[Read "Deadly Immunity," by Robert F. Kennedy Jr.]

A commentary on vaccine safety issues written by Robert F. Kennedy Jr. that appeared in Salon and Rolling Stone is rife with factual errors and distortions. Although there are numerous errors in Kennedy's piece that deserve correction and clarification, this letter focuses only on the errors pertaining to the Institute of Medicine.

Commentators and publications should take great care to ensure that their stories dealing with health provide parents with verifiable facts and reliable scientific evidence. Providing any less can lead to unwarranted fears and anxiety.

The Centers for Disease Control and Prevention and the National Institutes of Health asked the Institute of Medicine in 2000 to provide an independent and objective review of a series of vaccine safety concerns. This review resulted in a total of eight reports (for a full list see the IOM's Immunization Safety Review Web site). The charge to the study committee is a matter of public record and was discussed in a public meeting on Jan. 10, 2001.

Sponsors of IOM studies provide funding and their perspectives on what committees convened to carry out the studies should be charged to do. For example, in the case of IOM's studies of vaccines and autism, the CDC discussed what would make the committee's conclusions and recommendations most useful and understandable to policymakers, such as outlining specific research needs rather than giving a general recommendation that more research is needed. However, no organizations -- neither the CDC, nor the NIH, nor medical associations, nor parent groups, nor any others -- have control over committee members' conclusions and recommendations. The only factor that influences IOM committees' conclusions and recommendations is the evidence.

This is borne out by the results of IOM's first look at the thimerosal question in 2001. For those who are not aware, IOM visited the thimerosal question twice in its series on vaccine safety. At that time, the committee concluded that although there was not solid evidence linking thimerosal to brain disorders including autism, neither was the evidence sufficient to rule out a link. This would hardly seem to be a conclusion that would be reached by a committee that was "instructed to produce a report debunking the link," as Mr. Kennedy erroneously wrote. In addition, the committee in that report supported a call by the American Academy of Pediatrics for thimerosal to be removed voluntarily from vaccines.

Mr. Kennedy sought to buttress his misrepresentation of the purpose of the IOM studies by selectively using and fabricating quotations from the transcript of an organizational meeting held by the committee. He took words and phrases out of context, combined them and misrepresented them as direct quotes. The discussions recorded in this transcript focus on the committee members' efforts to make sure that they fully understood their charge and had reached agreement on the best way to conduct their study. Part of the discussions focused on the concerns, interests and stakes held by various parties, including both the CDC and parents.

For example, Mr. Kennedy used this partial quote from Dr. Marie McCormick: The CDC "wants us to declare, well, these things are pretty safe." Here is the full quote from Dr. McCormick: "I took away [from the previous day's discussion] actually an issue that we may have to confront, and that is actually the definition of what we mean by safety. It is safety on a population basis, but it is also safety for the individual child. I am wondering, if we take this dual perspective, we may address more of the parent concerns, perhaps developing a better message if we think about what comes down the stream as opposed to CDC, which wants us to declare, well, these things are pretty safe on a population basis." The full quote is part of a discussion of the need of parents to know if a vaccine is safe for an individual child (or if the vaccine is a danger for an ill child) compared with the public health community's need to know if vaccines pose risks to a whole population. In fact, Dr. McCormick proposed that the committee consider addressing the parental concerns about the health of an individual child -- not just the CDC's concern about population-wide effects.

In its 2004 report, the committee concluded that the totality of the evidence available indicated that neither thimerosal nor the measles-mumps-rubella vaccine is associated with autism. Contrary to Mr. Kennedy's assertions, the report clearly recommends that research continue to seek the cause or causes of autism, but that finite resources should be directed toward the most promising paths, and a significant investment in studies of the theoretical vaccine-autism connection is not useful at this time. The exact text of the IOM report reads: "The committee urges that research on autism focus more broadly on the disorder's causes and treatments for it. Thus, the committee recommends a public health response that fully supports an array of vaccine safety activities. In addition the committee recommends that available funding for autism research be channeled to the most promising areas." In addition, the committee recommended that trends in autism incidence in the United States be closely tracked now that thimerosal has been largely eliminated from routine immunizations. The report offers nine specific recommendations for research on autism, thimerosal and vaccine safety in general.

To reach its conclusions about the possible links between thimerosal and the MMR vaccine and autism, the committee reviewed a large amount of literature on this topic -- more than 215 references are cited in the 2004 report. The committee members based their conclusions on the whole body of the evidence, including toxicology studies, not just four epidemiological studies. The committee included several experts in epidemiology, pediatrics, clinical medicine, statistics and other relevant fields who reviewed and discussed in detail the strengths and weaknesses of the epidemiological, clinical and biological evidence pertaining to thimerosal and autism. Their detailed analysis of these strengths and weaknesses is included in their report, which is a public document available to anyone who wishes to read it. Mr. Kennedy did not offer a detailed analysis of why he views four of the studies reviewed by the committee as flawed.

In addition, Mr. Kennedy erroneously conflates the work of two entirely separate IOM committees. The IOM entered into a contract with the CDC in 2004 to assess the Vaccine Safety Datalink data-sharing program. (The VSD is a large, linked database of patient information such as vaccination histories, health outcomes and other characteristics that can be used to study whether health problems are associated with vaccinations.) As an independent, nongovernmental organization, the IOM cannot be forced to take on any particular study, nor was there pressure from Congress, parents or others to conduct this study. The Committee on the Review of the National Immunization Program's Research Procedures and Data Sharing Program did not review the work of the IOM Immunization Safety Review Committee, which wrote the report on vaccines and autism. The committee that reviewed the data-sharing program released a report in February 2005 that was directed to the CDC and others involved with the VSD, not to the IOM's Immunization Safety Review Committee.

Mr. Kennedy's piece does not mention that the IOM's studies on vaccines and autism -- like all IOM studies -- were peer reviewed (the equivalent of fact checking in the scientific community) by a group of outside experts before being approved for publication. The IOM, a branch of the National Academies, also follows stringent policies for reviewing the potential conflicts of interest or biases of individuals who are chosen to serve on its study committees. The members selected for the committee that reviewed vaccine safety issues had no links to the CDC, vaccine manufacturers, vaccine-related litigation or advocacy groups, and they had received no recent funding from the sponsors of the study. They received no remuneration for their work on the committee. The reputation of the National Academies for objectivity, integrity, independence and competence has been earned through numerous studies, and it is one of our most important assets.

Our corrections are only a subset of the many factual misstatements and errors contained in Mr. Kennedy's commentary. In the interests of accuracy and fairness -- and to alleviate unsubstantiated anxiety that this piece may have raised among readers -- we call upon Salon and Rolling Stone to publish this full letter and to run a complete correction of the many other errors contained in Mr. Kennedy's commentary.

-- Harvey V. Fineberg, M.D., Ph.D.

President, Institute of Medicine of the National Academies

Robert F. Kennedy Jr. responds:

Dr. Fineberg charges that I fabricated quotations by the participants of the January 2001 IOM meeting, but offers no evidence of "fabricated" quotes. I took all the statements from transcripts created by the IOM. Dr. Fineberg complains that I took Dr. Marie McCormick's words at the January 2001 organizational meeting out of context. He suggests McCormick wasn't trying to influence the study results when she said "the CDC wants us to declare, well, these things are pretty safe." But the context he adds does not contradict the way I used her quote.

Furthermore, additional statements made by McCormick ("We are not ever going to come down that [autism] is a true side effect of [thimerosal exposure]") and other meeting participants, including IOM study director Kathleen Stratton ("The evidence [will be] inadequate to accept or reject a causal relation") and Dr. Michael Kaback ("How we work our way out of the trap, I think is the charge"), make it clear that the panel members felt they were under pressure to exonerate thimerosal.

Dr. Fineberg complains that I didn't cite a 2001 report by the IOM that found a biologically plausible link between thimerosal and neurological disorders including autism, recommended removal of thimerosal from vaccines and advised that the links between thimerosal and the wide range of neurological disorders should be further researched. Dr. Fineberg argues that "this would hardly seem to be a conclusion that would be reached by a committee that was instructed to produce a report debunking the link." I did cite the 2001 IOM report in an early draft of "Deadly Immunity"; it was moved to a sidebar. But in 2004, the IOM contracted with the CDC to study the link between thimerosal and autism, and to look at studies that had appeared subsequent to the 2001 report. It was that committee's findings I wrote about in "Deadly Immunity."

Instead of conducting further studies of the possible causal links between thimerosal and a range of neurological disorders including speech delay, attention deficit disorder, hyperactivity, mental retardation and autism, as the 2001 IOM Immunization Safety Review Committee had recommended, the IOM narrowed its focus, at the request of the CDC, to study autism only. Although Dr. Fineberg says the committee cited 215 studies on links between thimerosal and autism, the report placed heavy emphasis on the four epidemiological studies I called flawed in "Deadly Immunity": studies of European countries where children had been exposed to only a fraction of the thimerosal concentrations administered to American children, and where they often undercounted the number of autistics.

The IOM's narrow focus on autism also allowed the committee to avoid the larger number of studies linking thimerosal to the wider spectrum of neurodevelopmental injuries. IOM chief staffer Kathleen Stratton effectively acknowledged that when she told me in an interview, "Clearly, mercury is very toxic. Clearly, ethyl mercury is neurotoxic. Clearly, ethyl mercury affects cell systems -- animals, human cells -- all those sorts of things, and clearly, when it was injected into newborn mice they had weird behavior. The point is, mercury is not good for you. Granted, thimerosal ... I mean it can't be good for you, right? And certainly, at some doses it's very bad for you. But the [only] question [we were charged with answering] is whether any of those animal or in vitro studies make a connection to autism."

I also question Dr. Fineberg's assurance that the IOM follows stringent policies for reviewing the potential conflict of interest of its panel members. Surely he is well aware of the many journalists and public officials who have raised questions about the web of ties among the CDC, the pharmaceutical industry and the research community. Fineberg's claims that the members selected for the Immunization Safety Review Committee had no links to the CDC and received no money from funding sponsors (the NIH and the CDC) are wrong. Committee member Dr. Rebecca Parkin is a former CDC epidemiologist. Dr. Gerald Medoff has a long history with NIH, having served on a Research Advisory Board for the National Institute of Allergy and Infectious Diseases, which along with the CDC contracted the IOM report. Additionally, his former employer, Washington University, receives significant funding from vaccine manufacturers. Dr. Steve Goodman received more than a dozen grants from NIH since 1993. Dr. McCormick's Harvard School of Public Health receives NIH funding as well as funding from pharmaceutical companies, and Dr. Bennett Shaywitz has received more than 50 NIH grants since 1988.

On the question of further research, Dr. Fineberg's own letter makes clear that the 2004 committee report recommended that research be funneled "to the most promising areas," which was almost universally understood to mean areas unrelated to the thimerosal-autism link. In fact, committee chairwoman Marie McCormick was explicit in a press conference after the report was issued: "The overwhelming evidence from several well-designed studies indicates that childhood vaccines are not associated with autism. We strongly support ongoing research to discover the cause or causes of this devastating disorder. Resources would be used most effectively if they were directed toward those avenues of inquiry that offer the greatest promise for answers. Without supporting evidence, the vaccine hypothesis does not hold such promise."

On the question of conflating two panels, my article described the February 2005 findings as the work of a "new" panel, but wrongly stated that it had been established to review the earlier committee's work. Instead, the new panel was convened to address continuing concerns about the Vaccine Safety Datalink data-sharing program. Its February 2005 findings were critical of the way the VSD had been used to study vaccine safety, which was central to the work of the Immunization Safety Review Committee's 2004 findings on thimerosal.

Dr. Fineberg claims that the IOM came under no pressure from the CDC to debunk the autism-vaccine link, despite the concerns voiced at the January 2001 meeting. But the CDC has failed to fully release the details of the contract (Task Order No. 74) that laid out the CDC's specifications to the IOM for the thimerosal-autism study, despite Freedom of Information Act requests. I challenge Dr. Fineberg to release the documents to support his claim, and further release all meeting notes, memorandums, e-mails, documents and transcripts from its Immunization Safety Review Committee from 2001 to 2004, including all closed-door meetings, which have been previously unavailable to the public. I also ask that Dr. Fineberg and the IOM join with other scientists, policymakers, legislators and advocates to demand that the CDC allow independent scientists to study the Vaccine Safety Datalink (which contains the vaccine and medical records of 6 million children). Instead, the CDC has closed the VSD to outside scientists, and has persistently defied federal law, congressional demands and appeals from the scientific community to open it.