Read Full Report

Date: Feb. 17, 2005
Contacts: Christine Stencel, Media Relations Officer
Chris Dobbins, Media Relations Assistant
Office of News and Public Information
202-334-2138; e-mail <>


Independent Oversight of Vaccine Safety Data Program Needed
To Ensure Greater Transparency and Enhance Public Trust

WASHINGTON -- Two new oversight groups are needed to ensure that the policies and procedures of the Vaccine Safety Datalink (VSD) and its data sharing program -- which is intended to give researchers access to patient data that will help them study vaccine safety issues -- are implemented as fairly and openly as possible, says a new report from the Institute of Medicine of the National Academies. The Centers for Disease Control and Prevention, which oversees VSD and the data sharing program, should create a new, independent committee to review researchers' proposals to use VSD data, monitor adherence to protocols, and advise the agency and its partners on when and how to release preliminary findings based on the data, the report says. In addition, CDC should create a new subcommittee of the National Vaccine Advisory Committee (NVAC), or tap an existing one, to enable stakeholders to review and provide input on the VSD research plan every year.

"Concerns about access and transparency have accompanied the development and functioning of the Vaccine Safety Datalink data sharing program, and consequently some people's trust in the reliability of findings from VSD studies has eroded," explained John C. Bailar III, chair of the committee that wrote the report and emeritus professor of health studies at the University of Chicago. "Taking steps to improve the independence, transparency, and fairness of VSD procedures will help enhance confidence in the data sharing program and in research based on this important tool for evaluating vaccine safety."

The Vaccine Safety Datalink is a large, linked database of patient information that was developed jointly by CDC and several private managed care organizations in 1991. It includes data on vaccination histories, health outcomes, and characteristics of more than 7 million patients of eight participating health organizations. Researchers from CDC and the managed care groups have used VSD information to study whether health problems are associated with vaccinations. The subsequent VSD data sharing program was launched in 2002 to allow independent, external researchers access to information in the database. Four successive versions of guidelines for external researchers who wish to use VSD data have been released; the most recent draft is posted in the Federal Register with a March 1 deadline for public comment.

Because of certain unique characteristics, the VSD data sharing program does not function in the same way that other data sharing programs do. In addition to the protections accorded to patient information by privacy laws, the participating managed care organizations maintain proprietary control over data they have supplied from 2001 onward. This means that these data are not available to external researchers working independently on most types of studies. Even so, because the VSD is a tax-supported public resource used to inform health policy decisions, the public deserves access to the data that influence such decisions, the report says; the public also is entitled to transparency and independence in the processes that permit or restrict access. The committee concluded that it is possible to facilitate public access and transparency while also protecting patient privacy.

The public should be involved in setting priorities for VSD research, given the controls placed on information available through the VSD data sharing program, the report says. To that end, a subcommittee of NVAC with representation from a wide variety of stakeholders -- from vaccine manufacturers to federal agencies to advocacy groups -- should provide feedback and input on the program's research plan each year. Although such a group would inevitably have biases and conflicts of interest, it is important that VSD officials hear from all interested individuals. To promote transparency, the subcommittee's deliberations should take place in open meetings.

A separate, independent committee should review research proposals to use VSD data to ensure that they meet the criteria for access. In the interest of fairness and openness, CDC needs to spell out more clearly these criteria, among which the technical feasibility of the proposal should be paramount, the report added. The committee also should oversee changes in research protocols for VSD studies conducted by researchers from CDC or the managed care groups. Members of this independent review committee should be chosen on the basis of their scientific and technical expertise and their lack of conflicts of interest or biases, or CDC's ability to balance these.

Another key function of the independent review committee should be to counsel VSD scientists and officials on when and how to release preliminary findings from studies based on VSD data. In some instances, public disclosure of preliminary findings based on incomplete data or analyses that have not undergone external peer review may be necessary to protect the public's health and safety. But early release can lessen people's confidence in the final results of a study if further data collection and analysis lead to different final conclusions. In nearly all situations, preliminary findings should undergo independent, external peer review before they are shared with the public or used as the basis of policy decisions, the report says.

Given that the managed care organizations maintain proprietary control over data collected beginning in 2001, independent, external researchers will have to collaborate with a partner from CDC or one of the health groups to gain access to these data. Although it is impossible to make collaboration mandatory, CDC should require each participating managed care group to designate an individual whose role is to facilitate and encourage partnerships.

Because any breach of patient data confidentiality could discourage the managed care groups from participating, CDC should make certain that its rules for addressing confidentiality violations are clearly spelled out in the program guidelines and agreement forms, the report says. In addition, the committee concluded that it is reasonable to expect independent, outside researchers to provide VSD officials with status reports on their studies. When scientists have findings ready for public release, they should be required to provide CDC copies of any manuscripts at least 30 days before submission for publication and copies of presentations to be given at public events at least 15 days in advance.

The study was sponsored by Centers for Disease Control and Prevention. The Institute of Medicine is a private, nonprofit institution that provides health policy advice under a congressional charter granted to the National Academy of Sciences. A committee roster follows.

Pre-publication copies of Vaccine Safety Research, Data Access, and Public Trust are available from the National Academies Press; tel. 202-334-3313 or 1-800-624-6242 or on the Internet at Reporters may obtain a copy from the Office of News and Public Information (contacts listed above).

[ This news release and report are available at ]

Board on Population Health and Public Health Practice

Committee on the Review of the National Immunization Program's
Research Procedures and Data Sharing Program

John C. Bailar III, M.D., Ph.D. (chair)
Professor Emeritus
Department of Health Studies
University of Chicago
Washington, D.C.

Garnet L. Anderson, Ph.D.
Co-Principal Investigator
Women's Health Initiative Clinical Coordinating Center
Fred Hutchinson Cancer Research Center

Stephen E. Fienberg, Ph.D.
Maurice Falk University Professor of Statistics and
Social Science
Carnegie Mellon University

Debra R. Lappin, J.D.
Senior Adviser
Public Health and Life Sciences Consulting
B&D Sagamore
Washington, D.C.

Mryon M. Levine, M.D., D.T.P.H.
Professor and Director
Center for Vaccine Development
School of Medicine
University of Maryland

Anna C. Mastroianni, J.D. M.P.H.
Assistant Professor
School of Law and Institute for Public Health Genetics
University of Washington

Colin L. Soskolne, Ph.D.
Department of Public Health Sciences
University of Alberta
Edmonton, Alberta

Elaine Vaughan, Ph.D.
Associate Professor
Department of Psychology and Social Behavior
School of Social Ecology
University of California


Andrea Pernack Anason, M.P.H.
Study Director