Courtrooms are generally not a good place to decide issues of science. I’ve said this more times than I can remember. Admittedly, courts can at times do pretty well with issues of science. The Vaccine Court is a good example, as is the Autism Omnibus decision, which ruled that the test cases brought before the Vaccine Court to determine if there was a plausible case to show a potential causative relationship between vaccines and autism. The court ruled against the test case complainants, even though the rules of evidence are those of a civil court, in which “50% and a feather” are all that is required for a ruling in favor of the complainant.
On the other hand, there are numerous examples that I can point to where courts got the science badly wrong. The one I always like to cite occurred in the 1980s and 1990s, where lawsuits against Dow-Corning over chronic diseases, such as autoimmune diseases and cancer, claimed to be caused by silicone breast implants, which resulted in Dow Corning agreeing to pay $3.2 billion in settlement in 1998 for tens of thousands of claims, resulting in the company filing for bankruptcy reorganization a few months later. There was only one problem. Science does did not support a link between silicone implants and it never did. Indeed, I remember that the Vaccine Court was created by the National Childhood Vaccine Injury Act of 1986 because of a flood of lawsuits over “vaccine injury.”
Unfortunately, yesterday I learned of a decision by the Court of Justice of the European Union that illustrates just how badly courts can get it wrong and is likely to open the way for more scientifically unsupportable rulings, at least in Europe:
The Court of Justice of the European Union ruled Wednesday that courts may consider vaccines to be the cause of an illness, even in the absence of scientific evidence confirming a link.
I can’t help but note that this is a very different thing than CNN’s profoundly clickbait headline that “Vaccines can be blamed for illness without scientific evidence.” A lot of news outlets were guilty of this.
Clickbait-loving editors will be clickbait-loving editors, I guess. Onward::
The EU’s highest court said that if the development of a disease is timely to the person’s receiving a vaccine, if the person was previously health with a lack of history of the disease in their family and if a significant number of disease cases are reported among people receiving a certain vaccine, this may serve as enough proof.
Yes, you read that right. Basically, the court said that correlation can equal causation. Of course, in some cases, that is true. After all, some correlations do indicate causation. However, as all scientists know, an observation of a correlation is a beginning, not an end. Observed correlations are hypothesis-generating. They are not sufficient to conclusively support a hypothesis. That requires more, a lot more, because correlations alone can be profoundly misleading. There can be confounding factors, such that an observed correlation to one factor (say, factor X) is actually a correlation to another factor (say, factor Y) that happens to be also correlated with factor X. Epidemiologists know this and do their damnedest to account for such confounders.
So what was the case that brought about this most curious ruling? A man with multiple sclerosis claimed that the hepatitis B vaccine caused his disease and sued:
The ruling stemmed from the case of a French man known as J.W. who was vaccinated against hepatitis B in 1998 and developed multiple sclerosis a year later. Multiple sclerosis is a neurological disorder in which the body’s own immune system attacks the brain and spinal cord. The disease scars nerve tissue and causes a range of symptoms, from vision problems to paralysis. J.W. died in 2011.
In 2006, J.W. sued pharmaceutical company Sanofi Pasteur, which produced the vaccine, blaming it for his decline in health.
The case was brought before the Court of Appeal in France, which ruled that there was no scientific consensus supporting a causal link and no evidence of a causal link between the hepatitis B vaccine and the man’s multiple sclerosis, therefore dismissing the action.
This judgment was appealed and brought to the French Court of Cessation, which took it to the European Court of Justice.
The Court of Justice said that “specific and consistent evidence” relating to timeliness, a prior healthy status, lack of family history and multiple cases may prove to be enough, according to a statement. J.W.’s case referred to the first three criteria.
So let’s take a look at the case and see if it even meets the court’s criteria—even one of them. Here’s the relevant document:
Between the end of 1998 and the middle of 1999 Mr J. W was vaccinated against hepatitis B using a vaccine produced by Sanofi Pasteur. In August 1999, Mr W began to present with various troubles, which led to a diagnosis of multiple sclerosis in November 2000. Mr W died in 2011. Earlier, in 2006, he and his family had brought legal proceedings against Sanofi Pasteur to obtain compensation for the damage they claim Mr W suffered due to the vaccine.
So wait a minute. When, exactly, was Mr. JW vaccinated against hepatitis B? That’s a several month window! Then, his symptoms didn’t start until months later, although it’s unclear from this description how many months. Then, it was more than a year before he was diagnosed with multiple sclerosis. This is hardly “specific and consistent evidence” relating to timeliness. It’s the same sort of weak gruel that we usually hear from parents convinced that vaccines caused their children’s autism. It’s the same sort of thin gruel that we hear from parents who think that Gardasil caused their daughter’s premature ovarian failure or even death, with long time frames between vaccination and the onset of symptoms. Moreover, there is plenty of evidence that the hepatitis B vaccine does not cause multiple sclerosis or other nerve demyelinating diseases.
Basically, the European Court of Justice’s ruling is very disturbing. Although it is not a decision on a specific case (that is, JW’s case), it does represent guidance for European courts in future cases claiming “vaccine injury.” In fairness, I do have to repeat that the clickbait headlines about the ruling are also misleading. For instance, as bad as this ruling is, it does not say that “vaccines can be blamed for any illness even with lack of evidence” or that “vaccines can be blamed for diseases without any proof.”
If anything, the ruling is far more puzzling than anything else and definitely appears to lower the standard of evidence for vaccine injury causation cases. It is a rather absurd ruling as well. Indeed, the news reports emphasize the absurdity of the ruling, with Dr. Paul Offit noting that using the court’s criteria, “you could reasonably make the case that someone should be compensated for developing leukemia after eating a peanut butter sandwich.” Now, I love and admire Dr. Offit to death, but unfortunately have to say that I find his example a bit too obvious—a bit too precious—a reductio ad absurdum, because even I, as staunch a proponent of vaccination and opponent of antivaccine nonsense that I am, don’t think that a court would, using the European Court of Justice’s ruling, find for the plaintiff in a lawsuit against a peanut butter company claiming that its product resulted in his leukemia. Personally, I think that Peter Openshaw, president of the British Society for Immunology and professor of experimental medicine at Imperial College London, nailed the issue more accurately:
“The scientific evidence does not support a link between the hepatitis B vaccine, or any other vaccine in current use, and multiple sclerosis,” he said. “To say that there is a link between any vaccine and multiple sclerosis and at the same time to admit that there is no scientific evidence of such a link is illogical and confusing to the public.”
Indeed, if you read the actual decision itself, I suspect that you, as I did, will find yourself scratching your head, because the decision is very poorly reasoned. To demonstrate that, it’s necessary to quote liberally:
In today’s judgment, the Court holds that evidentiary rules allowing the court, where there is not certain and irrefutable evidence, to conclude that there is a defect in a vaccine and a causal link between the defect and a disease on the basis of a set of evidence the seriousness, specificity and consistency of which allows it to consider, with a sufficiently high degree of probability, that such a conclusion corresponds to the reality of the situation, are compatible with the Directive. Such evidentiary rules do not bring about a reversal of the burden of proof which it is for the victim to discharge, since that system places the burden on the victim to prove the various elements of his case which, taken together, will provide the court hearing the case with a basis for its conclusion as to the existence of a defect in the vaccine and a causal link between that defect and the damage suffered.
OK, so the court can look at claims of correlation, but that doesn’t change the burden of proof on the plaintiff to prove all the elements of his case. So what’s changed? In any case, the court demonstrates a very poor understanding of science. First, the seriousness of the condition has nothing to do with whether it was caused by vaccines. These are two separate issues. Second, “temporal proximity” is useful for conditions with a rapid and clear onset, such as anaphylaxis. It’s not nearly as useful for diseases that often have an insidious onset, like multiple sclerosis. In other words, it’s not unreasonable to consider temporal proximity, but what that means for various diseases is so variable and vague that we rely on epidemiology to try to suss out correlations. Third, anecdotes are by their very nature seemingly plausible—and misleading. Large numbers of such anecdotes don’t necessarily make them more convincing. Then there’s this:
Moreover, excluding any method of proof other than certain proof based on medical research, could make it excessively difficult in many situations or, where it is common ground that medical research neither confirms nor rules out the existence of a causal link, impossible to establish producer liability, thereby undermining the effectiveness of the Directive and its objectives, which are to protect consumer health and safety and ensure a fair apportionment between the injured person and the producer of the risks inherent in modern technological production.
I find such language curious. I don’t know EU law here specifically, but the court seems to be implying that for a plaintiff to prevail the standard is evidence beyond a reasonable doubt when in civil cases it’s generally a preponderance of evidence (the aforementioned “50% and a feather”). In any event, there is no such thing as “certain proof based on medical research.” There never has been and never will be. What there are are probabilities, given that there is inherent uncertainty in all medical research. The questions that need to be determined in any given lawsuit are not whether there is “certain proof” but rather based in probabilities. How much, if at all, does a given exposure (e.g., a vaccine) increase the risk of disease, if increase the risk of disease it even does? How likely is it that the exposure claimed significantly contributed to the development of the disease claimed in this plaintiff? How likely does it have to be that the given exposure caused or contributed to the disease before compensation is warranted? We know from copious evidence that there is no good evidence that the hepatitis B vaccine causes multiple sclerosis or, more properly expressed, detectably increases the risk of multiple sclerosis above that of a population not exposed to the vaccine. In the case of JW, add that evidence to the very weak anecdotal evidence for causation in his case, and the only scientifically supportable conclusion is that it is incredibly unlikely that the hepatitis B vaccine caused JW’s multiple sclerosis.
What the Court of Justice seems to be doing is intentionally lowering the bar of evidence for product liability cases, while trying to convince itself that’s not what it’s doing. In other words, it seems to be trying to have it both ways:
In the present case, the Court considers that the temporal proximity between the administering of a vaccine and the occurrence of a disease, the lack of personal and familial history of that disease, together with the existence of a significant number of reported cases of the disease occurring following such vaccines being administered, appears on the face of it to constitute evidence which, taken together, may lead a national court to consider that a victim has discharged his burden of proof. That could be the case inter alia where that evidence leads the court to consider, first, that the administering of the vaccine is the most plausible explanation for the occurrence of the disease and, second, that the vaccine therefore does not offer the safety that one is entitled to expect.
On the surface, this sounds reasonable, but in reality it paves the way for correlations that don’t equal causation (for instance, that are spurious, coincidental, or due to confounders) to result in scientifically unjustifiable decisions in favor of plaintiffs claiming vaccine injury when there is no compelling evidence that vaccines caused whatever condition it is the plaintiff is suing over. Worse, the court uses as its template a case where the temporal correlation is not strong at all and represents it as potentially being sufficiently strong without additional evidence if there are a “sufficient number” of other cases reporting such a correlation. But what is a “sufficient number”? Actually, we already know how to define that number. It’s basically what epidemiologists do when they do epidemiological studies. Or, as Openshaw puts it:
“The only alleged evidence that would be worth taking seriously is the alleged numbers of other similar cases,” he said. “Those data should be capable of detailed case comparisons for consistency, and probably also orthodox epidemiological study.”
Exactly. If there are enough such cases, then science should be able to study them, compare them to the population at large, and determine if there is actually a real correlation and, if there is, whether that correlation is likely to represent evidence of causation. That’s what epidemiologists, not judges, are trained to do. Courts should let them do it.
Oh, and news organizations should stop representing this decision as allowing courts to blame vaccines for any disease without evidence. That doesn’t help. This ruling is muddled as hell, but it doesn’t say anything like that. It does, however, very much appear to lower the bar of evidence for vaccine injury causation claims. That’s bad enough.