COMMENT BY: Association of American Medical Colleges
SUBJECT: Against proposed revision to OMB Circular A-110
DATE: March 24, 1999

Dear Mr. Charney:

The AAMC appreciates the opportunity to respond to the Notice of Proposed Revision to OMB Circular A-11O (Federal Register, February 4, 1999). The proposed revision would make data generated under federal awards to medical schools, universities, hospitals, and other non-profit institutions accessible to the public under the Freedom of Information Act (FOIA). The Association of American Medical Colleges represents the nation’s 125 accredited medical schools, approximately 400 affiliated teaching hospitals and health systems, and 90 academic and scientific societies. Our member organizations, all of which are subject to Circular A-110, perform nearly half of all sponsored extramural research funded by the National Institutes of Health (NIH), and more than two-thirds of all NIH research awarded to academic institutions. Thus, this proposal is of great concern to the Association and its members.

Although the purpose of this letter is to provide constructive criticism of the proposed revision and to explicate the problems it will pose, the Association must state at the outset its unequivocal belief that the underlying statute is misguided and irreparably flawed. Despite OMB’s laudable attempt to refine the language to convey more precisely its legislative intent, the proposed revision to Circular A-110 would nonetheless have a severely deleterious effect on the vitally important scientific undertakings of our members and other organizations engaged in health research. Furthermore, the proposed revision will not, in the end, promote the Congress’ and the scientific community’s mutual interest in increasing the practice of data sharing under federally sponsored research. Specific areas of concern are discussed below:

The Inherent Unsoundness of the Underlying Statute

The proposed revision and underlying statute represent a dramatic alteration of current federal policies and are not merely an extension of existing regulation (as asserted by the statute’s proponents). Currently under Circular A-110, federal agencies retain the right to obtain data produced under federal awards and to transmit them to third parties "for Federal purposes" (Section _.36 (c)(2) [sic] ). Federal agencies also have the option to waive this right. The proposed revision would eliminate this waiver and would compel the agency to obtain data from non-governmental research laboratories at the request of any party, from anywhere in the world, for any purpose.

The proposed revision and underlying statute clearly intend that the FOIA’s protections against the release of some types of sensitive information remain in force. However, the redaction of sensitive information would be accomplished by the federal agencies only after the data are in their possession. Research universities, medical schools, and teaching hospitals would be obligated to provide all data related to certain research findings, including protected confidential and commercial data, to the sponsoring agency to be incorporated within federal records. This could have a profound impact on the collection of data related to clinical trials, epidemiologic studies, and other areas of medical research, since the use of patient medical records, public health data sets, and other data sources is governed by strict confidentiality agreements that require researchers stringently to observe and control the uses to which these data are subjected. Although federal agencies such as the NIH would certainly be diligent in efforts to apply fully the FOIA’s protection to these records, the records would nevertheless leave the institution’s control.

Consequently, medical schools and hospitals may need to undertake an extensive reassessment of research confidentiality agreements, and may, in some cases, be compelled temporarily to suspend lines of research or risk abrogating existing agreements. We are further concerned that the privacy exemption under the FOIA extends only to individuals, and would not adequately protect the rights and anonymity of entities such as community clinics, hospitals, schools, or other institutions participating in critical but sensitive aspects of public health research, such as mental illness, sexually transmitted diseases, alcoholism, or drug abuse. We believe that these changes could have a chilling effect on the willingness of individuals to participate in clinical research, or to authorize research uses of their medical records. It is ironic that in these respects the underlying statute flies in the face of the bipartisan efforts in both the 105th and 106th congresses to strengthen the privacy of medical information and restrict its disclosure through new federal legislation.

Similarly, the statute makes no allowance for research projects involving mixed federal and industry sponsorship at academic institutions. Much promising medical research involves cooperation with industry and the use of new technologies under development, including pharmaceuticals and medical devices. The concern that sensitive data could be made available to competitors would create disincentives for industry participation in research leading to patent applications, clinical trials of new therapeutics, or any research with commercial implications. While the FOIA exempts the release of commercially sensitive information, it is not known how adequately the protection would be extended to university-based research fully or partially sponsored by industry, as in the case of Cooperative Research and Development Agreements (CRADAs). The proposed revision would allow the premature release of data before their potential patentability could be adequately assessed, and would thereby preclude researchers from obtaining patents or other protections for intellectual property. It would thus undermine the Bayh-Dole Act and other federal policies that encourage development of intellectual property and commercial application of technologies derived from federally sponsored research. OMB has perhaps obviated this problem by limiting releases of data to those associated with "published" research (by which time any intent to patent has already been pursued), but this poses its own ambiguities as discussed below. In addition, some legal experts and Members of Congress have expressed the view that this and other OMB efforts to restrict the sweep of the original statute may not withstand legal challenge. The AAMC thus believes that these intellectual property concerns continue to be salient and germane.

Persistent Ambiguity in the Terms and Scope of the Revision

Although the OMB has attempted to ameliorate the ambiguity of the statute by restricting the FOIA’s application to data relating to published research findings used in developing federal policy or rules, the proposed revision to Circular A-110 leaves considerable uncertainty and latitude that would be difficult for agencies and research institutions to navigate. The term "data," for example, remains undefined by the proposed revision and can be vastly expansive and used differently by various individuals, institutions, and agencies. It is thus unclear to what types of records and materials the FOIA would apply, and whether they would extend to observers’ notebooks, data from secondary sources, administrative records, films, or even laboratory samples. Similarly, there is no definition of "publication," which might be interpreted to include peer-reviewed literature, lectures, communications, doctoral dissertations and theses, books, or other media. It is also unclear if an investigator is legally compelled to release all data collected in the course of a lengthy or exhaustive study, or only those data that, in retrospect, proved critical to the conclusions of a specific published paper in question. Finally, the proposal does not specify who would have final authority to make these determinations: the investigator, the research institution, the funding agency, the regulatory agency, the Department of Justice (as provided under the FOIA), the courts?

The question of identifying data used in federal policy making is also ambiguous. It is not always feasible to link research data directly or causally to federal policies and rules, since regulatory agencies seldom cite specific research findings in their rulemaking. More often, these agencies rely on an extensive and illimitable array of findings (only some of which are produced by academic or other institutions with federal funding). For example, a new federal regulation concerning tobacco use could rest upon an extraordinarily vast international body of literature on tobacco-related illnesses. Indeed, most federally funded biomedical research is supported under the auspices of the U.S. Public Health Service in the expectation that almost any aspect of the research could support revised health policies and regulations, as well as new therapies and cures.

The Burden of Cost in Implementation

The proposed revision would dramatically compound the costs of research at academic and other institutions. It would divert the time and effort of laboratory personnel to comply with unpredictable and possibly frequent requests for data and provide necessary documentation. Additional staff effort would be required to process data requests and maintain records certifying compliance. Institutions would almost certainly need to employ additional legal and administrative personnel in much the same way that federal agencies must maintain dedicated staff in their freedom of information offices. While the provision allows for a fee to cover additional costs to the Federal Government for obtaining data, revenues derived from this fee would devolve to the U.S. Treasury, and would not remunerate the sponsoring agency or the institution. Finally, inordinate or tendentious requests for data could involve research institutions in protracted litigation. In fact, the unremunerated burden of compliance, together with the invitation for litigation, make the provision an ideal mechanism for intimidation and harassment of researchers and their institutions, as we believe has already occurred in some instances under state law.

The AAMC is notably concerned that few of our academic institutions would be able to increase the recovery of indirect administrative costs from federal grants to pay for the burden of complying with FOIA requests, as the reimbursement of administrative costs under federal awards has been capped at 26 percent of the direct costs. Many of our institutions already incur administrative costs that meet or exceed this level. Consequently, the increased administrative burden would be almost certain to come at the expense of the institution’s own financial resources. The proposed revision, if implemented, would make it clearly untenable for federal policy to continue the existing arbitrary cap on indirect administrative costs reimbursement.

In Support of Data Sharing

The AAMC believes the responsibility to promote an open exchange of research data (and reagents) is a shared responsibility of scientists, funders, and journals. We also believe it appropriate that certain obligations to share scientific data attach to the acceptance of federal funds to conduct research. Data sharing policies are most appropriately established and effectively enforced by research sponsors and publishers, and have been promulgated by the NIH, the National Science Foundation (NSF), and a number of prestigious scientific journals. Sound policies to promote data sharing by federally funded investigators should be a responsibility of these and other federal funding agencies, in extensive consultation with the Congress, the Administration, the scientific community, and the public.

We recognize that, at the complex interface of scientific research and the formulation of policy or regulations, there may be issues that demand examination and resolution. We are convinced however, that exposing all federally funded research to access under the FOIA is an inappropriate and ineffective way to address these issues. The Freedom of Information Act is far too blunt an instrument to accomplish the free and open exchange of scientific data. Its use could indiscriminately retard the progress of important lines of research, bleed resources from productive activity, and subject the processes of independent scientific inquiry to adversarial intimidation. It could actually encourage scientists to restrict their communication and discussion of research activities to preserve the autonomy necessary to their research.

We recommend that the underlying statute be repealed or, at a minimum, suspended pending further deliberation. We urge the OMB to review the implications of the proposed revision and the underlying statute most carefully, and not to implement the revision until the consequences are better understood by the Administration and the Congress. The AAMC encourages the OMB to return to Congress to seek reversal or suspension of this statute until these issues can be thoughtfully addressed in an appropriate forum, perhaps under the auspices of the National Academy of Sciences or its Government-University-Industry Research Roundtable. The AAMC originally advised granting a 90-day comment period for this proposed revision (correspondence to Director Lew, December 9, 1998). Circumstances still warrant an extension of this comment period, or a second comment period, for more thorough response by the affected publics.

Sincerely,

Jordan J. Cohen, M.D.