Too good to be true

Are carelessness, mistakes and cheating in research and publishing becoming more and more common, or have we simply got better at detecting them?

In April 2000, Peter Kranke et al. published an article in the American journal Anesthesia & Analgesia: «Reported data on granisetron and postoperative nausea and vomiting by Fujii et al. are incredibly nice!» (1). The title was not intended as a compliment. The following year, they found a number of discrepancies between data published by Fujii and data published by others (2). Nothing happened. Unease and wonder over research results are one thing; doing something about the suspicions is another thing altogether. So Yoshitaka Fukii continued to publish.

In April 2010, the British journal Anaesthesia featured an editorial by R. Andrew Moore et al., discussing the negative effect that carelessness, cheating and mistakes in research can have on medical practice (3). The article illustrated how difficult it is to prove that data have been falsified or fabricated, but showed at the same time how discrepancies could be revealed when studies were compared in the right way in correctly performed systematic review articles. Moore drew on Fujii’s and other data to demonstrate his point. The present editor-in-chief, Steven Yentis, described subsequent events in an editorial in Anaesthesia that was first published on the internet on 8 March 2012 (4): «The Journal received a number of responses to Moore et al.’s editorial, amongst them one from a reader who bemoaned the fact that the evidence base remained distorted by that researcher’s work, and challenging anaesthetic journal editors to do something about it. This all happened just a few months after an unprecedented international collaboration between anaesthetic journal editors and publishers, that had led to retraction of almost 90 published papers … by Joachim Boldt» (5).

Yentis went on to describe how he – in contrast to his American editorial colleagues ten years earlier – chose to pursue the matter. Instead of publishing the letter to the editor immediately, he challenged its author to make as thorough an analysis of Fujii’s work as possible – so thorough that it might be used to confront the researcher and the institution at which he worked. This is what the correspondent, John Carlisle, has now done (6). It hasn’t been simple. As Steven Yentis puts it: «(It has taken) 19 months, 18 reiterations, two consultations with the Committee on Publication Ethics (COPE) and three statisticians…» But the results are impressive – and frightening. The analysis of 168 of Fujii’s papers shows clearly how far his results deviated from what might reasonably have been expected if the research had proceeded correctly. An editorial explains in detail the method that was used (7).

The most recent development in this matter is that all the 23 editors-in-chief who have published Fujii’s papers sent a letter dated 9 April 2012 to the officials responsible at the nine Japanese institutions from which his work has come and asked for answers to three concrete questions (8):

This is in complete accordance with COPE’s guidelines for collaboration between journals and research institutions on cases concerning ethics and integrity in research and publication (9).

COPE has been pursuing the case closely over the years, and we have attempted to give the editors-in-chief advice and support along the way. COPE is an organisation for editors and publishers of scientific journals that has now existed for 15 years (10). The organisation has evolved from a small, but important club in which editors could discuss difficult cases, into a large organisation with over 7 000 scientific journals and publishers. All the members commit themselves to following COPE’s ethical guidelines and Codes of Conduct (11). In other respects the objective of the organisation is the same as it has always been: to contribute practical advice and develop guidelines that enable editors, individually and jointly, to ensure that the body of research literature is correct and reliable.

The manner in which the anaesthesiology community has handled this particular case is exemplary. The reality is unfortunate, however. As matters stand now, it looks as though 193 articles from the period 1991 – 2011 have been or may be withdrawn. This means that the knowledge base will have to be reviewed again. We can only hope that the «invalid» articles have not had serious consequences for patients who have undergone anaesthesia during these years. Consequences there have undoubtedly been. Because medical practice is knowledge-based, it is of necessity subject to constant adjustment as new knowledge becomes available. This means that the knowledge base must be correct.