Imprimir Republish


Data manipulation

Fraud in a study about vaccine reopens discussion about research practices

LAURA DAVIÑAA series of articles published in January of this year in the British Medical Journal (BMJ), a highly regarded English medical publication, showed confusing evidence of the manipulation of data and non-ethical conduct in a late 1990’s study, which had a devastating impact on public health in several countries and put the lives of thousands of children at risk by suggesting that vaccination against measles, mumps and rubella (MMR) might cause the child to develop autism. With these BMJ texts, an issue that is as old as the scientific method itself is back in the limelight: how can one reduce the risk of fraud?

The story that is now being told in detail by the journalist Brian Deer, who investigated the case of the MMR vaccine for seven years and gained access to the medical records of those who took part in the study, exposes yet again the frailties of a knowledge production system that, despite some self-correcting capabilities, is not infallible. “Brian Deer’s series of articles illustrates many of the ways in which science may be corrupt,” wrote Douglas Opel and Douglas Diekema, from the Seattle Children’s Research Institute, and Edgar Marcuse, from the Seattle Children’s Hospital, in an editorial dated January 18, in the BMJ. “Above all, Deer shows that the mechanisms that are meant to ensure the integrity of research failed completely.”

The work that would turn out to be the result of data manipulation, omissions of responsibility and ethical deviations started to disseminate fear of the MMR vaccine (against measles, mumps and rubella) 13 years ago. In its February 18, 1998 edition, The Lancet, one of the world’s most influential medical journals, aired the apparently alarming results of a study conducted by the gastric surgeon Andrew Wakefield. In this article, he and another 12 authors narrated that one week after being given the vaccine, 12 children in England started showing gastrointestinal disorders along with impairment of their mental development akin to that of autism.

At the time, Wakefield stated that the symptoms of the children characterized a new syndrome, which he named regressive autism, because it arose after a phase of normal development. Although the article in The Lancet stated that “we have not proven the association between the vaccine against measles, mumps and rubella and the syndrome described,” Wakefield undertook to confirm the connection. With the support of the institution that he worked for, the Royal Free Hospital in London, he prepared a press conference and distributed a video to TV networks in which he championed the link between the vaccine and autism.

“The main reason for the press conference was not the possible connection between the intestinal and developmental disorders – it was Wakefield’s supposition that the MMR vaccine, used in the United States since the early 1970’s and in Great Britain since a decade before that, might be the element responsible for the dramatic increase in autism rates,” explains the journalist Seth Mnookin in the book The panic virus: a true story about medicine, science, and fear. Mnookin states, in his work released this year, that Wakefield stuck to his statement of having found the measles virus – something denied by other studies – in the intestinal tract of children with irritable bowel syndrome, to show a possible biological pathway connecting the vaccination to the intestinal disorder and to autism.

Although experts questioned the data at the time, the damage was done. Fear that the vaccine might cause autism spread through several countries with the help of anti-vaccination support groups and the careless work of the press. Result: the ratio of vaccinated children dropped to 80% in Great Britain in 2003, well below the 95% recommended by the World Health Organization, so that in 2008 measles once again became an endemic disease in England and in Wales.

LAURA DAVIÑAInvestigating the case, Brain Deer managed to get, in 2004, the first indications of fraud in Wakefield’s work, which he published in the newspaper The Sunday Times. As a result, the British General Medical Council opened proceedings against Wakefield and the other authors, which made it possible to reconstruct the farce.

Deer obtained proof that Wakefield had acted deliberately the whole time. The cases described in the article concerned children whose parents believed that their offspring had developed autism after vaccination, but had not received a medical diagnosis. They had been referred to Wakefield by an anti-vaccination association, Justice Awareness and Basic Support, whereas the correct procedure would have been to look for cases using a random population sample as the starting point or to receive patients referred by other medical centers. The surgeon also received payment from the lawyer Richard Barr, who was looking for evidence to sue the vaccine manufacturers.

According to these articles, Wakefield was not against vaccinating children in general, only against the MMR vaccine, since he had taken out a patent for a measles vaccine. “With no evidence, Wakefield stated for years that throughout the world, physicians, along with public health authorities, not only knew that the vaccines caused children terrible problems, but concealed this information for their own benefit,” Deer told Pesquisa FAPESP.

Retracted article
Although the medical council ruled on the ethical suitability of Wakefield’s research, it was Deer that showed that the clinical symptoms presented in the article in The Lancet did not correspond to those described by the children’s parents. Despite the evidence of fraud, it was only after the ruling of the council, which revoked Wakefield’s medical license in 2010, that The Lancet cancelled the 1998 article – it can still be read online, but it exhibits in red the word “retracted.”

Although the case was concluded, certain concerns remain in the air, given that similar problems may be more common than one might imagine. One such concern is: What drives people to manipulate research results? The physician William Saad Hossne, head of the National Research Ethics Committee (Conep) in Brazil from 1996 to 2007 and coordinator of the development of the rules of clinical research in the country, believes that the reasons are manifold. “The number of researchers is increasing exponentially; competition is increasingly strong and there is a search for recognition. Moreover, the projects are more complex and involve more people,” he states.

“Nowadays, the researcher must be productive, which fosters a more flexible research culture,” says Sueli Dallari, who teaches at the School of Public Health at the University of São Paulo (USP). “In studies with humans, when they record the data, researchers are often less strict than they should be,” states the researcher, who used to be a member of Conep and sat on the research ethics commission of USP’s Clinicas Hospital.

Concern about fraud, according to Hossne, led countries such as the United States, Germany and Denmark to establish, in recent decades, institutions that aim to guarantee the integrity of research. In Brazil, Conep, established in 1990, regulates, approves and monitors trials involving humans. Conep’s resolutions 196 and 251, for example, determine that research data must be stored and remain available for consultation for at least five years. “The fact that someone might ask to examine the data helps to control quality,” says Hossne.

According to Sueli, to improve this control further, it would be necessary to have ethics commissions able to monitor the execution of the projects. “As the number is large,” she says, “one might check on a few to see whether they actually did what they set out to do and in the way in which they set out to do it.”