eLetters to:

ARTICLE: Thomas Verstraeten, Robert L. Davis, Frank DeStefano, Tracy A. Lieu, Philip H. Rhodes, Steven B. Black, Henry Shinefield, Robert T. Chen for the Vaccine Safety Datalink Team Safety of Thimerosal-Containing Vaccines: A Two-Phased Study of Computerized Health Maintenance Organization Databases Pediatrics 2003; 112: 1039-1048 [Abstract] [Full text] [PDF]

eLetters: Submit a response to this article

eLetters published:

Comments on Verstraeten et al, Safety of Thimerosal-Containing Vaccines from Nov 5, 2003 Pediatrics

Neal A. Halsey, Daniel A. Salmon and Lawrence H. Moulton   (17 December 2003)

Safety of Thimerosal-Containing Vaccines: Response to Halsey et al

Frank DeStefano, Philip H. Rhodes and Robert L. Davis   (9 January 2004)

Study Misses Link Between Thimerosal and Neurodevelopmental Disorders

Mark R. Geier, David A. Geier of Medcon, Inc.   (23 February 2004)

Safety of Thimerosal-Containing Vaccines -- Response to Geier, et al

Frank DeStefano, Philip H. Rhodes   (2 March 2004)

False information in DeStefano/Rhodes response to M. Geier

Brian S. Hooker   (24 March 2004)

The Safety of Thimerosal-containing Vaccines -- Response to B.S. Hooker

Frank DeStefano   (26 March 2004)

Re: The Safety of Thimerosal-containing Vaccines -- Response to B.S. Hooker

Brian S. Hooker, Ph.D., P.E.   (30 March 2004)

The Safety of Thimerosal containing vaccines: Response to Destefano

Lyn H. Redwood   (31 March 2004)

Safety of Thimerosal-Containing Vaccines -- Response to L. H. Redwood

Frank DeStefano   (5 April 2004)

Comments on Verstraeten et al, Safety of Thimerosal-Containing Vaccines from Nov 5, 2003 Pediatrics 17 December 2003
Neal A. Halsey Johns Hopkins Bloomberg School of Public Health - Institute for Vaccine Safety, Daniel A. Salmon and Lawrence H. Moulton Send letter to journal: Re: Comments on Verstraeten et al, Safety of Thimerosal-Containing Vaccines from Nov 5, 2003 Pediatrics nhalsey{at}jhsph.edu Neal A. Halsey, et al.	To the editor, Verstraeten et al. (1) are to be commended for trying to determine if there were adverse effects from administering multiple vaccines that contained thimerosal. However, we would like to point out a few aspects of the analyses and their interpretation that invite further elaboration of these data. The results differ somewhat from those presented to the Institute of Medicine (IOM) in 2001 (2), where a statistically significant dose- response association was observed between thimerosal exposure by three months of age and neurodevelopmental delay assessed by combining several diagnostic categories. Such an analysis is reasonable since there is variability in coding of diagnoses for children with developmental delay and since a related compound, methylmercury, is known to be associated with multiple effects on neurological development. Analysis by separate diagnostic category may have substantially reduced the power to find important relationships. In addition, the selection criteria used in the published article appear to have been more lax than in the IOM presentation, as the latter was based on fewer outcomes. Several important issues were not adequately described. Were all DTaP, Haemophilus influenzae type b, and hepatitis B vaccines assumed to contain the thimerosal preservative? Some vaccines were available that did not contain this preservative (3). Were diagnoses that were not made by a specialist (i.e. not validated) excluded from analyses? Primary care physicians are capable of diagnosing attention deficit disorder (ADD) without input from a subspecialist. Were analyses conducted to determine if there are differences in associations by gender? Males are more susceptible to the effects of methylmercury and elemental mercury than females (4). The authors correctly hypothesize that concerned parents who obtained vaccines on time for their children may have been more likely to seek evaluations for delayed development and note the association between thimerosal exposure and increased visits for upper respiratory infections and well-child care in the second year of life. However, parents of children with mild developmental delay might have been more likely to seek evaluations for minor illnesses and well-child visits because of increased concern about their children's well-being. The authors discount the positive association in HMO B between thimerosal exposure and language delays because no association was found in HMO C. The power to detect outcomes in HMO B was much greater than for HMOs A and C as the HMO B population was 8.3 and 6.6 times larger than HMOs A and C respectively. Also, HMO B had more variability in exposures to thimerosal, based on the information provided to the IOM. Discrepant findings could be due to differences in background exposure to mercury from other sources and coding of outcomes, and do not negate the statistically significant associations found. Some analyses for outcomes of interest were not conducted when there were fewer than 50 outcomes in HMOs A and C. These analyses should have been included with appropriate statistical analysis rather than excluded altogether. The authors correctly suggest that detailed neurodevelopmental assessments of children who were randomized to receive vaccines with different amounts of thimerosal will provide useful data. However, such studies could have limitations. Few studies of sufficient size will have enough variability in thimerosal exposure to reflect the full range of exposures that have occurred. Also, intensive therapeutic intervention for young children with delayed speech or language problems can result in significant improvements, and studies in rats suggest that some adverse effects of lead exposure early in life can be partially ameliorated by intensive retraining (5). Therefore, detailed histories of past problems should be included in comparison studies conducted at 5-7 years of age. Public confidence in vaccines and the National Immunization Program could be enhanced if there was greater independence in vaccine safety assessments from the highly successful program to promote immunizations (6). We believe an independent organization, perhaps the IOM, should convene a panel with expertise in neurodevelopmental delay, effects of methylmercury exposure, and statistical methods to review the data and conduct additional analyses if indicated. References 1. Verstraeten T, Davis RL, DeStefano F, Lieu TA, Rhodes PH, Black SB, Shinefield H, Chen RT; Vaccine Safety Datalink Team. Safety of thimerosal-containing vaccines: a two-phased study of computerized health maintenance organization databases. Pediatrics 2003;112(5):1039-48. 2. Verstraeten T. Vaccine Safety Datalink (VSD) Screening Study and Follow-up Analysis with Harvard Pilgrim Data. Presented at (IOM Meeting) Thimerosal-Containing Vaccines and Neurodevelopmental Outcomes; 2001 Jul 16; Cambridge, Massachusetts. 3. Thimerosal in vaccines: a joint statement of the American Academy of Pediatrics and the Public Health Service. MMWR 1999;48:563–565. 4. Grandjean P, Weihe P, White RF, Debes F. Cognitive performance of children prenatally exposed to "safe" levels of methylmercury. Environ Res 1998 May;77(2):165-72. 5. Guilarte TR, Toscano CD, McGlothan JL, Weaver SA. Environmental enrichment reverses cognitive and molecular deficits induced by developmental lead exposure. Ann Neurol 2003;53(1):50-6. 6. Salmon DA, Moulton LH, Halsey NA. Enhancing Public Confidence in Vaccines through Independent Oversight of Post-Licensure Vaccine Safety. AJPH 2004; in press.
Safety of Thimerosal-Containing Vaccines: Response to Halsey et al 9 January 2004
Frank DeStefano, Medical Epidemiologist National Immunization Program, Centers for Disease Control and Prevention, Atlanta, GA, Philip H. Rhodes and Robert L. Davis Send letter to journal: Re: Safety of Thimerosal-Containing Vaccines: Response to Halsey et al fxd1{at}cdc.gov Frank DeStefano, et al.	To the editor: Dr. Halsey and colleagues raise several important questions about our study and we appreciate the opportunity to provide additional details. Several factors entered into our decision not to include the general category of “any neurodevelopmental disorder” in the results presented in the article, even though we had included such a category in a presentation of preliminary results to the Institute of Medicine (IOM). In the preliminary analysis, children with diagnoses of speech or language delay constituted a large proportion of the combined “any neurodevelopmental disorder” category. Thus, the combined category results were primarily a reflection of the speech and language delay results. Moreover, the combined category contained several heterogeneous conditions with variable age distributions and we were advised by some reviewers to delete this category. In terms of the selection criteria, they were not more lax for the published report than for the IOM presentation. Actually, the criteria were more stringent for the published report. After the IOM presentation, we obtained more complete data on exclusionary conditions from HMO A. We also applied the exclusion criteria in a stricter fashion by excluding children who ever received one of the exclusionary codes rather than requiring that the code had to have been received in the first month of life as was done for the IOM presentation. Contrary to the suggestion by Halsey et al, we did not make assumptions about the thimerosal contents of the vaccines. We used documented information on vaccine type and manufacturer to determine the thimerosal content of specific vaccines. For example, one of the DTaP vaccines did not contain thimerosal and was assigned a value of zero in calculating mercury exposure. The diagnoses included in the analyses could have been made by any health care provider. We did not restrict to diagnoses made by a specialist. We also did not validate all the diagnoses by reviewing charts. We performed chart review validation for a sub-sample of speech delay, ADD, and autism diagnoses to gauge the validity of these diagnoses. The chart-review findings for these three conditions, however, were not used in the analysis. All the results for all conditions evaluated were based on the ICD-9 codes in the computerized files routinely maintained by the HMOs. Although we did not present gender-specific results in the article, we have performed such analyses. We found no large or systematic differences in effects by gender. We agree that parents of children with mild developmental delay might have been more likely to seek evaluations for minor illnesses and well child visits. This is another example of possible health-care seeking differences between different groups of children and reinforces the need to cautiously interpret the results of our study. We did not mean to discount the positive association at HMO B between thimerosal exposure and language delays because no association was found at HMO C. We do not feel, however, that this result by itself could be considered sufficient to establish a causal association. We are currently conducting a follow-up study that includes children from these same HMOs to more rigorously evaluate possible neurodevelopmental effects of thimerosal exposure. Children are selected according to different levels of thimerosal exposure and undergo an extensive standardized battery of neuropsychological tests that includes a thorough evaluation of speech and language function. The examiners are not aware of the children’s thimerosal exposure level. We believe that the follow-up study will provide more reliable data for drawing causal inferences than the analysis based on limited computerized diagnostic codes. The limitation to outcomes with 50 or more cases was an a priori decision at the time the protocol was written. The decision was based on preliminary power estimates. We appreciate the suggestion to obtain detailed developmental and treatment histories. In the follow-up study that we are conducting, we are reviewing participants’ medical charts and interviewing their parents regarding developmental and treatment histories for all outcomes that we are assessing. The study is powered to detect meaningful differences within the range of thimerosal exposures experienced by infants vaccinated in the 1990s. We would welcome any additional review of our study, but we are uncertain about the value of such a review at this time. The study underwent extensive review internally at the Centers for Disease Control and Prevention, by external experts, by an IOM committee, and by peer reviewers for Pediatrics. Several helpful suggestions from throughout the review process were incorporated into the final analysis. Although other reviewers may be able to suggest improvements to the analysis, we do not think that any analysis would be able to fully overcome the inherent limitations of computerized health services data for evaluating causal associations with complex neurodevelopmental conditions. One of the initial reasons for our study was to provide guidance in selecting conditions to be evaluated in a more rigorous follow-up study that we are currently conducting. In the follow-up neuropsychological testing study, we are evaluating several neurodevelopmental outcomes, including any outcomes that showed any positive associations at any stage of our study from preliminary results to the final published manuscript. The follow-up study is scheduled to be completed in 2004. We suspect that it is unlikely that an independent re-analysis of the computerized HMO data could be completed much before when the more definitive results from the follow-up study are expected to be available.
Study Misses Link Between Thimerosal and Neurodevelopmental Disorders 23 February 2004
Mark R. Geier, Geneticist and Vaccinologist The Genetic Centers of America, David A. Geier of Medcon, Inc. Send letter to journal: Re: Study Misses Link Between Thimerosal and Neurodevelopmental Disorders mgeier{at}comcast.net Mark R. Geier, et al.	Letter to the Editor: The recent article, “Safety of Thimerosal-Containing Vaccines: A Two- Phased Study of Computerized Health Maintenance Organization Databases,” by Verstraeteten et al. [1], which failed to find a consistent association between thimerosal in childhood vaccines and neurodevelopmental disorders, has a number of issues that need to be further addressed. First, the head author, Dr. Thomas Verstraeten, has for the past several years worked for GlaxoSmithKline, a vaccine manufacturer of thimerosal-containing vaccines. In addition, Nancy Pekarek, a company spokeswoman for GlaxoSmithKline, has written that Verstraeten, since leaving the Centers for Disease Control and Prevention (CDC), has worked as an adviser as the study was finalized and prepared for publication. Presently, GlaxoSmithKline, potentially, faces a large number of lawsuits on the very issue that the paper discusses. Second, this very study was the topic of secrete-closed meetings between members of the CDC and other government organizations, as well as members of the vaccine manufacturers held at Simpsonwood, Georgia from 7-8 June 2000. The transcript of this meeting has been obtained under the Freedom of Information Act. This transcript reveals that the study initially found statistically significant dose-response effects between increasing doses of mercury from thimerosal-containing childhood vaccines and various types of neurodevelopmental disorders. The transcript documents that the data was real and statistically significant for many types of neurodevelopmental disorders, but that the meeting participants expressed that the data had to be “handled.” Despite, discussion about how to “handle” the data, some participants expressed concern that the work that had already been done would be obtained by others through the Freedom of Information Act. In this event, even if professional bodies expressed the opinion that there was no association between thimerosal and neurodevelopmental disorders, it was already too late to do anything. In addition, other participants expressed that the vaccine manufacturers were in a horrible position to be able to defend any lawsuits alleging a relationship between thimerosal and neurodevelopmental disorders, since no one would say with the available data that there was no relationship between thimerosal and neurodevelopmental disorders. Even Verstraeten, in an email following the Simpsonwood meeting, expressed surprise that the data was to be manipulated, stating that ones desire to disprove an unpleasant theory should not interfere with sound scientific methods to evaluate the relationship between thimerosal and neurodevelopmental disorders. Third, there are also significant issues about the methods used to determine the mercury dose that children received from thimerosal- containing vaccines. The authors, in Table 1 of their manuscript, completely fail to mention that there were large numbers of thimerosal- free Diphtheria-Tetanus-acellular-Pertussis (DTaP) vaccines administered to children in the Health Management Organizations (HMOs) analyzed. Thimerosal-free DTaP vaccine has been produced by GlaxoSmithKline since 1997. We have personally analyzed the Vaccine Safety Datalink (VSD) database determining that approximately one-third of the children receiving DTaP in the VSD from 1997 through 2000 were immunized with this vaccine, and that the children received thimerosal-free DTaP vaccines in various combinations, with some receiving four doses of thimerosal-free DTaP, some receiving three doses of thimerosal free DTaP and one dose of thimerosal-containing DTaP, some receiving two doses with and two doses without thimerosal, some receiving three with and one without thimerosal, and some receiving all four doses of thimerosal-containing DTaP. In order to evaluate whether Verstraeteten et al., did or did not take this into account, we analyzed Table 1 from their study for the possible cumulative mercury exposures at the various ages of immunization. At one month, the possible mercury exposure was 12.5 micrograms of mercury according to the authors, which is appropriate because there was no potential thimerosal- free DTaP vaccine to take into account. At 2-3 months, the possible cumulative mercury exposure was 37.5-75 micrograms of mercury according to the authors. These potential possible cumulative mercury exposures could be generated by DTP and Hib vaccine separated or combined, or by thimerosal-free DTaP vaccine and Hib (i.e. both DTPH or thimerosal-free DTaP vaccine and Hib vaccine, resulted in children being exposed to 25 micrograms of mercury). At 5-6 months, the possible cumulative mercury exposure was 75 or 125 micrograms according to the authors. The fact that the authors only list these two potential possible cumulative mercury exposure doses show that the authors failed to take into account the thimerosal-free DTaP vaccine made by GlaxoSmithKline, since children receiving one thimerosal-containing DTaP followed by one thimerosal-free DTaP vaccine, in addition to their two doses of hepatitis B vaccine and two doses of Hib vaccine received 100 micrograms of mercury, a mercury dose not mentioned in the table. At 6-7 months, the possible cumulative mercury exposure was 112.5 micrograms of mercury or 187.5 micrograms of mercury according to the authors. These potential possible cumulative mercury exposures show overwhelmingly that there is a significant error in the study. The intermediate mercury values children were exposed to also included: two thimerosal-containing and one thimerosal-free DTaP vaccine, with three doses of hepatitis B vaccine and three doses of Hib vaccine, for a total of 162.5 micrograms of mercury; and two thimerosal-free DTaP and one thimerosal-containing DTaP vaccine, with three doses of hepatitis B vaccine and three doses of Hib vaccine, for a total of 137.5 micrograms of mercury. These calculations indicate that Verstraeteten et al. did not take thimerosal-free DTaP vaccine into account in their study, or if they did, then their paper, as it stands, is replete with inaccurate information. Additionally, the fact that the VSD data contained large numbers of children who took thimerosal-free DTaP vaccine and large numbers of children who took thimerosal-containing DTaP vaccine allows a much more direct and powerful way to do the study by comparing these two groups, since this type of analysis would allow for overall evaluation of the effects of increasing doses of mercury from thimerosal in comparison to considerably lesser doses of mercury from thimerosal. We have done just such a study in VSD and found an association between increasing doses of thimerosal and neurodevelopmental disorders. We have previously epidemiologically examined the Vaccine Adverse Event Reporting System (VAERS) for children receiving thimerosal-containing DTaP vaccines in comparison to thimerosal-free DTaP vaccines and the US Department of Education dataset, and both showed an overall and dose-response statistically significant link between increasing doses of thimerosal and neurodevelopmental disorders [2-5]. It also has been observed that children with autism fail to excrete mercury in their hair and show large increases in the amount of mercury in their urine following chelation therapy in comparison to controls [6,7]. These finding are particularly troubling in light of the fact that many authors including Slikker [8] from the Food and Drug Administration have published that thimerosal crosses the blood-brain and placental barriers and results in appreciable mercury content in tissues including the brain, and because it has been shown by Baskin et al. [9] that micromolar concentrations of thimerosal are capable of causing significant damage to neurons. A recently published report from Northeastern University, the University of Nebraska, the USDA, and the Johns Hopkins University has found that thimerosal at picomolar concentrations is a potent neurotoxin since it inhibits the insulin growth factor-1 and the dopamine-stimulated methlyation synthase pathways providing a potential molecular mechanism of how the link between thimerosal in vaccines and neurodevelpmental disorders, reported in our studies, actually increased the incidence of autism and how thimerosal in vaccines through its interaction with the D4 receptor gene may even account for the increase in ADHD as well [10]. It also is in keeping with the many hundreds of peer-reviewed articles published over many decades and from many fields of medicine and science reporting on the harmful effects of thimerosal in humans, animals, isolated neurons, and other systems. Fourth, there is also a significant issue regarding the inclusion of children who received whole-cell Diphtheria-Tetanus-Pertussis (DTP) vaccine and DTaP vaccine. The Institute of Medicine of the United States’ National Academy of Sciences has determined that the evidence is consistent with a causal relationship between whole-cell DTP vaccine and permanent brain damage [11, 12]. In addition, despite the claim by Verstraeteten et al. that encephalopathies following whole-cell DTP occur only rarely, and therefore, this would be unlikely to have influenced the results of the study, some authors, such as Strom [13] reported that 1 in 6,000 children developed a neurological reaction and 1 in 17,000 children died or were left with a permanent neurological defect, and Pollock and Morris [14] who reported that 1 in 8,500 children died or had a neurological disorder following whole-cell pertussis vaccination. Therefore, it is clear that the assumption by Verstraeteten et al. that whole-cell DTP vaccine would have limited effects upon the results of their study seems incorrect, but rather points to a serious confounder present in their study that makes evaluation of the effect of thimerosal more difficult to discern. In conclusion, because of a number of very serious issues have been raised and the critical importance of the issue as to whether thimerosal causes neurodevelopmental disorders, we respectfully request that Verstraeten et al. consider withdrawing this study. In order to restore the badly damaged confidence in our much needed vaccine program, it is necessity that past errors be admitted, and that open investigations be conducted on vaccines issues. It is also essential that future vaccine decisions are made by physicians and scientists without even the appearance of conflicts of interest. Dr. Mark R. Geier has been a consultant and expert witness in cases involving vaccine adverse reactions before the National Vaccine Injury Compensation Program and in civil litigation. David A. Geier has been a consultant in cases involving vaccine adverse reactions before the National Vaccine Injury Compensation Program and in civil litigation. References 1. Verstraeten T, Davis RL, DeStefano F, et al. Safety of thimerosal- containing vaccines: A two-phased study of computerized health maintenance organization databases. Pediatrics. 2003;112:1039-1048. 2. Geier MR, Geier DA. Neurodevelopmental disorders following thimerosal-containing vaccines: a brief communication. Exp Biol Med. 2003;228:660-664. 3. Geier MR, Geier DA. Thimerosal in childhood vaccines, neurodevelopment disorders, and heart disease in the United States. J Am Phys Surg. 2003;8(1):6-11. 4. Geier DA, Geier MR. An assessment of the impact of thimerosal on childhood neurodevelopmental disorders. Pediatr Rehabil. 2003;6:97-102. 5. Geier DA, Geier MR. A comparative evaluation of the effects of MMR immunization and mercury doses from thimerosal-containing childhood vaccines on the population prevalence of autism. Med Sci Monit. 2004;10(3):PI33-139. 6. Bradstreet J, Geier DA, Kartzinel JJ, Adams JB, Geier MR. A case- control study of mercury burden in children with autistic spectrum disorders. J Am Phys Surg. 2003;8:76-79. 7. Holmes AS, Blaxill MF, Haley BE. Reduced levels of mercury in first baby haircuts of autistic children. Int J Toxic. 2003;22:277-285. 8. Slikker W. Developmental neurotoxicology of therapeutics: survey of novel recent findings. Neurotoxicology. 2000;21:250. 9. Baskin DS, Ngo H, Didenko VV. Thimerosal induces DNA breaks, caspase-3 activation, membrane damage, and cell death in cultured human neurons and fibroblasts. Toxicol Sci. 2003;74:361-368. 10. Waly H, Olteanu H, Banerjee R, et al. Activation of methione synthase by insulin-like growth factor-1 and dopamine: a target for neurodevelopmental toxins and thimerosal. Mol Pyschiatry. (in press). 11. Howson CP, Howe CJ, Finebery HV, eds. Adverse Effects of Pertussis and Rubella Vaccines. Institute of Medicine. Washington, DC: National Academy Press; 1991. 12. Stratton KR, Howe CJ, Johnston RB, eds. DPT Vaccine and Chronic Nervous System Dysfunction: A New Analysis. Institute of Medicine. Washington, DC: National Academy Press; 1994. 13. Strom J. Is universal pertussis vaccination always justified? Br Med J. 1960;2:1184-1186. 14. Pollock TM, Morris J. A 7-year survey of disorders attributed to vaccination in North West Thames region. Lancet. 1983;1:753-757.
Safety of Thimerosal-Containing Vaccines -- Response to Geier, et al 2 March 2004
Frank DeStefano, Medical Epidemiologist Centers for Disease Control and Prevention, Philip H. Rhodes Send letter to journal: Re: Safety of Thimerosal-Containing Vaccines -- Response to Geier, et al fxd1{at}cdc.gov Frank DeStefano, et al.	To the editor: The Centers for Disease Control and Prevention (CDC) and all the co- authors stand behind the science and findings of the study, “Safety of Thimerosal-Containing Vaccines: A Two- Phased Study of Computerized Health Maintenance Organization Databases” (1). Although Geier and Geier try to discredit the study by impugning the integrity of the investigators, they have identified no substantive deficiencies with the study’s methods, analysis, or results (2). Their only substantive comment was about possible inaccurate coding of the thimerosal content of one DTaP vaccine that did not contain thimerosal. We grant that Table 1 in the manuscript could be interpreted as indicating that we may have ignored thimerosal-free vaccines, but this was not the case. We did not make assumptions about the thimerosal contents of the vaccines. We used documented information on vaccine type and manufacturer to determine the thimerosal content of specific vaccines. The table was meant to illustrate the most common combinations of vaccines and amounts of (estimated) mercury received at several ages. There were a vast number of possible vaccine combinations adding up to various amounts of mercury at different ages. A complete table of all possibilities would have been unwieldy for a journal article. The two major vaccines containing DTP, DTaP, HiB or HepB that were mercury-free were the GlaxoSmithKline (GSK) DTaP (HMOB only) and a Merck Hib-HepB combination (HMO A). Both became available late in the study and did not constitute a major portion of vaccines received in the first 7 months of age except for the latter portions of the birth cohorts. There were some additional vaccines used at HMO B in the context of clinical trials that were thimerosal-free, although the numbers of children involved were very small. All of these vaccines were assigned a mercury content of zero in the calculations of cumulative mercury exposure in our study. At all times throughout the conduct of our study we looked for comparisons that would involve well-vaccinated children who had received vaccines with different amounts of thimerosal. The use of a thimerosal- free DTaP produced by GSK was one of several such potential opportunities. However, as noted above, in this instance the number of children and amount of follow-up time involved was insufficient to allow for a useful separate analysis. Our paper did, however, feature the results of another opportunity of this kind, i.e. the switch in usage at HMO B from separate DTP and HIB vaccines to a combined DTPHIB vaccine (1993) and then back to separate DTaP and HIB vaccines (1997) (1, Appendix Table 3). These results indicated that the children receiving the larger amounts of thimerosal did not have a higher risk of developmental delays. The Geiers question the appropriateness of Dr. Verstraeten’s involvement in the study because he was employed by GSK after he left CDC. Dr. Verstraeten worked at CDC during the critical time when the study was designed and the data were analyzed. Once he began working at GSK, his involvement in the study was limited to reviewing drafts of the manuscript. After he left CDC, he did not have any further access to the data and was not involved in any subsequent data analyses. The Geiers express concerns about the deliberations that CDC held with outside consultants at the Simpsonwood retreat center to review preliminary study results. In addition to the Simpsonwood review, the study underwent extensive review internally at CDC, by external organizations including representatives of SafeMinds, by an Institute of Medicine (IOM) committee, and by peer reviewers for Pediatrics. Following many helpful suggestions from other scientists, researchers, and organizations, improvements were made in the databases, research methods, and statistical procedures used to analyze the data. It is accepted and sound scientific practice, especially with complex and important research issues, to seek and use the advice and recommendations from both internal and external reviewers to strengthen a study as much as possible before publishing a final paper. In this case, four major improvements were made as a result of the reviews of preliminary findings: • The methods used to analyze the data were refined and improved. • Errors in the data (e.g., mistakes in medical records, errors regarding the thimerosal content of certain vaccines) were corrected to make the results more accurate. • Suggestions by reviewers were incorporated into the analysis to address and reduce potential biases (e.g., differences in health care seeking behavior among parents that could lead to some children going to their physician more often and therefore being diagnosed more often). • More children with diagnoses of interest were identified as the study progressed and the children at the HMOs became older. The Geiers mention that they have done an analysis of VSD data and found an association between increasing doses of thimerosal and neurodevelopmental disorders. We are not able to comment on this assertion since the study has not been published and they provided no data in support of the statement. Perhaps they may be referring to an analysis that they presented at an IOM Immunization Safety Review Committee meeting on February 9, 2004 in which they compared risk of autism between children who received thimerosal-containing and thimerosal-free DTaP vaccines (3). Our analysis of the same data, however, indicated that the Geiers’ results can be explained by confounding by age rather than any true risk (4). The Geiers’ analyses of the Vaccine Adverse Events Reporting System (VAERS) and U.S. Department of Education data contain numerous conceptual and scientific flaws, omissions of fact, inaccuracies, and misstatements that are detailed in an extensive critique by the Academy (5). The other studies mentioned by the Geiers that measured hair mercury levels, urinary mercury excretion, and in-vitro toxicological effects of ethylmercury may have relevance to the biological plausibility of the hypothesis that thimerosal-containing vaccines could be associated with neurodevelopmental disorders. Some of this evidence had been reviewed in 2001 by the IOM Immunization Safety Review Committee, which concluded at that time that the hypothesis was biologically plausible (6). The more recent studies mentioned by the Geiers were also reviewed at the IOM Immunization Safety Review Committee meeting on February 9, 2004. We will be interested to see how the committee interprets these studies and their implications for the possible association between thimerosal-containing vaccines and neurodevelopmental disorders, including autism. Finally, the Geiers take issue with our interpretation that possible encephalopathy from whole-cell pertussis vaccines would occur too infrequently to have had a material impact on our results. The most comprehensive review conducted on this topic by IOM found that the risk for acute encephalopathy after pertussis vaccination was at most 10.5 per million (7). We stand by our conclusion that encephalopathy after pertussis vaccination is unlikely to have had a meaningful impact on our results. Sincerely yours, Frank DeStefano, MD, MPH Philip H. Rhodes, PhD References: 1. Verstraeten T, Davis RL, DeStefano F, et al. Safety of thimerosal -containing vaccines: A two-phased study of computerized health maintenance organization databases. Pediatrics. 2003;112:1039-1048. 2. Geier MR, et al. Study misses link between thimerosal and neurodevelopmental disorders. Pediatrics – P3Rs for Verstraeten el al., 112:1039-1048. 3. Geier MR, Geier D. Autism and thimerosal-containing vaccines: analysis of the Vaccine Adverse Events Reporting System (VAERS). Immunization Safety Review Committee Meeting 9: Vaccines and Autism. The National Academy of Sciences, Washington, DC, February 9, 2004. 4. Davis RL. Vaccine Safety Datalink Study: Autism Outcome. Immunization Safety Review Committee Meeting 9: Vaccines and Autism. The National Academy of Sciences, Washington, DC, February 9, 2004. 5. Study fails to show a connection between thimerosal and autism. American Academy of Pediatrics 2003. (http://search.aap.org/aap/CISPframe.html?url=http://www.cispimmunize.org/pro/doc/Geiersummary.doc) 6. Stratton K, Gable A, McCormick MC, eds. Thimerosal-Containing Vaccines and Neurodevelopmental Disorders. Institute of Medicine. Washington, DC: National Academy Press; 2001. 7. Howson CP, Howe CJ, Fineberg HV, eds. Adverse Effects of Pertussis and Rubella Vaccines. Institute of Medicine. Washington, DC: National Academy Press; 1991.
False information in DeStefano/Rhodes response to M. Geier 24 March 2004
Brian S. Hooker, Research Scientist Autism Healing Network Send letter to journal: Re: False information in DeStefano/Rhodes response to M. Geier brian{at}dream-big.us Brian S. Hooker	Dear Editor The coauthors of the Verstraeten et al. 2003 Pediatrics study erroneously report in their P3R response to Dr. Mark Geier, that the SAFEMINDS coalition was involved in the review of the different phases of the CDC study currently under web review. Unfortunately, at no time were the SAFEMINDS individuals involved in review of this document, neither were any presumed recommendations by SAFEMINDS ever enacted. This is based on a conversation I had with Lyn Redwood, President of the SAFEMINDS organization (3/23/04). In fact, concerning the CDC study, Ms. Redwood was quoted as follows in Insight on the News - National Issue: 12/23/03 CDC Study Raises Level of Suspicion: Lyn Redwood, a registered nurse, mother of an autistic child and president and cofounder of www.SafeMinds.org (Sensible Action for Ending Mercury-Induced Neurological Disorders), a nonprofit organization dedicated to ending devastation caused by the needless use of mercury in medicines, tells Insight that "there are so many problems with the study, but over time you can see how all the manipulations of the data slowly bring down the signals for neurological disorders. I think they were trying to get lower numbers. It must be very hard to admit that a program that was designed to eradicate infectious disease has resulted in an epidemic of a whole new kind of disease. But to think that we weren't given a choice when the regulators and manufacturers knew these products contained mercury is inconceivable." Redwood says with a sigh, "On a scale of one to 10, I give the CDC study a big fat zero. I think it started out good, but when they saw the early numbers it scared the hell out of them. I don't have any faith in the CDC doing a decent study of this matter. It's like having the tobacco industry monitor cigarettes for safety. From a parent's perspective and from a health-care professional's perspective it's maddening that we can't get products that are safe, and yet we're forced by law to use them. They need to just get the thimerosal out. It's barbaric." (Insight on the News - National Issue: 12/23/03 CDC Study Raises Level of Suspicion By Kelly Patricia O Meara) Thus, the information that Dr. DeStefano and Dr. Rhodes posted on the Pediatrics P3R website on 3/2/04 (in response to Dr. Mark Geier's peer-reviewed comment) is false. I would like to give the investigators the benefit of the doubt and say that this misstep was merely an error, but unfortunately, it is completely clear that this misrepresentation was intentional. Given the track record of this publication and the evident gross institutional and commercial conflict of interest, it would be best if the entire submittal were withdrawn as soon as possible.
The Safety of Thimerosal-containing Vaccines -- Response to B.S. Hooker 26 March 2004
Frank DeStefano, Medical Epidemiologist Centers for Disease Control and Prevention Send letter to journal: Re: The Safety of Thimerosal-containing Vaccines -- Response to B.S. Hooker fxd1{at}cdc.gov Frank DeStefano	In a previous response to a comment on our article, we stated: “… the study underwent extensive review internally at CDC, by external organizations including representatives of SafeMinds, by an Institute of Medicine (IOM) committee, and by peer reviewers for Pediatrics. Following many helpful suggestions from other scientists, researchers, and organizations, improvements were made in the databases, research methods, and statistical procedures used to analyze the data.” The statement is clear that the reviews were conducted during preliminary stages of the analysis and that the suggestions of a number of reviewers were incorporated into the final analysis. Ms. Redwood is the President of SafeMinds. She spent the better part of a day at CDC reviewing preliminary analyses and results with CDC researchers working on the study. The analyses of cumulative exposure by 7 months of age in the published manuscript were included at her request.
Re: The Safety of Thimerosal-containing Vaccines -- Response to B.S. Hooker 30 March 2004
Brian S. Hooker, Ph.D., P.E., Research Scientist Autism Healing Network Send letter to journal: Re: Re: The Safety of Thimerosal-containing Vaccines -- Response to B.S. Hooker brian{at}dream-big.us Brian S. Hooker, Ph.D., P.E.	In deference to my gracious colleagues at SafeMinds, I hereby retract my previous P3R submittal of March 24, 2003.
The Safety of Thimerosal containing vaccines: Response to Destefano 31 March 2004
Lyn H. Redwood, RN, MSN, CRNP Pres. Safe Minds Send letter to journal: Re: The Safety of Thimerosal containing vaccines: Response to Destefano tlredwood{at}mindspring.com Lyn H. Redwood	Re: Verstraeten, T, Davis, RL, DeStefano F, et. al. Safety of thimerosal containing vaccines: a two phased study of computerized health maintenance organizations databases. Pediatrics 2003; 112: 1039-1048 and postings on “Post-publication Peer Review” (P3R) site. The Coalition for SafeMinds (SafeMinds) has reviewed the above referenced article and subsequent postings on the Academy’s P3R online site and wish to respond to statements made regarding our organizations involvement in this investigation. First, in regards to Dr. DeStefano’s comment “...the study underwent extensive review internally at CDC, by external organizations including representatives of Safe Minds, by an Institute of medicine (IOM) committee, and by peer reviewers for Pediatrics. Following many helpful suggestions from other scientists, researchers, and organizations, improvements were made in the databases, research methods, and statistical procedures used to analyze the data.” The suggestion here that the authors made good faith efforts to seek reviews from and respond to concerns by “external organizations” such as Safe Minds could not be further from the truth. Although Safe Minds addressed selected question to the CDC regarding the VSD analysis when it was under review by the Institute of Medicine in 2001 and later conducted our own independent investigations of the CDC’s research methods, we were in no respect invited reviewers. Our comments were never solicited, the authors’ responses to our questions were perfunctory and our most salient concerns have never been addressed. The suggestion to the contrary is misleading in the extreme. Taking the authors’ intent at face value, however, we submit that it is not yet too late to subject the VSD research to the kind of open review to which they claim it has (or perhaps should be) submitted. We propose that the CDC 1. Submit the results of the VSD analysis to examination in a full, fair and open review discussion session including representatives from the CDC, Safe Minds and other independent experts in epidemiology and toxicology; 2. Submit to an external audit of the research ethics and supervisory protocols involved in preparing sequential versions of the VSD research for public review and ultimate publication; 3. Open the VSD database to full, unconstrained, and unmonitored examination by independent investigators; and 4. Disclose any and all publication peer review commentary, including reviews from Pediatrics and any others journals to which any earlier drafts may have been submitted. We realize these are unusual steps. Nevertheless, in light of the manifest need for legitimacy in the face of the controversy surrounding this work, we believe unusual openness is required. We look forward to the CDC’s response. For those who wish to review our formal critique of Verstraeten et al, it may be viewed at: http://www.safeminds.org/verstraeten/VSD.SafeMinds.critique.pdf Respectfully submitted, Lyn Redwood, RN, MSN, CRNP President, The Coalition for SafeMinds 14 Commerce Drive, PH Cranford, NJ 07016
Safety of Thimerosal-Containing Vaccines -- Response to L. H. Redwood 5 April 2004
Frank DeStefano, Medical Epidemiologist Centers for Disease Control and Prevention Send letter to journal: Re: Safety of Thimerosal-Containing Vaccines -- Response to L. H. Redwood fxd1{at}cdc.gov Frank DeStefano	As we have stated previously, the study underwent extensive review internally at the Centers for Disease Control and Prevention, by external experts, by an IOM committee, and by peer reviewers for Pediatrics. At each step of the review process, we carefully considered the suggestions of all reviewers and incorporated changes into the analysis that would result in the most scientifically valid and unbiased comparison of the risk of neurodevelopmental outcomes according to different levels of exposure to thimerosal-containing vaccines. We stand behind the science and the findings of the study. We regret that Ms. Redwood feels that we did not adequately address her and SafeMinds’ concerns. We spent considerable time with her individually reviewing the preliminary analyses. The fact that we incorporated an entire set of additional analysis (i.e., associations according to cumulative exposure by 7 months of age) at her request indicates that we gave her suggestions more than perfunctory consideration. Regarding access to the data, in 2002, CDC's Vaccine Safety Datalink (VSD) Project established a data sharing program to allow external researchers to conduct new studies of vaccine safety or to reanalyze study -specific datasets from published VSD studies. In compliance with federal regulations, access by external researchers to the VSD data files or to datasets from VSD published studies requires review and approval by the appropriate institutional review boards (IRBs) of the relevant HMOs. The IRBs have the responsibility to protect the confidentiality and privacy of their members’ medical records and to adhere to the rules and regulations of their specific institution; therefore, each of the HMO IRBs must review any request for access to the VSD data files that contain information on its HMO members. External researchers access the approved VSD data at the Research Data Center (RDC) located at the National Center for Health Statistics (NCHS) in Hyattsville, Maryland. Further details about the VSD data sharing process can be found at http://www.cdc.gov/nip/vacsafe/vsd/VSDGuidelines.doc. As we have previously stated, we would welcome any additional review or re-analysis of our study, but we are uncertain about the value of such a review at this time. Although other reviewers may be able to suggest improvements to the analysis, we do not think that any analysis would be able to fully overcome the inherent limitations of computerized health services data for evaluating causal associations with complex neurodevelopmental conditions. One of the initial reasons for our study was to provide guidance in selecting conditions to be evaluated in more rigorous studies that we are currently conducting. In the follow-up neuropsychological testing study, we are evaluating several neurodevelopmental outcomes, including any outcomes that showed any positive associations at any stage of our study from preliminary results to the final published manuscript. The follow-up study is scheduled to be completed this year. We believe that our and everyone else’s energies would be most constructively spent by focusing on completing the current studies that have been specifically designed to provide more conclusive evidence on possible associations between thimerosal-containing vaccines and neurodevelopmental outcomes.