The committee acknowledges that some readers may have concerns about two aspects of the report. First, the committee does not make conclusions about how frequently vaccine adverse events occur. Secondly, the committee concluded, for most analyses, that the evidence is inadequate to accept or reject a causal relationship and some readers might interpret the committee’s language in different and inaccurate ways. The committee offers concluding comments to address these two issues.
This report is not intended to answer the question “Are vaccines safe?”. The committee was not charged with answering that question. Other bodies make that determination and contribute to ongoing safety monitoring, including governmental agencies, care providers, and industry, as they determine the benefits and risks of marketing a product. At all levels, policy determining vaccine use requires a balancing of risks and benefits. As described in Chapter 1 and the Preface, that is outside the bounds of this committee’s assignment. It should also be noted that where the committee has found evidence of a causal relationship, it does not make conclusions about the rate or incidence of these adverse effects.
Determining the rate of specific adverse events following immunization, in the general population or a subset thereof, is challenging. It would be possible, for example, to estimate a rate of the occurrence of a specific adverse effect in a vaccinated population or susceptible subgroup of interest. This could be done using a summary relative risk or absolute risk difference (e.g., estimated from a set of consistent reports reviewed by the committee) if there were large population-based studies of the occurrence of the adverse event in unvaccinated individuals (e.g., in the general popu-
lation or susceptible subgroups of interest) who do not substantially differ from those vaccinated on any known, important confounders (e.g., age and exposure to other vaccinations or other agents or factors known to cause the adverse event). None of these preconditions is fully met for the adverse events reviewed in this report.
The committee also notes here that large epidemiologic studies that report no cases of the adverse event of interest in vaccinated study participants, if included in our analyses, raise particular concerns. If at least some cases of the adverse event occurred in a study’s unvaccinated comparison population, an upper limit of the 95% confidence interval (CI) for the study’s relative risk or absolute risk difference could be estimated, but one would be unable to rule out a possibly increased risk unless the vaccine was significantly protective against that particular adverse effect. Also, including such studies may have exacerbated problems with detection biases unless precautions were taken to ensure equal surveillance for the adverse event in the unvaccinated and vaccinated populations being compared.
Discussion of the adverse events where the committee concluded that there is evidence to support causation illustrates more fully the challenge of specifying rates, although for some estimates can be provided.
MMR vaccine: The committee concluded that the evidence favors acceptance of a causal relationship between measles, mumps, and rubella (MMR) vaccine and febrile seizures. Approximately 4 percent of children will experience a febrile seizure by 5 years of age (Marin et al., 2010). Fever may occur following MMR vaccination, and some children who have fever following MMR may have a febrile seizure. It is important to note that simple febrile seizures are benign and have no permanent sequelae. For example, children with simple febrile seizures have no greater chance of getting epilepsy or experiencing long-term brain damage than children who do not have febrile seizures.
Three of the studies the committee examined provided both a number of children vaccinated with MMR (the denominator) and the number of febrile seizures considered to be attributable to MMR (Farrington et al., 1995; Griffin et al., 1991; Vestergaard et al., 2004). Children who receive the MMR vaccine are at risk for febrile seizures 8–14 days after vaccination (Marin et al., 2010). About one additional febrile seizure occurs during the 30 days after vaccination among every 3,000–4,000 children who receive MMR vaccine, compared with children who are not vaccinated (Marin et al., 2010). Using the number of febrile seizures attributed to MMR vaccine, and dividing by the number of children in the cohort, each of the other studies provides a similar rate, between one in 1,000 and one in 4,000 doses.
Varicella vaccine: The varicella vaccine accounted for five of the affirmative causality conclusions. All were caused by infection of persons with
the varicella vaccine strain, usually in immunodeficient persons. Varicella vaccine is a live virus vaccine that is contraindicated in people with known, severe immunodeficiency, including severe combined immunodeficiency, other congenital immunodeficiencies, and immunodeficiency arising from long-term immunosuppressive therapy or from chemotherapy for hematologic or solid tumors. The evidence for the causal relationships for adverse events from infection by the vaccine virus came from case reports, so there was no cohort or background population to allow calculation of a rate, even among the population of people who have demonstrated immunodeficiencies.
Although the lifetime rate of shingles in the population has been estimated (9–10 percent for those under 45 years of age, 22–32 percent in older persons) (Chapman et al., 2003), the rate of shingles and other infection-related adverse events associated specifically with the varicella vaccine virus are not known for several reasons. First, while the rate of shingles can be estimated (see Chapman, above), in most cases the virus is not characterized, meaning no test is done to determine whether the virus is wild or vaccine type. Second, while the Food and Drug Administration (FDA) and the Centers for Disease Control and Prevention (CDC) oversee a large database of reports, the Vaccine Adverse Event Reporting System (VAERS), those reports are often incomplete and do not always have the information that would document the vaccine strain or the presence or absence of immunodeficiency. However, it appears likely to the committee that the risk of vaccine-strain varicella infection and subsequent serious disease to persons demonstrated to be immunocompetent is exceedingly low, while the risk to those with severe immunodeficiency is real, which is what the CDC and FDA have concluded by deciding that varicella vaccination is contraindicated in such persons. And, of course, immunocompromised individuals benefit greatly from a high level of immunity to varicella within the community.
Anaphylaxis: Although it is also difficult to estimate rates for very rare conditions, the committee concluded that evidence supports the association of anaphylaxis with certain vaccines in certain circumstances, but the number of events related to each specific vaccine is not known. Rates can be estimated from surveillance studies, but often specific details are missing, and each case cannot be linked with certainty to vaccine. For example, Bohlke et al. (2003) (an uncontrolled study from the Vaccine Safety Datalink [VSD] network) reported three cases of anaphylaxis after administration of 848,945 doses of MMR vaccine. However, two of those children received vaccines in addition to MMR, and the confidence interval for the calculated rate per million doses was very wide (rate per million doses = 3.5, 95% CI, 0.7–10.3). Lastly, regarding this example of a rare condition, not only is the number of true anaphylactic reactions to vaccines not known, but also
the “denominator” of persons susceptible to anaphylaxis (rather than the general population of persons to be vaccinated) is unknown.
Anaphylactic reactions to several vaccines are likely caused by the presence of components introduced during manufacturing, such as egg protein, milk protein, or gelatin. When a specific inciting component of the vaccine has been identified and the manufacturers find ways to remove or drastically reduce the amount of the reactive antigen (e.g., egg protein in influenza vaccine), the number of reports of anaphylaxis in spontaneous reporting systems has decreased. It appears likely to the committee that the risk of anaphylaxis caused by vaccines is exceedingly low in the general population. The risk is obviously higher in people with known and demonstrably severe allergies to certain vaccine components, such as eggs or gelatin.
An affirmative finding for causality was determined for a very mild condition (oculorespiratory syndrome) subsequent to certain influenza vaccines used only in two seasons in Canada. The committee made no attempt to determine the rate of this condition.
Finally, the committee determined that evidence supported an association with what the committee considered to be injection-related events: deltoid bursitis and syncope. These injection-related events are known to be caused by many things other than vaccine administration and are likely often unreported. Estimates of the rates caused by vaccination are similarly not available, as population-based studies have not been conducted.
The seriousness of any particular adverse effect is a complex question, taking into account such factors as the degree and duration of disability and the type of health care needed as a result, recognizing that any individual who experiences an adverse effect may regard it as serious. All of these considerations have social and ethical components as well. Deeming this calculus to be too complex to define with particularity, the committee elected to defer to common understanding within the health care community for assessment of the seriousness of any particular adverse effect.
An issue that is likely to be of concern to some readers regards the very stringent approach our committee has taken. For the majority of adverse events the committee was asked to examine, the committee concludes that the evidence is inadequate to accept or reject a causal relationship. Some might interpret that to mean either of the following statements:
- Because the committee did not find convincing evidence that the vaccine does cause the adverse event, the vaccine is safe.
- Because the committee did not find convincing evidence that the vaccine does not cause the adverse event, the vaccine is unsafe.
Neither of these interpretations is correct. “Inadequate to accept or reject” means just that—inadequate. If there is evidence in either direction that is suggestive but not sufficiently strong about the causal relationship, it will be reflected in the weight-of-evidence assessments of the epidemiologic or the mechanistic data. However suggestive those assessments might be, in the end the committee concluded that the evidence was inadequate to accept or reject a causal association.
The committee does want to emphasize many of the adverse events examined are exceedingly rare in the population overall, and in most instances any particular adverse event, be it arthritis, meningitis, or any of the other vaccine–adverse events that the committee considered, are not preceded by immunization. The committee chose cautious and scientific language for our conclusions, because, especially with rare events, it is not possible to prove a negative (i.e., the vaccine did not and cannot cause the event). The committee cannot say that in a certain person at a certain time, some event cannot happen; there is much about biology that is not known.
The committee tried to apply consistent standards when reviewing individual articles and when assessing the bodies of evidence. Some of the conclusions were easy to reach; the evidence was clear and consistent or, in the other extreme, completely absent. Some conclusions required substantial discussion and debate. Inevitably, there are elements of expert clinical and scientific judgment involved.
The committee used the best evidence available at the time. The committee hopes that the report is sufficiently transparent such that when new information emerges from either the clinic or the laboratory, others will be able to assess the importance of that new information within the approach and set of conclusions set forth in this report.
The committee hopes this summary of the thinking of the committee is helpful to the reader.
Bohlke, K., R. L. Davis, S. M. Marcy, M. M. Braun, F. DeStefano, S. B. Black, J. P. Mullooly, and R. S. Thompson. 2003. Risk of anaphylaxis after vaccination of children and adolescents. Pediatrics 112(4):815-820.
Chapman, R. S., K. W. Cross, and D. M. Fleming. 2003. The incidence of shingles and its implications for vaccination policy. Vaccine 21(19-20):2541-2547.
Farrington, P., S. Pugh, A. Colville, A. Flower, J. Nash, P. Morgan-Capner, M. Rush, and E. Miller. 1995. A new method for active surveillance of adverse events from diphtheria/ tetanus/pertussis and measles/mumps/rubella vaccines. Lancet 345(8949):567-569.
Griffin, M. R., W. A. Ray, E. A. Mortimer, G. M. Fenichel, and W. Schaffner. 1991. Risk of seizures after measles-mumps-rubella immunization. Pediatrics 88(5):881-885.
Marin, M., K. R. Broder, J. L. Temte, D. E. Snider, and J. F. Seward. 2010. Use of combination measles, mumps, rubella, and varicella vaccine: Recommendations of the Advisory Committee on Immunization Practices. Morbidity & Mortality Weekly Report 59(RR03):1-12.
Vestergaard, M., A. Hviid, K. M. Madsen, J. Wohlfahrt, P. Thorsen, D. Schendel, M. Melbye, and J. Olsen. 2004. MMR vaccination and febrile seizures: Evaluation of susceptible subgroups and long-term prognosis. Journal of the American Medical Association 292(3):351-357.