Prevention of fraud

Although it might be hoped that investigators would be less inclined to commit research and publication misconduct if they knew the chances of getting caught and punished were increased, ultimately the route to prevention must be through education and reaffirmation of the still widely held belief in the principle of research honesty. All students entering a period of research should not only receive instruction on research design, methodology, and laboratory techniques but also guidance on the fundamentals of research and publication ethics (Box 18.3). Research should be protocol driven, and contributors and collaborators should define their roles before the work is commenced. Any protocol changes should be discussed and agreed by all of the participants. Statistical advice should be sought, where appropriate, during the planning process, and it goes without saying that all studies involving human subjects should be submitted to an appropriate research ethics committee. There is a sense that at least some of the instances of research misconduct could have been avoided by closer supervision of the project. Supervisors can find themselves increasingly distanced from the laboratory bench or the patients in the clinic, which can make it difficult to check the veracity of the primary data. Review of research results should involve examination of the primary research record, which may be the laboratory record book or patient's medical notes. Impeccable record keeping is an essential component of good research.

Box 18.3 Can research and publication misconduct be prevented?

• Research training

• Research ethics

• Publication ethics

• Protocol driven

• Establish contributors and collaborators

• agree protocol

• agree presentation of results

• Define methodology for data analysis

• statistical advice

• Ethical approval

• Project and personal licence (Home Office)

• Supervision

• communication

• ensure good clinical practice

• record keeping

• The publication

• Disclose conflict of interest

• Disclose previous publications

• Approval by all contributors

• Submit to one journal at a time

• Assume research data audit

The ethics of publication are as important as research ethics. All authors should see the final manuscript before submission and each sign the declaration confirming the originality of the work. An article should only be submitted to one journal at a time. Authors should disclose any potential conflicts of interest, particularly if the opinions expressed in the manuscript could be influenced by financial incentives through sponsors. Authors should also disclose the nature of any related publications that have been published recently or that are under consideration by another journal. There are times when it is entirely reasonable to publish some previously published data, but this should be done with the full knowledge of the editor and reviewers, so that a full assessment can be made as to the extent of any potential overlap. In the future, original research may be submitted to random audit prior to publication and it is therefore wise that authors assume that they may be asked to produce their primary data to support the work described in a manuscript.

Was this article helpful?

0 0

Post a comment