Letters

Vaccines and their real or perceived adverse effects

Authors' conclusions are at odds with investigators'

Studies of adverse effects of vaccination have duty to present full picture

Authors' reply

Authors' conclusions are at odds with investigators'

EDITORIn responding to Jefferson's editorial about vaccination and its adverse effects Classen and Classen say that present immunisation schedules may exacerbate the development of type 1 diabetes and that early immunisation with a cocktail of vaccines may prevent the development of diabetes. ^{1 2}

The authors cite their unpublished reanalysis of the Finnish trial of Haemophilus influenzae type b vaccine.³ Their conclusions are at odds with the analyses and conclusions of the investigator (JT) presented at a National Institutes of Health workshop and were published recently.⁴ The Finnish investigators' review of the data could not attribute the risk of type 1 diabetes in Finnish children to differences in the timing of childhood vaccination in the vaccine trial. The incidence of type 1 diabetes in Finland has risen over several decades, and the risk seems to have been almost linear since the early 1950s. Given the disparity of conclusions, it was suggested that Classen and Classen's analysis be sent to an independent statistician. We are not aware that this has been done.

The authors quote the aim of the National Institutes of Health workshop as being "to discuss our data" when in fact the title was "an evaluation of the possible role of vaccines and infectious diseases in insulin dependent (type 1) diabetes mellitus." Two independent reviews of available data specific to vaccines were presented by the Cochrane reviewers and the Johns Hopkins Vaccine Safety Institute; neither found an association such as that reported, and both indicated concerns about methodological issues in statistical analysis and the design and conduct of these studies. The conclusion of the workshop, presented in June 1998, was that studies in humans do not indicate an increase in type 1 diabetes attributable to any vaccine or the timing of immunisation.

Dr J B Classen has filed a European and American patent on his schedule that may have widespread implications should his views be implemented. He applies for a patent on immunisation schedules to be administered from birth, both to reduce the likelihood of type 1 diabetes developing and to protect individuals from communicable diseases.⁵ The safety or efficacy of one particular paediatric schedule (annex 5 of the application⁵) has not been shown.

The applications can be obtained for a small fee from the Search and Advisory Service, Patent Office, Cardiff Road, Newport, Gwent NP9 1RH. The full list of references for this letter is on the BMJ's website.⁶ We agree that the public should be made aware of what is going on.

T O Jefferson, Honorary visiting fellow. 
Cochrane Centre, Oxford OX2 7LG toj1{at}aol.com

Regina Rabinovich, Chief. 
Clinical and Regulatory Affairs Branch, Division of Microbiology and Infectious Diseases, National Institute of Allergy and Infectious Diseases, Bethesda, MD 20892, USA RR28k{at}nih.gov

Jaako Tuomilehto, Research professor. 
Diabetes and Genetic Epidemiology Unit, National Public Health Institute, 00300 Helsinki, Finland jaakko.tuomilehto{at}ktl.fi

Studies of adverse effects of vaccination have duty to present full picture

EDITORThe letter from Classen and Classen has the potential to raise concerns about the long term effects of vaccination but is grossly misleading.¹ There is now a considerable body of literature on the relation between diabetes and vaccination. While both have increased, there is no convincing evidence to suggest that there is a causal, as opposed to temporal, relation.

The Classens describe a collaborative study with Professor Jaako Tuomilehto looking at the effect of Haemophilus influenzae type b vaccine on the incidence of diabetes. The implication is that this was a joint study. I do not know the details, but this was certainly not the case. Dr John Classen presented his as yet unpublished work at a seminar on 20 March 1998 in Baltimore, Maryland. He suggested that the incidence of diabetes was lower in children who received a single dose of the vaccine at 14 months than in those who received doses at 3, 4, 6, and 14 months. The workshop panel concluded that the analytical methods were incorrect. Furthermore, data were available from Professor Tuomilehto showing that follow up over 10 years showed no difference in the incidence of diabetes between children who had received one dose of vaccine and those who had received four doses. The workshop panel examined evidence from several sources and concluded that "there is no evidence that any vaccines have increased the risk of type 1 diabetes in animals or humans."

Those reporting studies of the adverse effects of vaccination, whatever their conclusions, have a duty to present the full picture. We applaud the sentiments underlying the Classens' letter but regret that they didn't live up to them.

Helen Bedford, Research fellow. 
Department of Epidemiology and Public Health, Institute of Child Health, London WC1N 1EH h.bedford{at}ich.ucl.ac.uk

David Elliman, Consultant in community child health. 
St George's Hospital, London SW17 0QT

Authors' reply

EDITORJefferson et al confuse readers about our research. Haemophilus influenzae type b immunisation was initiated in Finland to prevent 7 deaths and 7-26 cases of severe disability (mean benefit 23.5 children) per 100 000 immunised.¹ Our analysis on a collaboration monitoring Finnish data found that children receiving 4 doses of the vaccine had a higher rate of type 1 diabetes than did unimmunised children by more than twice the mean number of children expected to benefit from immunisation. Jefferson et al attempt to downplay this, stating that type 1 diabetes is rising in Finland. In fact, the incidence of type 1 diabetes in ages 5-9 was stable from 1983 to 1992 until the children vaccinated with H influenzae type b reached this age and type 1 diabetes rose by a rate exceeding the mean benefit of the vaccine. We conclude that the potential risk exceeds the potential benefit, showing the need for safer immunisation technology.

The Cochrane review that Jefferson et al mention omits important references, including a preliminary study by the Centers for Disease Control and Prevention. This study supports our data that hepatitis B immunisation starting after 2 months is associated with a more than 60% rise in type 1 diabetes. The authors criticise us for not having our analysis peer reviewed, but our collaborators asked us not to publish the data to allow all researchers to review them. Several trials of the vaccine were performed in Finland, and we are completing the analysis of these.

Jefferson et al misrepresent us by citing patents drafted by lawyers. We believe that starting immunisation in the first month will optimise the trade off between preventing infection and inducing autoimmunity.² Studies are feasible to show that immunisation starting at 4 weeks compared with immunisation starting after 8 weeks will decrease the risk of type 1 diabetes.³ Killed vaccines are routinely given in the United States as early as 6 weeks, and starting at 4 weeks is acceptable to many. Finnish researchers indicate that H influenzae type b immunisation may begin at 4 weeks,⁴ which may decrease the risk of type 1 diabetes. It is feasible to compare the effect of hepatitis B and BCG vaccines on type 1 diabetes when given starting at birth and starting after 8 weeks since these are routinely given then.

French officials stopped hepatitis B immunisation of schoolchildren while continuing neonatal immunisation, reportedly because of the increased risk of autoimmune neurological diseases after immunisation of schoolchildren. French courts have demanded that health officials make safety a priority. Will Jefferson et al remain opposed to our research when the United Kingdom adopts this policy?

In answer to Elliman's letter, we will comment on the meeting in Baltimore when we see a published report from it. To our knowledge no consensus was ever reached. There are clearly differences of opinion regarding safety of vaccines. American law prohibits the marketing of pharmaceutical products if safety has not been demonstrated. Our research has shown that this criterion has not been met with the H influenzae vaccine, so changes are clearly necessary. Some public health officials believe in promoting the vaccine until they are convinced that it is causing harm, but this is not consistent with American law. American laws may not be accepted in many countries, and the risk-benefit of immunisation will be different in developing countries. We believe that our analysis of the data is consistent with American law though understand that many may oppose this standard. A complete list of references and a longer reply is at http://vaccines.net.

John Barthelow Classen, President. 
Classen Immunotherapies, 6517 Montrose Avenue, Baltimore, MD 21212, USA classen{at}worldnet.att.net

David C Classen, Infectious diseases physician. 
Division of Infectious Diseases, LDS Hospital, Salt Lake City, UT, USA

^a Competing interests: John Classen holds shares in Classen Immunotherapies.

^b Competing interests: None declared.