VAERS stands for the Vaccine Adverse Event Reporting System - it is run by the CDC (Centers for Disease Control) as one method of monitoring vaccine safety. This is a fairly basic and standard practice, similar to the less formal reports of drug adverse events to the FDA after drugs are on the market.
Such voluntary reporting systems are an early warning system. Their purpose is to indicate a possible new trend that might indicate a previously unrecognized side effect or risk. Such reporting systems are also important because drugs are typically studies in thousands of individuals before going to market, but then they may be used by orders of magnitude more people, perhaps millions. Therefore there may be side effects or risks that are statistically too small to show up in studies of thousands of people, but will show up when given to millions.
Such systems are also used as part of the precautionary principle. The FDA, for example, may place a warning on a drug, or even remove it from the market, based upon a possible association with an adverse outcome, even if we can't be sure of a real association or a cause and effect.
What voluntary reporting systems are not, however, are scientifically rigorous assessments of true risk. They are useful for generating, but not testing, hypotheses.
The primary weakness of voluntary reporting systems, like VAERS, is that they are voluntary - people take it upon themselves to report what they believe may be a vaccine side effect. Such reporting is therefore subject to reporting bias. A news story warning about the risks of the flu vaccine will result in a spike in VAERS reports of flu vaccine side effects.
Any trend in VAERS reporting, therefore, has to be interpreted with caution, and in fact cannot be scientifically interpreted. A possible signal in VAERS would need to be followed up with some rigorous data - systematically looking at the incidence of the alleged side effect in a population and correlating it with vaccine status, for example.
A recent article in Mothering Magazine however, ignored all this and treated VAERS reporting as if it were scientific data. The author, Stephen Rubin PhD, analyzed recent VAERS data and interpreted the results entirely as if trends represented actual trends in side effects, rather than potentially artifactual trends in reporting. The result was an irresponsible piece of fear-mongering.
First of all, it should come as no surprise that reports of vaccine side effects correlate with getting vaccines. Why would someone think that they or their child had a vaccine side effect if they didn't recently get a vaccine? It is also easy to imagine many artifacts in the reporting. Perhaps parents (especially new parents) are more likely to worry about vaccine side effects (and report them) after their child gets their first series of vaccines, but are more mellow with later vaccines. We should also expect that ages at which other infantile or childhood diseases occur with result in spikes in VAERS reporting as they are misinterpreted as vaccine side effects.
It's important to note that VAERS accepts all reports - reports are not filtered based upon any assessment of how plausible they are or whether or not they are likely to be a real vaccine effect. Anyone can report anything and it goes into the database.
Rubin writes, for example:
"The graph shows a serious spike in Influenza-related reports during the last two years. Suddenly, there are about three times as many adverse events being reported following Flu shots. Of course, concerns about the H1N1 Flu caused more people to be vaccinated during these years, but were there really three times as many people vaccinated? The numbers do not show that."
"Something is causing an increase in the number of reactions to the Flu shot, and it isn’t just that more people are getting it. We should all wonder why nearly half of recent VAERS reports involve people who have gotten a Flu shot."
Or - perhaps all the media hype about the risks of the "swine flu" vaccine resulted in an increase in self-reporting. Rubin does not even consider this possibility. He treats the data as if it were a recording of actual side effects, not voluntary reports of side effects.
VAERS remains a very important source of information for the CDC and others to monitor the vaccine program. However, it continues to be abused in a pseudoscientific way by those in the anti-vaccine community and also by those who are simply naive about the nature of VAERS and the limitations of self-reporting.