In an unusual authorship dispute, an Indian researcher has accused her US collaborator of 'scientific misconduct' for submitting a manuscript for publication without her approval.
Mrudula Phadke was the principal investigator (PI) of an Indo-US clinical trial on how to prevent HIV transmission from mother to child during breast feeding. The project was funded by the US National Institutes of Health in Bethesda and conducted at the B. J. Medical College in Pune, of which Phadke was the dean.
Phadke who is currently vice-chancellor of Maharashtra University of Health Sciences in Nashik says as PI of the project she did not give permission for publishing the report in the form it was written. She thought the conclusion drawn was inconsistent with the findings and was based on 'enormous misrepresentation and misinterpretation of the Indian data.' She and two of her co-workers have distanced themselves from the study's conclusion published recently1.
"The controversy is unfortunate," Robert Bollinger, Phadke's US counterpart and associate director of the Johns Hopkins Center for Global Health in Baltimore told NatureIndia . "I am afraid this has left a bad impression." Bollinger, whose association with Indian pubic health researchers stretches for over three decades, said he is "worried about the impact of this episode on the enthusiasm of sponsors and international collaborators."
The dispute arose over the interpretation of the results of the clinical trial. During the trial the Indo-US team sought to find out if daily nevirapine given to newborns for six weeks could be more effective than the current practice of single dose at birth in reducing HIV transmission. In all, 370 babies born to HIV infected mothers received single-dose and another 367 were given nevirapine for six weeks. The Indian data was pooled with that from similar trials done in Ethiopia and Uganda to provide appropriate sample size for statistical analysis.
According to the study design, babies in the two groups were to be tested for HIV at six months. Analysis showed that the six-week regimen (in comparison to single dose) did not significantly reduce infection at 6 months of age but the difference was significant at 6 weeks. Interpreting these results, the study authors concluded that the lack of a significant reduction at 6 months suggested that a longer (than 6 weeks) course of nevirapine 'might be more effective.'
Phadke argued that the conclusion was incorrect as it was drawn from the results at 6 weeks and not 6 months — the 'primary end point' of the study. Phadke in an email, informed Bollinger that his paper "is indeed an example of misrepresentation of data and hence it is a scientific misconduct." She also wrote to India's health secretary that cost to infant health and to society would be enormous if the study's recommendation for extended nevirapine therapy was adopted.
"When you are asked to give nevirapine for 6 weeks to an apparently healthy new born and when the drug also produces adverse reactions in 38 - 39% babies, it is time to think," Phadke said. "A drug meant to be a prophylaxis should be used for short period and be highly effective with least side effects. Nevirapine has not fulfilled these criteria. I am 200 per cent sure the US scientists are just trying to bulldoze us. They think they can dump any (drug) on us."
Bollinger argues that the results of "the analyses are scientifically sound", especially when seen together with trial results from Malawi where nevirapine was given for 14 weeks with significant reduction in HIV infection and mortality in infants at nine months of age.
Bollinger also dismisses as nonsense Phadke's charge that she was given only half the data needed for an independent analysis. "Phadke asked for the data of study participants which we could not share for confidentiality reasons," he said. "Excepting this, a complete copy of the computerized research data base was given to her many months ago and I do not think she could find a statistician to analyze it."
Bollinger says dispute with Phadke and efforts to get her concurrence delayed the publication for nine months during which more babies were dying (due to HIV infection). "It is a sad story for Indian science." Bollinger said that when all efforts to resolve the disagreement reached a stalemate, they resolved to submit the manuscript to the journal by majority vote of the manuscript's writing team as suggested by the NIH. Phadke argues that publication cannot be by consensus or voting. "If the PI does not agree, no body should publish."
"Disputes among authors are one of the most difficult situations for editors to deal with," editors of the journal 'The Lancet' wrote in a commentary accompanying the study. The editors felt publication of these important data, with complete disclosure, was the best course of action.
The dispute has been particularly difficult and demoralizing for many, Bollinger said. "It would be a terrible shame if such episodes discourage future collaborations and limit the opportunities for junior Indian investigators," he added. "Despite these concerns, I am still optimistic."
