COMMENT BY: The Biotechnology Industry Organization
SUBJECT: Against proposed revision to OMB Circular A-110
DATE: April 5, 1999

Dear Mr. Chamey:

These comments are submitted on behalf of the Biotechnology Industry Organization (BlO) and the Healthcare Leadership Council (HLC), which have interests in protecting research data from inappropriate use or disclosure. BlO represents over 850 biotechnology companies, academic institutions and state biotech centers and related organizations in 46 states and more than 26 countries. BlO members are involved in research and development of health care, agricultural, industrial and environmental biotechnology products. HLC represents 50 of the nation’s leading health care companies, including the biomedical research, health plan, hospital, medical device and provider sectors. We recognize that OMB is merely implementing a statutory directive and we strongly support its adherence to the expressed purpose of the law as set forth in the legislative history. Specifically, we note that the sponsor, Senator Richard Shelby (R. Ala.), emphasized the importance of the legislation for "ensuring that the public has access to all studies used by the Federal Government to develop Federal policy." We urge you to focus on narrowly targeting implementation of the law to fulfilling the legislative purpose. The Public Has Legitimate Interests in Data Generated in Federally Supported Research.

We agree in principle that the public has legitimate interests in data generated in federally supported research, including ensuring that federal funds are not misspent. Historically, however, data generated in grant-supported research has remained at the institution where the research is conducted. As a result, the data are not "agency records" subject to disclosure under the Freedom of Information Act (FOIA), as confirmed by the U.S. Supreme Court in 1980.

Moreover, federal regulations already provide authority for funding agencies to make research data and results available to third parties to further the government’s interests. In addition, grants administration policies of the Public Health Service require that primary data and unique research materials (such as cell lines, DNA, and genetic mapping information) developed with PHS funds must be made available after publication to qualified scientists on request. Some biomedical journals (for example, Cell) also require that scientists publishing research reports based on biological materials submit samples of the relevant cell lines, DNA, etc. to national depositories which, in turn, make them available to other researchers who want either to further the research or to test its validity.

In sum, most data necessary to replicate or otherwise test a biomedical scientist’s claims already are made available after publication of the research results. We believe that is as it should be and this new law should have no place in overturning existing data protection arrangements.

Premature Release of Data Could Jeopardize Research.

Almost 20 years ago, the HEW Ethics Advisory Board issued a report on "The Request of the National Institutes of Health for a Limited Exemption From the Freedom of Information Act." The requested exemption was intended to protect data accumulating in ongoing clinical trials from premature release. The Ethics Advisory Board, a panel of eminent individuals from diverse disciplines and backgrounds, concluded that release of incomplete or trend data could make it impossible to continue clinical trials because prospective participants might be unwilling to enroll in a study and accept the possibility of random assignment to treatments once trend data become known. As a result, research to determine the safety and efficacy of new drugs or devices might be ended prematurely, and unsafe or ineffective treatments could be introduced into medical practice. Premature release of data from epidemiological studies or other biomedical research could result in similar distortions of the findings that could mislead physicians or confuse the public.

RECOMMENDATION: OMB should not compel release of research data until the project or study data have been analyzed and the research team has published its findings.

FOIA Exemptions Should Apply.

As you know, the FOIA provides exemptions that protect certain information from compulsory release. These include, among others, proprietary information (trade secrets and commercial or financial information) (Exemption #4), and medical files and similar information disclosure of which would constitute a clearly unwarranted invasion of personal privacy (Exemption #6). It is urgent that OMB confirm that all exemptions in the FOIA will apply to disclosures mandated by Circular A-110. Without the protection of individuals’ privacy assured by Exemption #6, it will be difficult to recruit participants for research projects. And without the protection of commercial trade secrets provided by Exemption #4, commercial entities are likely to be disinclined to provide research support or investigational products to university-based scientists and clinicians. The risk would be too great that an agency might disclose trade secrets, reveal confidential business information, or destroy patent opportunities.

Even for researchers without commercial funding, release of certain information that is now protected could have serious consequences. For example, personal information such as a scientist’s social security number, home address, and home telephone number, currently are protected by Exemption 6 to the FOIA. If that exemption is inapplicable to requests submitted pursuant to Circular A-110, activists from various public interest groups that use extreme tactics would be in a position to aim their terrorist activities at scientists’ homes and families in addition to their laboratories. There is no indication that the sponsors of the bill intended such a result.

Finally, although most medical research files identify individual human participants by a unique code, some may not (for example, if patient records are included for purposes of medical history or eligibility for inclusion in the research). To protect confidentiality of research subjects, grantee institutions should be permitted to redact personal identifiers (such as name, address, and social security number) before forwarding copies of research files to the funding agency. The resources required, however, to locate and mask all personal identifiers could be enormous, and should be fully reimbursable under Circular A-110. More importantly, the master code linking research subjects’ names with the unique alphanumeric identifier assigned to them should be exempt from production even to the funding agency, so long as the grantee indicates in a cover letter that it has withheld the relevant document or computer print-out.

RECOMMENDATION: OMB should state emphatically that all exemptions provided by the FOIA apply to requests for research data pursuant to Circular A-110. In addition, grantees should be permitted to redact personal identifiers prior to submission of research records to the funding agency and to withhold lists or computer files linking research subjects’ names with their assigned codes, so long as they notify the agency that they have taken such action.

Research Data Generated Prior to the Effective Date of the A-110 Revisions Should Not Be Subject to Compelled Disclosure.

Retroactive application of a new law is strongly disfavored in American jurisprudence. Moreover, nothing in the statute requires retroactive application. The strongest argument against it, however, may be that it would vitiate promises of confidentiality made by the Department of Health and Human Services to applicants for research grants, and by academic institutions to commercial sponsors of research. Instructions accompanying application forms for NIH grants (PHS 398 and PHS 2590) advise applicants that "[g]enerally not available for release to the public are competing grant applications (initial, competing continuation, and supplemental) for which awards have not been made; evaluative portions of site visit reports; and summary statements of findings and recommendations of review groups." Although brief summaries of work in progress and preliminary research may be released after funding, applicants are not warned that underlying data will be demanded and disclosed. Such disruptive and invasive measures should not be taken without advance warning.

Similarly, virtually all research agreements negotiated between industry and academic institutions rely on promises of confidentiality of research ideas and data, at least until formal publication of the results. The proposed amendments to Circular A-110 would destroy such agreements for all industry-sponsored research that also receives federal support.

RECOMMENDATION: Revisions to A-110 compelling disclosure of grant-supported research data should not apply to: (I) data generated prior to the effective date of the revision; (2) data generated pursuant to agreements between academic institutions and commercial entities signed prior to the effective date of the revision; or (3) data generated in research supported by a federal grant awarded prior to the effective date of the revision. Funding agencies should be directed to revise grant application instructions and forms to warn applicants that their research data may be subject to compulsory disclosure pursuant to the FOIA and OMB Directive A-110.

"Publication" Should Be Defined.

We strongly endorse OMB’s proposal to permit FOIA access to research data only after the findings have been published. To implement this directive, however, OMB must define "published" or "publication." We believe that research data should be protected from FOIA release until the results of the research have been published by those who conducted the study. Principal investigators (research team leaders), like other authors, have a right to control the timing and manner of the initial presentation of their work. Premature, compelled release of their data would deprive them of that right on which patents, copyright, and other intellectual property interests depend. This is particularly true of draft manuscripts being prepared for publication. In this connection, we note that guidelines for the conduct of intramural research at NIH state that the principal investigator is responsible for determining when research data are ready for publication.

We recognize that some of our colleagues have urged OMB to define "publication" as publication in a peer-reviewed journal. Unfortunately, that would have the incongruous effect of protecting the data of those who publish on the internet or in non-peer-reviewed journals, while requiring release of data only from those who publish their reports in peer-reviewed journals. Instead, we believe the preferred approach is to define publication as the first formal and complete presentation of research and its findings in whatever medium the principal investigator chooses to use.

RECOMMENDATION: OMB should adopt the following definition of "publication:" Publication" means dissemination of a formal report of research results by the principal investigator and his or her co-authors. In addition to papers published in scientific journals, "publication" includes formal, full length oral presentations at scientific meetings (complete with tables and graphs, as appropriate) and publication of organized reports by electronic means. Abstracts or summaries of research submitted to a funding agency in grant applications or progress reports do not constitute publication for purposes of compulsory release pursuant to A-110. Similarly, brief references to the study in any medium, testimony in judicial proceedings compelled by subpoena, and reports by journalists, are excluded from the definition of "publication.

"Data" Subject to Compulsory Release Under A-110 Should Be Defined.

Defining "data" is key to avoiding ambiguities as to what must be released under Circular A-110. In the biomedical sciences, "data" include a broad spectrum of materials including: entries in lab notebooks; printouts from computers and other electronic equipment (e.g., those measuring blood flow or electrical activity of the heart, brain, or peripheral nerves); chemical compounds; bioassays; DNA sequences; slides containing blood, tissue samples, and other biological materials obtained for medical or research purposes; X-Ray, CAT scans, and MRI films; data remaining in computers; computer programs; and a variety of biological materials stored in refrigerators and freezers (such as, whole organs, fertilized eggs, tissue specimens, blood and other bodily fluids, cell lines, DNA, bacteria, and viruses). It does not appear that release of all such materials was contemplated when the statutory language was adopted by the conferees.

RECOMMENDATION: The definition of "data", for purposes of release under A-110 should be limited to information maintained in handwritten, printed or electronic form that is readily duplicated or photocopied. If a requ