaware of the safeguards in place to ensure that information used in health research is kept confidential. Besides consent to specific studies, I support greater public education about health research, through pamphlets distributed to patients at academic medical centers and other research sites, as well as through public forums.
Even when health research does not involve direct intervention with subjects, it is vital to preserve the distinction between research and treatment, and between being a subject and being a patient. Contacting patients to inform them that researchers wish to use data about them, and thus to make them subjects, helps to reassure patients that information about them is treated respectfully. It also gives them the opportunity to make their own choices about how, and for what purposes, information that could identify them may be used. The individual’s informed choice is a central tenet of bioethics.
A few studies have shown that a high percentage of patients, when asked whether they would permit their records or samples to be used for health research, say yes. Of course, these people were asked. If they were asked a different question—”Would it be okay with you if health researchers used your medical records or blood or tissue samples, in ways that could identify you, without asking you or letting you know?”— at least some might answer differently.
Even if many people would agree, if asked, I still think it’s worth the time and trouble to do so. The argument that this effort is just too costly and time-consuming could be applied to every aspect of research, including study design and data security. All aspects of research deserve the attention and investment needed for sound science and good ethics.
If researchers universally sought consent for such studies, certainly some people might say no. This is not surprising, since little effort has been made to educate and inform the public about the research enterprise. If, after increased public discussion, researchers still have concerns about selection bias, then perhaps they have been unsuccessful in persuading particular potential subjects of the value of the project in question, or have not succeeded in allaying potential subjects’ confidentiality concerns. That is part of the researcher’s job, whenever the researcher uses data that could identify individual research subjects.
Balance society’s needs with patients’ rights.
As we strengthen laws to protect patient confidentiality, we must be careful not to impose unnecessary restrictions that have unintended consequences. I believe there are practical and scientific justifications for conducting certain kinds of research without direct consent, under carefully monitored conditions. These include studies where there will be no contact or interaction with the patient, and where any risks have been determined through independent review to be minimal.
As patients, we can all agree that confidentiality is important, but we also benefit from health care advances that can only be gained by analyzing medical records in the aggregate. Epidemiologic studies have yielded valuable insights on the etiology, natural history, diagnosis, and treatment of disease. Studies of this nature are most meaningful if they truly represent the population in question. If outcomes data are compromised by difficulties in gaining access or by self-selected exclusions, we are left with a biased, potentially dangerous picture. For instance, if patients with bad outcomes were less inclined to allow their records to be included, we might continue performing procedures that deserve to be halted.
The experience of the Mayo Clinic in complying with a new Minnesota state law suggests that most patients support the use of their records in medical research (NEJM, Nov. 1997). The law, effective January 1997, requires institutions to seek prospective authorization for use of records in research. To date, fewer than 4 percent of nearly 400,000 patients have objected to use of their records for future unspecified research. It has been argued that agreement by the overwhelming majority (96 percent) of potential subjects supports the concept of “constructed consent” from the general population, in which consent is so broad as to be inferred for all.
What about those 4 percent who object? This may be another instance where our democracy aims to best serve the common denominator. Federal regulations have established a system of research ethics committees or Institutional Review Boards (IRBs) that protect the interests of all subjects by reviewing studies on a case-by-case basis. Informed consent is a principle that we who serve to protect subjects take very seriously. But IRBs should be called upon to balance the needs of society against concerns for individual privacy by determining when consent is truly necessary and meaningful.
Seeking consent for such studies can also be costly and, when thousands of records are involved, impracticable. For example, Mayo spent over $1 million in the first year alone to solicit authorization, follow up with non-responders, and establish tracking systems to comply with the new legislation. We need to ask if we, as a society, are willing to spend a large piece of scarce research funding on these types of efforts, if it means diverting those dollars from actual research.
We accept that patients at a teaching hospital will receive some of their care from physicians still in training. Should this acceptance be extended to scientists working in pursuit of knowledge, another mission of that same academic center? Education is a central component of informed consent, and we must do a better job of educating patients, both to the importance of research utilizing records and to the potential use of their records. Patients should understand that their identities are never revealed in studies like this and that access remains limited to agents of the institution, bound by the same promise of confidentiality as their immediate care giver and subject to peer review. I am not naïve enough to think that abuses could never occur, but biomedical scientists investigating legitimate questions are not a likely source.
Society has a vested interest in the cost-effective advances that medical records research has provided, and can continue to provide. But we should also ask ourselves a very personal question. If we learn tomorrow that we have breast or prostate cancer, or heart disease, or diabetes, do we want our provider’s approach supported by the most accurate data available, or are we willing to accept the limitations that unwarranted restrictions would impose?
Catherine House was formerly a staff contributor for Endeavors.
Nancy M.P. King is associate professor of social medicine, cochair of the UNC Hospitals Ethics Committee, and a member of the IRB.