Data-sharing dilemmas: allowing pharmaceutical company access to research data.
Subject: Medical research
Medicine, Experimental
Pharmaceutical industry
Authors: Anderson, James R.
Schonfeld, Toby L.
Pub Date: 05/01/2009
Publication: Name: IRB: Ethics & Human Research Publisher: Hastings Center Audience: Academic Format: Magazine/Journal Subject: Health Copyright: COPYRIGHT 2009 Hastings Center ISSN: 0193-7758
Issue: Date: May-June, 2009 Source Volume: 31 Source Issue: 3
Product: Product Code: 8000200 Medical Research; 9105220 Health Research Programs; 8000240 Epilepsy & Muscle Disease R&D NAICS Code: 54171 Research and Development in the Physical, Engineering, and Life Sciences; 92312 Administration of Public Health Programs SIC Code: 2833 Medicinals and botanicals; 2834 Pharmaceutical preparations
Accession Number: 220469490
Full Text: Dilemmas sometimes arise in interactions between the organizations that conduct clinical research and the pharmaceutical companies that sponsor the trials and are interested in study participants' information. These dilemmas can occur when new uses for research data to which participants have not specifically consented emerge after the study is complete, or they may be the result of a compromised consent process or form. Solving them is important because allowing pharmaceutical companies access to this data can dramatically improve their understanding of how certain drugs work--which will, in turn, lead to better treatment. We contend that in some cases, it is ethically acceptable to share data with pharmaceutical companies by appealing to the principles of beneficence and respect for persons. These principles require us to look not just at how well the process and documentation of consent conform to regulatory requirements, but also at how well they minimize harm and maximize benefit for research participants and future beneficiaries of the research.

The principle of respect for persons in research entails enabling the potential subject to decide on research participation with as much information as is available at the start of the study. This includes (but is not limited to) information about the risks and benefits of participation, alternatives to participation, and what will be done with the research data. This latter point involves providing details about the individuals and organizations that will have access to the subject's information both during and after the study. By including this information in the consent process, potential subjects can decide whether the release of their data in the ways specified is acceptable to them.

However, sometimes identifying in advance all the ways that research data may be used is just not possible. In cases where investigators or sponsors want to use data in ways other than those specified in the consent form, an issue of authorization arises: can data legitimately be used in ways that subjects have not specifically authorized?

We were recently presented with just such a challenge involving a clinical trial assessing a new regimen for the treatment of high-risk leukemia. The study evaluated a new combination chemotherapy regimen: subjects would have agent X added to the therapy if and only if their tumors displayed a particular genetic defect, defect A. The company providing agent X realized at the outset that a positive outcome of the trial might result in an application to the Food and Drug Administration (FDA) for expanded indication approval for agent X. Since the company was only interested in data on those subjects receiving agent X as part of their chemotherapy regimen, the consent form limited the company's access to data from only those trial subjects. Thus, the consent form indicated that "Representatives of [the company], which supplied [agent X], will have access to data only for patients that are to receive [agent X] on [study]."

Agent X indeed appeared to improve outcomes for patients with leukemia and defect A. However, the company wished to include information in its FDA application regarding any new or increased incidence of known adverse events by comparing the adverse event experience of trial participants receiving agent X to those who received backbone therapy alone. Yet consent form language provided the company access only to data from the participants scheduled to receive agent X.

Because there was no institutional review board (IRB) approval to do so, it appeared that information from patients receiving backbone therapy alone could not be provided to the company. However, given that the request for data for the purpose of adverse event comparison is in the best interest of all trial participants, we considered ways in which these data could be provided without harming participants or violating the parameters of consent.

One option was to recontact those individuals who received backbone therapy and/or their legal guardians, specifically asking for the release of the extra information. However, this option was not practical for several reasons. For one thing, the study had been conducted at many institutions, so obtaining further consent would be logistically difficult. Also, since the study had been conducted, some patients had died of their disease, so their further consent would be impossible to obtain. And--for both these reasons and others--local investigators would likely not be willing to make the effort to contact former subjects again.

Another option we considered was to seek institutional IRB approval without additional consent under 45 CFR 46.116(d). This provision of the research regulation allows an IRB to waive or alter the requirements to obtain informed consent provided the IRB finds and documents that:

1) the research involves no more than minimal risk to the subjects;

2) the waiver or alteration will not adversely affect the rights and welfare of the subjects;

3) the research could not practicably be carried out without the waiver or alteration; and

4) whenever appropriate, the subjects will be provided with additional pertinent information after participation.

Thus, a request could be made to the various IRBs involved for the release of the adverse event data from individuals treated with backbone chemotherapy, with personal identifiers removed. Investigators could argue that this request meets the first three requirements above and, therefore, continues to demonstrate respect for persons. Study investigators could then meet the fourth requirement by preparing an information sheet to be shared with trial participants and/or guardians regarding the pharmaceutical company's access to the data in order to maintain good communication with subjects.

Another challenge involving the provision of information can arise when the use is specified in the consent form, but documentation of a subject's consent is somehow compromised. This is often discovered through the routine audits that research studies sponsored by pharmaceutical companies and cooperative groups undergo.

Routine audits confirm that data provided by individual investigators are accurate (e.g., that a value is 182, not 128), that patients are eligible for the trial by corroborating information from the medical record (e.g., documentation that kidney function was appropriately assessed prior to study enrollment), and that the process of consent conforms to federal regulations by validating the documentation in the research record (e.g., a valid witness signature appears on the consent form).

Sometimes these routine audits find discrepancies that may make an individual ineligible for the trial. For example, if auditors can find no record of the order for a test required prior to study enrollment, the individual may be unable to participate even if the test was done; the absence of documentation is enough for many research groups to declare an individual ineligible. And in cases where the investigator or witness signature is missing from the consent form, or the consent signature is dated after therapy was begun, many clinical trial groups would classify the individual as ineligible for the study. They would do so because this scenario suggests deviation from the federal regulations that govern informed consent.

Yet when a pharmaceutical company is interested in using the results of a clinical trial for an FDA application, it is likely to be interested in all available information regarding drug exposure, adverse events, and outcomes, regardless of the eligibility status of the prospective trial participants. That is, even if an individual was deemed ineligible for one of the reasons outlined above, the company would probably still wish to have access to that individual's data. This is good practice not just in order to complete the data, but because communicating this information is the best way to protect human subjects. Without such reporting procedures, important safety information might not be available. For example, information about dosing adjustments or significant adverse events (like a death attributed to the new agent) might be withheld because the individual is now administratively ineligible to participate. This could result in trial data that incompletely characterize subject safety, which seems to us to violate the principles of beneficence and respect for persons. In fact, ensuring that important trial information is reported is the best way to support the principle of beneficence, which in turn ensures a favorable risk-benefit relationship in research. In addition, respect for persons is not violated provided the subject's understanding of the data-sharing arrangements is documented in the consent form.

Therefore, it is our view that data from individuals ineligible to participate can be shared with pharmaceutical company partners provided the consent form disclosed that the information would be shared with the company. Consider two possible scenarios. In the first, an audit reveals that a consent form is missing or unsigned. In that case, consent to share information with the company cannot be confirmed, and therefore, the data cannot be ethically shared. In the second case, the consent documentation is inadequate rather than absent. For example, the signed consent form is not dated, or the witness signature is missing. In this circumstance, there appears to be evidence that the subject understood and agreed to the data-sharing plan with the company. While the documentation problems are important, they do not compromise the subject's consent to share information with the company. Therefore, it is appropriate to share these individuals' data with the company, as they have already consented to doing so.

These examples demonstrate that it is important for investigators and IRBs to evaluate carefully requests for deviations from established consent plans. While it is always preferable to specify in advance the purposes for which data will be used, there may be justified uses of research data that were not detailed in the consent documents. Similarly, while we encourage investigators to make every effort to obtain proper consent documentation for all research participants, we would advise IRBs to think critically about the use of data where consent documentation is merely inadequate. Maintaining focus on the principles of beneficence and respect for persons will enable IRBs to make decisions that best minimize harm and maximize benefits to participants on the trial.


James R. Anderson presently serves on data and safety monitoring committees for research protocols conducted by Amgen, Merck, and the National Institutes of Health, for which he has received an honorarium and reimbursement for travel expenses.

* James R. Anderson, PhD, is Professor, Department of Biostatistics, College of Public Health, University of Nebraska Medical Center, Omaha, NE, and Toby L. Schonfeld, PhD, is Director, Center for Humanities, Ethics and Society, and Associate Professor and Vice-Chair, Health Promotion, Social and Behavioral Health, College of Public Health, University of Nebraska Medical Center, Omaha, NE.

James R. Anderson and Toby L. Schonfeld, "Data-Sharing Dilemmas: Allowing Pharmaceutical Company Access to Research Data," IRB: Ethics & Human Research 31, no. 3 (2009): 17-19.
Gale Copyright: Copyright 2009 Gale, Cengage Learning. All rights reserved.