Skip Navigation
Return to NIEHS | Current Issue
Increase text size Decrease text size

NIEHS Convenes Independent Expert Panel to Consider Use of CDC Vaccine Database to Investigate Thimerosal/Autism Connection

By Ernie Hood
June 2006

The NIEHS convened a panel of scientific experts Thursday, May 4, at the Institute's Research Triangle Park, North Carolina campus to determine the feasibility of using a large database maintained by the Centers for Disease Control and Prevention (CDC) called the Vaccine Safety Datalink (VSD) to investigate the potential connection between exposure to the vaccine preservative thimerosal and autism. The panel was comprised of specialists in clinical and epidemiological research, biostatistics, neurotoxicology, and risk assessment.

NIEHS Director David Schwartz set the tone for the day's proceedings in his welcoming remarks. "We are very committed to this area of research," he said. "We believe it is essential that we identify the causes of autism...and that we explore all possibilities to try to decrease the risk of developing this disorder among the children in our country." Schwartz acknowledged the high level of expertise represented among the panel members and summarized the panel's mission. "We have assembled a group of individuals that we believe have a strong interest in getting to the roots of autism, but are independent and unbiased, and can look at the Vaccine Safety Datalink and identify the strengths, weaknesses, and utility of the VSD in terms of autism research. The charge to the committee is to evaluate the utility of the VSD as a resource to better identify risk factors related to autism, and to develop recommendations for the next steps."

The workshop's morning session was devoted to presentations that provided detailed background information to the panel members on the VSD-its technical aspects, considerations on research collaborations and data sharing, and current uses of the database, including past and ongoing research studies addressing thimerosal and autism. The balance of the day was taken up by discussion and brainstorming among the panel members.

Discussion initially focused on the feasibility and utility of the VSD for conducting a relatively straightforward "ecologic" study, looking at trends over time to determine whether changes in autism rates were correlated with the removal of thimerosal from most US vaccines in the late 1990s. Allowing for a lag time between thimerosal exposure by means of routine childhood immunizations and diagnosis of autism, an ecologic study should reveal an eventual concomitant drop in autism rates if thimerosal exposure was in fact a major causative factor. Most of the panelists agreed that given the limitations inherent to the VSD, it would not be the best source of data for such a study.

The panel also expressed concern that some of the limitations in the VSD must be addressed prior to further consideration and recommendation of any study. Panel members' concerns centered on limitations in the data's applicability to address specific questions regarding autism and vaccines, on variation in the quality of the VSD data, and on the need for an evaluation of the accuracy of the information in the VSD. For example, screening, developmental assessment, diagnostic and coding practices appear to vary from HMO to HMO, and the diagnosis of autism itself has changed over time. Panelists noted that the database appears to have less-than-complete information on referrals for outside evaluations and continuity of enrollment, both of which could result in underreporting of autism cases. Concern was also expressed about the ability to identify and access VSD records containing information on subjects' families, particularly data regarding maternal prenatal vaccinations and exposure to mercury-containing drugs, which could represent alternate or additional routes of exposure to mercury by newborn infants.

If these and several other questions about the VSD can be effectively resolved or circumvented, the panel recommended that a variety of potential future study designs should be considered. Several members agreed that it would be fruitful to consider a study in high-risk populations, such as the siblings of children diagnosed with autism or autism spectrum disorders. There also appeared to be consensus that data regarding prenatal exposure to thimerosal would be crucial to the validity of study conclusions, and that obtaining data from individual medical records would be necessary in order to reduce confounding.

Although the strengths and weaknesses of several potential study designs were discussed, panel members seemed particularly intrigued by a proposal to re-examine and expand upon a VSD-based thimerosal vaccine safety study previously conducted by the CDC (Verstraeten et al. Safety of Thimerosal-Containing Vaccines: A Two-Phased Study of Computerized Health Maintenance Organization Databases. Pediatrics. Volume 112, No. 5. November 2003.) It was suggested that updating this retrospective cohort study by adding data from more recent years and refining the methodologies employed will address some of the limitations of the previous study.

The panel was in broad agreement about the need for vigorous and transparent oversight of any eventual study, regardless of its final form. Members were particularly anxious to ensure public representatives of autism advocacy groups from the autism community be fully included in any entity empowered to develop and oversee the research initiative.

Several public advocacy groups were represented at the workshop, and were given the opportunity to address the expert panel on behalf of parents, families, caregivers, and children with autism. Many spoke movingly of their personal situations, and of the urgent need to aggressively deal with the autism epidemic facing the nation. They also expressed great frustration and skepticism about the objectivity of prior government-lead efforts to investigate the thimerosal/autism connection, as well as optimism that the convening of this expert panel by the NIEHS would result in a fresh, independent, and unbiased consideration of the issue.

A more detailed report of the panel meeting and deliberations will be completed within 90 days. The report will be made available to the public for comment, will be shared with the Autism Coordinating Committee for the National Institutes of Health (NIH). The actions eventually taken are likely to involve partnerships among NIH institutes and other federal agencies.

The Expert Panel was chaired by Irva Hertz-Picciotto, Ph.D. of the University of California, Davis. The other panel members were: Scott Bartell, Ph.D. of Emory University; Thomas Burbacher, Ph.D. of University of Washington; Julie Daniels, Ph.D. of University of North Carolina at Chapel Hill; Philip Davidson, M.D. of University of Rochester School of Medicine and Dentistry; Pamela Factor-Litvak, Ph.D. of Columbia University; Craig Newschaffer, Ph.D. of Johns Hopkins University; and Chirayath Suchindran, Ph.D. of University of North Carolina at Chapel Hill.

The Workshop's Planning Committee included: Cindy Lawler, Ph.D.; G. Jean Harry, Ph.D.; Christopher Portier, Ph. D.; and Sheila Newton, Ph.D.; of the NIEHS, and Frank DeStefano, M.D. of the CDC.

"NIEHS Well Represented at..." - previous story Previous story Next story next story - "DERT Papers of the..."
June 2006 Cover Page

Back to top Back to top