A prominent medical journal is retracting a controversial Canadian scientist's 1989 study on the effect of infant formula on the incidence of allergies, calling the case a "major failure of scientific governance."

The BMJ, formerly the British Medical Journal, announced today it's retracting the study it published in 1989 by Dr. Ranjit Kumar Chandra.

Chandra's influential study looked at the effects of breastfeeding and formula for infants who were at high risk of eczema because of their family history. It concluded women who breastfeed should avoid common allergenic foods and those who use formula should choose a hydrolyzed type of formula.

Memorial University of Newfoundland completed an inquiry into Chandra's research in 1995, which concluded that "scientific misconduct has been committed by Dr. Chandra." But the university never published the inquiry report or alerted journal editors.

CBC-TV's The National aired a three-part documentary called The Secret Life of Dr. Chandra in 2006, after which Chandra sued the CBC for libel. This summer, a jury ruled that the documentaries were true. The BMJ obtained the university's internal investigation report after it was entered as evidence at the trial. 

"He had all the data analyzed and published even before we had the data collected," nurse Marilyn Harvey, Chandra's research assistant, told CBC at the time.

Ivan Oransky

'If someone makes up data out of whole cloth, it's very difficult to tell whether you are reading a novel or a scientific paper,' says Dr. Ivan Oransky, co-founder of the Retraction Watch blog. (CBC)

The BMJ said it's retracting the paper because of the 1995 report and "because of the convincing evidence given in the CBC television programs and the court case that the work of R.K. Chandra is not to be trusted."

University ignored earlier fraud

This is not the first study of Chandra's to be retracted. It was the BMJ that uncovered the first incidence of fraud for Chandra's own patented vitamin pill, which the university ignored. 

When Chandra submitted a paper to the BMJ 15 years ago about his vitamin pill, it set off alarms. Richard Smith was the journal's editor then.

Smith said the journal's statistical reviewer concluded Chandra's vitamin's data had "the hallmarks of being entirely invented."

"From my point of view, the university's the real villain of the piece, unfortunately. I mean there will always be fraudsters. Wherever there's human activity, there's misconduct, but the university should have taken this much more seriously," said Smith, who also wrote an editorial published with the eczema retraction.

"It seems to me that universities should be about integrity and truth and as far as I can see this university has completely ignored both of those."

The editorial labelled the case a "major failure of scientific governance" and called for a public inquiry, saying the saga highlights a collective failure to defend the integrity of science.

Fictional data

The retracted paper has been cited more than 100 times by other papers, said Dr. Ivan Oransky, co-founder of Retraction Watch blog and a professor of medicine at New York University. 

"Had it been retracted [before], that would have been a big red flag, literally a red water mark on the paper," said Oransky.

"If you're relying on something that turns out to be fraudulent, turns out be fake, well that's a problem for science, it is a problem for those of us who follow science and for the public, in this case in terms of public health."

The peer review process of publishing scientific and medical research is only one limited filter, said Oransky. He cautioned that publications shouldn't be considered sacrosanct.

"It is very difficult if someone makes up data out of whole cloth, it's very difficult to tell whether you are reading a novel or a scientific paper."

​While retractions have increased to 400 from 40 between 2001 and 2010, Oransky said it's unclear whether scientific fraud or misconduct is also on the rise, since it also now easier to catch fraud.

Prof. Richard Marceau, Memorial's vice-president of research, said the university has established a number of policies and procedures to prevent, investigate and, if necessary, punish research misconduct.

"We at Memorial University have very high standards. We maintain them, we enforce them. We have progressed over the years. We have learned a great deal," Marceau said. 

Asked about the journal's categorization of the university as a villain, he said, "Once a paper is published and found to be wanting, everybody is a victim in this story, the journal, the university, readers."

Chandra now focuses on his vitamin business in India. He did not respond to emailed requests for comment from CBC News.