Jul 29 2013

Legal Courts And Science

Facebook is like a graveyard in a zombie movie, where old news items rise from the dead to have a second life. I am often asked about news items that are burning up Facebook, only to find that they are years old, but never-the-less they have to be addressed all over again. ]

One such item (actually a few items) is a 2012 news report about the Italian courts awarding money to the Bocca family a large reward because it concluded their 9-year-old son acquired autism from the MMR vaccine.

History here is a useful guide. The courts have historically often been out-of-step with the science, tending to err on the side of awarding compensation for possible harm. For example, until about the 1920s it was thought that physical trauma could cause cancer. Animal studies and epidemiological evidence, however, showed that there was no causal connection. Recall bias and increased surveillance were likely the cause of the apparent association.

However, this did not stop the courts from awarding damages to individuals based upon the notion that minor trauma caused their cancer – even decades after the scientific community rejected this notion. A legal paper on this topic demonstrates the difference in approach:

The author feels that the legal concept of causation is best satisfied by the sequence of events test, which results in a shifting of the burden of proof to the defendant.

In other words – if the cancer happened after the trauma it is reasonable for legal purposes to assume that the trauma caused the cancer (the post hoc ergo propter hoc logical fallacy), and then to shift the burden of proof to the defendant so that they now have to prove that the trauma did not cause the cancer.

Another example is that of silicone breast implants and autoimmune disease. A series of class-action suits were brought against Dow Chemical and Dow Corning  for cases of lupus and other autoimmune disease  alleged caused by ruptured implants. Many multi-million dollar judgments were made against the companies, and Dow Corning filed for bankruptcy to protect itself against the slew of suits.

This was not based on any solid scientific evidence that breast implants actually caused autoimmune disease, but rather on the difficulty in ruling out that it did. Even completely negative studies do not rule out a connection – the more data we have, rather, the smaller a possible risk can be. We can never, however, rule out a tiny risk or a rare reaction. This, of course, opens the door for legal cases claiming to be that one rare case.

The most recent review of the epidemiological evidence I can find is from 2004, and they conclude that there is no evidence linking silicone breast implants and autoimmune disease.

These historical cases, and others, demonstrate that the legal standard for concluding a causal connection to harm is different than the scientific standard – it is lower, and errs on the side of concluding that harm did occur. This is deliberate, the philosophy being that the courts do not want possible victims to go uncompensated just because the science is unsettled. However, this keeps the door perpetually open for claims of rare cases, because scientific data will never be sufficient to rule out rare causal events.

For antivaccinationists who do not want to give up on their claim that vaccines cause autism, court rulings provide the occasional propaganda boon. They use the courts as a proxy because they have lost their case in the scientific arena.

I have discussed the evidence for vaccines and autism many times, and a fairly thorough reference of that evidence can be found here. The MMR vaccine is not causally linked to autism, and neither is thimerosal or vaccines in general.

Ironically, when the vaccine court in the US was asked to directly address this question they concluded that MMR and thimerosal, separately or in combination, do not cause autism. However, in individual cases in which neurological injury is alleged the courts do sometimes conclude that “compensation is appropriate.” The word choice is also deliberate – they are not making scientific conclusions, only saying that by the rules of the court the family should be compensated, acknowledging that those rules set the bar much lower than scientific consensus.

Also, in many of these cases the courts are not finding that the vaccines caused autism, but neurological injury that may have some autistic features. This is not a quibble – the idea is that vaccines can rarely cause encephalitis – swelling of the brain – which can cause brain damage. When encephalitis occurs after a vaccine, that is a table injury, meaning that the mere association is sufficient to justify compensation (in line with the legal standard recommended above). Having encephalitis that causes brain damage which produces some features that resemble autism is not the same thing as having autism. It is therefore not legitimate to conclude from these cases that vaccines cause autism.

This doesn’t stop the antivaccinationists from doing just that.


The scientific data is largely in – to a high degree of confidence we can conclude that vaccines do not cause autism. The data can never demonstrate that the risk is zero, but if it does exist it is very small. It is tiny enough that the benefits of getting vaccinated vastly outweigh the possible risks.

The courts have the burden of deciding in individual cases whether a potential victim should be compensated. They have to deal with often conflicting or inconclusive science, or the residue of rare but possible risk that can never be eliminated. The courts tend to err on the side of victims. Such decisions, however, do not trump the scientific evidence. They are grist, however, for the antivaccine propaganda mill.



10 responses so far

10 Responses to “Legal Courts And Science”

  1. ChrisHon 29 Jul 2013 at 10:50 am

    However, in individual cases in which neurological injury is alleged the courts do sometimes conclude that “compensation is appropriate.”

    I often come up with those that say since the courts award from vaccine injury to the tune of some big number that it is proof positive vaccines are dangerous. Though I try to explain the numbers a bit differently:

    In over twenty years the National Vaccine Injury Compensation Program has awarded less then 3400 cases. In that same time span several billion doses of vaccines have been administered in the USA (kind of an estimate based on four million births, and between 20 and 55 vaccines, the latter number from an anti-vaccine site). That is about one injury in around a million doses, which makes them safer than the diseases.

    I have not yet gotten a real response. Some may claim that the vaccines cause more injury, but when I press for scientific evidence (not legal) I get nothing. Well, not exactly, several like to point to conspiracy theories between the CDC, FDA, any government and Big Pharma.

  2. tedwon 29 Jul 2013 at 9:01 pm

    Just a note about courts getting the science wrong (as background, I’m an electrical engineer, attorney, and now a high school physics/math teacher). There are really two culprits in the U.S. (I don’t know about the Italian legal system), the judge and the fact finder (the jury or, in a non-jury trial, the judge). If a party challenges the probative value (validity) of proposed expert evidence, the judge makes the initial determination about whether such expert scientific evidence sees the light of day in the court room. There are two tests floating around in the U.S. legal system (each jurisdiction (federal or state courts) adopts which one applies via rules of evidence or case law). Under either test, the judge must consider the extent to which such evidence is generally accepted by the scientific community. Under one test (Frye Test–a number of states), the court should only allow such evidence that is generally accepted by the scientific community. Under the other test (Daubert Test–federal courts and some states), the judge considers to what extent it is accepted by the scientific community, as well as whether it is peer reviewed, whether the theory is falsifiable, and the error factors if known. In theory, either test should keep pseudo-science evidence out of a trial. However, it is always one judge making the decision for each case before him or her–a mere human with biases (on appeal, the appellate court will scrutinize the judges ruling based on an “abuse of discretion” standard). Consequently, a judge could allow nonsense into evidence. Once evidence is allowed into the trial, it is, then, up to the jury to decide the final outcome–the jury can do as it pleases. However, one of the old sayings in the legal system is, “bad facts make bad law.” In other words, a sympathetic plaintiff might cause the court and/or the jury to bend over backwards to find in favor of such a plaintiff.

    The other issue is the precedence of any ruling. In theory, each case is based on the facts presented. So, if one case goes for the plaintiff on an MMR-Autism claim, this is not precedent for another case, because the facts could be different. This is true whether or not the second case is brought before the same court or one in another town, county, or state. In the case of Dow and the breast-implants, Dow simply made a decision that it would not have any better “luck” with subsequent lawsuits. It seems, all too often, companies would rather pay out than fight the battle, as a matter of economics.

  3. Kawarthajonon 29 Jul 2013 at 10:36 pm

    Another important issue to consider is the standard of proof in civil Court – the balance of probabilities. This is a fairly low level of evidence to prove a case (i.e. 51%), so you’re not really needing to work too hard to prove that the vaccine did harm.

    If these cases were ever tried in criminal Court (i.e. to prove criminal negligence on the part of the vaccine company execs or something like that), the standard of proof would be more like 99% and the anti-vaccine nuts would lose every time.

    Science and the Courts overall have a mixed relationship. The Courts have consistently upheld the ban on teaching creationism as a science in school, but often rule against science in other situations. It is interesting to me that a system based entirely on logic (law) can fail to uphold logic (science) in many cases.

  4. Bruce Woodwardon 30 Jul 2013 at 3:45 am

    Following the whole creationsim thing in the US and A, I always get the impression that the decisions on upholding bans and relevant rulings are dependant on the judge. There is always uncertainty around the initial verdicts too where there is no uncertainty in the science.

  5. evhantheinfidelon 30 Jul 2013 at 4:59 am

    Hey, Dr. Novella, I was wondering if there was research into the effects of observing brain scans on the brains of those who are interpreting them?

  6. Bill Openthalton 31 Jul 2013 at 7:16 am


    … decisions on upholding bans and relevant rulings are dependant on the judge.

    I have a couple of legal eagles in the family, and they tell me this is a feature. The legal texts are guidelines and determine boundaries, but the judges should appreciate the specifics of each case and rule accordingly. I always got frustrated by the inability of lawyers to look at the facts and say how the ruling would be, until I understood it does depend on which judge is hearing the case. It took me longer to understand that this is actually a good thing.

  7. Bruce Woodwardon 31 Jul 2013 at 7:30 am


    Noticed the error on reading back my own quoted comment.

  8. Bill Openthalton 31 Jul 2013 at 8:56 am


    In French it would have been correct :)

    If ever you doubt your spelling & pronunciation skills, there’s always “The Chaos” by Gerard Nolst Trenité.


  9. Bruce Woodwardon 31 Jul 2013 at 10:02 am

    Thanks Bill.

    I used to be so much better, but MS Word Autocorrect has made me lazy in my old age.

  10. Bill Openthalton 31 Jul 2013 at 10:59 am

    When I was young, I spent a fair chunk of my time learning how to use mathematical tables (slide rules were too expensive for schoolboys). As a result, I got quite adept at mental arithmetic, a skill I all but lost making a career out of programming and using computing devices (from the Texas Instruments SR-56 to Beowulf clusters).

    Obviously, substantial mental resources were devoted to knowing how to calculate and how to spell, and thanks to computers, today’s humans, relieved from these tedious chores as they are, can use their brains for nobler activities…

    … like watching re-runs of NCIS.


Trackback URI | Comments RSS

Leave a Reply

You must be logged in to post a comment.