The perils of dishonest science

Margaret Somerville
12 February 2010
Reproduced with Permission

Last week, I received several calls from the media asking about the ethics issues raised by two stories that were in the news.

One was that a prestigious medical journal, The Lancet, had retracted a 1998 study linking childhood vaccinations with autism. The other was that there had been an outbreak of pertussis -- "whooping cough" -- in the Kootenay Boundary area of British Columbia which has the lowest rate of immunization of children in the area covered by BC's Interior Health Authority.

There is a connection between these stories because some Canadian parents have been reluctant to have their children immunized because of fears about autism.

The editor of The Lancet described the publication of the seriously flawed study that linked the measles, mumps and rubella vaccination with the onset of symptoms of autism as a "collective failure."

As well as the journal's editors and reviewers, he included in "the collective" who failed, the government, institutions and scientists. And "the failures" extended to dishonest and misleading statements in the retracted article, major conflicts of interest (the researcher was being paid by lawyers acting for the parents of the autistic children, who were suing the vaccine manufacturers for damages), and scientifically invalid research methodologies (for instance, an inadequate number of research subjects to give statistically valid results), all of which also constitute serious breaches of ethics.

In addition, subsequent research had not been able to replicate the findings reported in the article. Some journals now require that certain research be repeated by independent investigators to verify results, before they are willing to publish the findings.

As a result of this article, there was a decline in the number of children being vaccinated, for example, in Britain, and an increase in the childhood diseases they would have protected against. In short, both physicians and members of the general public rely on such articles to make treatment decisions.

On-going damage

Despite the retraction, the damage caused by the article may not have been stemmed. It's been reported that Canadians in the anti-vaccination movement are not convinced by the evidence that there is no proof that vaccines are associated with autism, as the article had wrongly claimed there was. Rather, they want proof that vaccines are not associated with autism and they say there is no such evidence.

Turning to the second story, whatever their reasons, parents in the Kootenay Boundary area are not responding to health authorities' pleas to have their children immunized and, as a result, there has been an outbreak of whooping cough. Nineteen cases have been diagnosed in the last eight weeks.

Here are some of the ethics issues I was asked to address in relation to these two stories. Keep in mind that when undertaking an ethical or legal analysis, there is no such entity as a neutral question. The questions we choose both to ask and not to ask, structure the outcomes of our enquiries.

The ethics of publication

Why did The Lancet wait over a decade to retract this article? Was this delay unethical or even illegal (negligence)?

That depends on all the facts. Even if publishing the article in the first place was not negligent, certainly, having published it, The Lancet had assumed a duty, both ethically and legally, to take reasonable care to warn people of the risks of relying on it, once they became aware of its inaccuracies. Refusing vaccination is not a risk-free decision and, in rare cases, children can even die from the infectious diseases the vaccinations would protect against.

Is unethical science and the publication of misleading results rare?

Not as rare as we thought 10 years ago. In recent times, there has been an increasing number of retractions from some of the world's most prestigious scientific and medical journals, to the extent that the editors of these journals have agreed on an ethics code to guide them to try to avoid similar mistakes in the future. One only has to search "publication ethics" on Google to see the explosion of reports and articles on this topic and the concern that serious breaches of ethics have caused.

Science and medicine are now major news stories on a daily basis. Everyone personally relates to and identifies with health stories -- "this could be me or someone I love." Over 50 per cent of hits on the internet are related to health. The temptations for researchers, their institutions, prestigious journals and even governments to be the "first cab off the rank" with news of breakthrough medical science are great, to say nothing of the prestige and money involved. We need safeguards to counter these realities.

But we also need to keep in mind that there can be wrongful blocking of the publication of research, not just wrongful publication. A famous example involves the prestigious surgeon, William Halsted, who believed that radical mastectomy (a very mutilating operation which he had pioneered in the late 19th and early 20th century) was the only acceptable treatment for breast cancer. For decades, researchers who challenged that belief were shunned and had great difficulty publishing research that showed segmental mastectomy (lumpectomy that is much less damaging) was also effective. (By the way, the "non-recurrence of the cancer" results that Dr. Halsted claimed for his operation were never able to be duplicated by other surgeons.)

Eminent American psychiatrist-ethicist, the late Jay Katz, provides wise advice in this regard: "Halsted should serve as a reminder that orthodoxy and authority are powerful forces that tend to obliterate awareness of uncertainty; they do not easily bow to contrary claims of science: that in the search for truth professionals must constantly scrutinize their certainties against their uncertainties of existing medical knowledge." We all, physicians and patients, must face up to our unavoidable ignorance and the unavoidable uncertainties that entails, and despite those uncertainties try to make wise decisions. False certainty is, indeed, especially dangerous in medicine.

Modern trust

What influences our decisions about matters such as vaccination?

In making risk-taking decisions -- and both deciding for or against vaccination are risk-taking decisions -- we should keep in mind that psychologists tell us that we tend to choose to do the opposite of our last bad decision. So engineers who decide that a dam is safe and doesn't need repair, but the dam breaks and hundreds of people in the town below are drowned, looking at the next dam will be likely to decide that it is unsafe. Parents' stories that vaccination caused their child to become autistic could operate as surrogate bad decisions for other parents.

The nature of trust has also changed. Earned trust has replaced blind trust.

Blind trust ("Trust me because I know what is best for you") is a paternalistic concept that depends on authority, status and power. In the past, medicine was governed by blind trust. If, even in the 1960s, a physician said to a mother, "Now we are going to vaccinate little Johnny," most women would not have questioned the physician's decision.

Now medicine is governed by earned trust ("Trust me because I will act in your 'best interests' and show that you can trust me"). That means that if trust is not earned it is not present. And earning it requires honesty about risks, harms, benefits and uncertainty regarding these, which, in turn, requires sharing of information and shared decision making. So now there is an opportunity for parents to say, "No thanks," to vaccination, however unwise that decision might be.

Should parents be forced to have their children vaccinated, both for the child's protection and for the benefit of other children?

The basic presumption is that parents have a right to decide what medical interventions may and may not be undertaken on their child. However, if their refusal of treatment amounts to child abuse or neglect, a court can step in and take decision making authority away from the parents. Refusal of vaccination would only constitute such a situation where there was a high incidence of a serious, highly contagious disease -- for instance, something in the nature of the Ebola virus, which is fatal, and there was a protective vaccine available.

Under public health protection acts even adults can be forcibly treated when an infectious disease they have threatens the public's health. But this is a strictly limited exception to respect for people's rights to autonomy and self-determination, and to their rights to "security of the person" and liberty. Although parents' rights to refuse medical treatment for their children are not as extensive as with regard to their own treatment, they are still the basic presumption.

But isn't it unfair that unvaccinated children benefit from the reduction of risk that results from vaccinated children?

This reduction in risk is called "herd immunity" -- vaccinating a child protects not only that child, but also helps to protect other children. And yes, it is unfair, but that doesn't mean we are ethically justified in forcing vaccination on children whose parents refuse it.

There are, however, ways that can be used to encourage vaccination. For instance, a school's requiring that, in order to be admitted as pupils, children must have proof that they have been vaccinated against certain diseases.

As often happens in ethics, in deciding about the "rules" that should govern childhood vaccinations, we have to walk a fine line in balancing respect for individual rights and protection of the community.