A randomised controlled trial submitted to a journal showed that a nutritional supplement could dramatically improve one aspect of the health of the elderly. The study was a follow up to a trial reported in an international journal eight years previously. Why had there been so much delay? Why were the results reported in this study not reported in the previous study? There was only one author and, if true, the results were extremely dramatic. The paper was sent for statistical review. The reviewer suggested that the paper bore all the hallmarks of being entirely invented. The results were unbelievably dramatic for the kind of health problem reported. The president of the university was asked to investigate.
_ Some countries are less rigorous than the UK regarding research conduct. _ Show concern rather than ask for definitive evidence. _ Refer the matter to the author’s institution. _ If there is no response contact the country’s national regulatory body.
The institution had investigated and had found no problem. But the editor remained unconvinced and sent the paper for further statistical review. The journal’s ethics committee also felt the results were unbelievable and suggested that the editor approach the institution again, asking for a further explanation, using the new evidence from the statistician and the ethics committee to highlight their concerns. It was not up to COPE to question the institution’s process of investigation. As the paper had been rejected the raw data could not be requested. But it was an editor’s responsibility to protect the integrity of scientific publication. The editor contacted the institution again requesting further information on this judgment. The journal is seriously considering publishing something on its unhappiness with the process. The university informed the editor that the lead author had resigned from his post. The author responded, explaining that it was “not practical” to respond to queries about the paper, because “all my papers are in storage and some pertaining to this study were mislaid. ” The university now wishes to “close the book on this matter” unless the editor could suggest another approach. Though the paper has never been formally rejected or withdrawn, a very similar paper has been published, raising the issue of duplicate submission of the paper.