Array

Cause or Coincidence

Updated: Agosto 28, 2006

As with any drug, there are risks and side effects with vaccines, although serious side effects are mostly rare. Side effects are symptoms and signs that occur either locally—such as pain or redness at the injection site—or which may be systemic—such as headache or fever—that are known to follow a particular immunization or administration of drugs. We use the term “adverse event” to mean something that occurred at about the time a vaccine was given, which could have been caused by the vaccine or could have just have occurred by coincidence.

Thus, when an adverse event occurs after vaccination, it needs to be determined whether the adverse event was caused by the vaccine or whether it was just coincidental in time with the administration of the vaccine—that is, it was going to happen anyway.

For example, many vaccines are given to children at the ages when developmental and other problems are being recognized for the first time. Because something happened at about the same time that a vaccine was given does not mean that the vaccine caused the problem.

Although vaccines have saved millions of lives around the world, some have blamed them for causing conditions that are not completely understood despite the fact that there is no scientific evidence that the vaccines caused the condition—for example, asthma, autism, diabetes type 1, multiple sclerosis, and sudden infant death syndrome (SIDS) among others.

How can researchers determine if a vaccine causes a particular adverse event or not?

Establishing causality

It is often difficult, time consuming and expensive to answer these types of questions. However, as vaccines are recommended for all children as they are developing, it is critical that the studies be done.

There are a number of factors that are considered in trying to determine whether a vaccine is the cause of an adverse event or disease.12

  • Time of onset: The onset of the disease must follow vaccination. If symptoms of the disease occurred before vaccination, then the vaccine is not the cause.
  • Virus isolation: In the case of a live virus vaccine, a causal relationship between vaccine and disease may often be inferred if the virus is recovered from a normally sterile body site (for example, blood or cerebral spinal fluid).
  • Uniqueness of the clinical syndrome: Causation may be inferred if the adverse event or disease only occurs following vaccination and does not occur in persons who did not receive the vaccine. Causation may also be inferred if the adverse event occurs a second time after repeat exposure to the same vaccine.
  • Biological mechanism: Biologic mechanisms that could potentially explain how the vaccine might cause the adverse event are not sufficient to prove that the vaccine caused the problem. However, when there is other evidence of an association, such as epidemiologic studies, biologic mechanisms that could explain the association scientifically could help in establishing whether the vaccine caused the problem.
  • Epidemiological studies: Epidemiological studies often provide the most important evidence as to whether a vaccine caused the problem because most adverse events are not unique clinical syndromes (that is, they also occur in people who did not receive the vaccine). Epidemiological studies determine whether the risk (rate) of the illness of concern is higher in vaccinated persons compared to unvaccinated persons. A higher risk among vaccinated persons could mean that the vaccine caused the problem.

Epidemiology and Causation

Epidemiology is the study of how disease is distributed in a population and of the factors that influence or determine this distribution. Epidemiology helps to identify the causes and risk factors of a disease in a community. Causality is often inferred from epidemiological studies by using the following criteria:

  • Strength of association: The greater the difference in rates between the vaccinated and unvaccinated, the more likely there is a causal relationship.
  • Consistency of association: The more studies that show similar results using different populations and differing study methods, the more likely there is a causal relationship.
  • Dose response: When it can be shown that there is increasing risk for the adverse event with increasing dose, the more likely there is a causal relationship. This is usually more of a consideration for chemicals or drugs and is not often relevant for vaccines.

Epidemiological studies are useful to identify the cause of a disease in the general population. It is harder, however, for epidemiology to find the cause of rare diseases in a small percentage of the population; and epidemiology cannot prove the negative—that means it cannot prove that a vaccine does not cause a disease.

An Example: Rotavirus (RotaShield®) vaccine

The process of evaluating whether a particular vaccine causes an adverse event can start with repeated observations or reports of an adverse event. Then researchers conduct epidemiological studies to determine the risk of the event for vaccinated people compared to unvaccinated people. Still other researchers will try to determine potential biologic mechanisms for the vaccine to cause the event.

This process can be illustrated with a rotavirus vaccine, RotaShield®, licensed by the Food and Drug Administration (FDA) in August 1998.3 In July 1999, after approximately 1 million children had been immunized with the vaccine, the Centers for Disease Control and Prevention (CDC) recommended that use of the rotavirus vaccine be temporarily suspended. The CDC was concerned that the vaccine might be causing a serious bowel disease called “intussusception,” since 15 cases of this condition had been reported in children who had received the vaccine. Subsequent epidemiological studies reported that an infant’s risk of developing intussusception indeed increased after receiving RotaShield®.

  • Four different epidemiologic studies employing very diverse methodologies demonstrated an increased risk of intussusception occurring predominantly during the first week after the first dose of the vaccine. The timing of the adverse event, that it was scientifically plausible, and the strength and consistency of the evidence led scientists to conclude that there was a causal association.3 This example shows the types of evidence used to establish causality.
  • Although there is consensus among scientists about the causal association, some have hypothesized that vaccine triggered intussusception in intussusception-prone infants—that is, these children might have developed intussusception later in infancy anyway.4 This example demonstrates the complexity of interpreting causality associations.

The vaccine was withdrawn from the market by the manufacturer in October 1999.

Coincidental Associations

While epidemiologic studies can help establish a causal association between a vaccine and an adverse event, they cannot prove with absolute certainty that an adverse event that follows immunization represents only a coincidence. These tools can only infer that coincidence is the most likely explanation.

Epidemiologic studies cannot absolutely prove coincidence (reject causation) because there can always be very rare occurrences that were not detected in the study population, or because the vaccine only accounted for a very small proportion of the adverse events. Thus, the strongest interpretation that can be made from epidemiologic studies is that the evidence favors rejection of causation when the risk for vaccinated children cannot be distinguished from the risk for unvaccinated children.

Thus, the Institute of Medicine (IOM) uses two categories for favoring causation—one favoring causation—the evidence is supportive but not definitive—and one establishing causation—the evidence is definitive. On the other hand, because epidemiologic evidence can never be 100% certain, there is only one category for favoring coincidence—that is “the evidence favors rejection of a causal relationship”.

For example, in the case of vaccines and autism, multiple large epidemiological studies done in different countries and using different methods provide a preponderance of evidence that “favors rejection of a causal relationship between the MMR vaccine and autism” as well as “favors rejection of a causal relationship between thimerosal-containing vaccines and autism.”5

Another element that may help infer coincidence is the finding of alternative causes for the disease or absence of the disease. For example, a few years ago it was noticed that as rates of hepatitis B immunization at birth were increasing, cases of SIDS were decreasing. However, the hepatitis B vaccine was not protecting against SIDS; increasing use of the vaccine was just coincidental with a successful campaign aimed at parents to place their infants to sleep on their backs.6

References