Last week saw the publication of the third version of Good Publication Practice for reporting company-sponsored research (imaginatively, we're calling it GPP3). To be useful, guidelines must reflect the current situation so they need periodic updating. The first GPP guideline (called GPP for pharmaceutical companies) appeared in 2003 and was revised to form GPP2 in 2010. It is probably obvious to anybody working in the healthcare or publishing sectors why updates were needed, as these industries have been affected by a slew of regulations, technological innovations and new business models. Society's expectations of corporate behavior, demands for greater transparency, and views about what constitute best practice are also changing, so the guideline needs to reflect these.
So, what's new about GPP3?
The biggest change is that we have introduced a set of principles. One problem with revising guidelines is that they tend to get longer rather than shorter with each update, and there's a danger of losing sight of the original core concepts in the verbiage or ending up with a document that is too long or detailed for some users. It is also impossible to "future-proof" any guideline and predict all the new situations that may arise before it's time for another update. We therefore felt it would be helpful to add some guiding principles which we imagine won't change radically over time, which should serve as a useful introduction to GPP, and might help respond to future challenges.
Examples of why GPP2 wasn’t “future-proof” and needed some updates are recent changes in the publications environment that have emerged or amplified in the last five years such as journal requirements for data sharing, and the US Physician Payments Sunshine Act. Although this Act doesn’t directly mention publications, public posting of information about payments to doctors (and about “transfers of value” which might include editing and writing assistance) may affect public perceptions of investigators’ and physicians’ attitudes towards publications developed with companies. Other differences between GPP2 and GPP3 are: new sections on plagiarism and data sharing, more detail about which studies should be published, and a new table on authorship.
How did we develop GPP3?
The first GPP guidelines arose from a meeting held in 1999 to discuss publication issues and concerns of drug companies, journal editors, and investigators. The meeting was organized by the Council of Biology Editors (since renamed the Council of Science Editors). Afterwards, a small group of people from a handful of drug companies got together informally and drafted some guidelines. We shared the draft guidelines with those who had been at the meeting and some major pharmaceutical companies where we had contacts. The guidelines were rejected by two major journals before being published in Current Medical Research & Opinion in 2003. One of the unfavorable reviewer comments was that the guideline was of limited interest and should be published in “an industry journal” (whatever that was). However, we didn’t give up hope and by 2010 the guidelines were firmly establishedand adopted by many companies as well as the journals that had initially rejected them. The International Society for Medical Publication Professionals (ISMPP) had been formed in 2005 and offered to coordinate an update, ably led by Chris Graf of Wiley. GPP2 was produced after much more extensive (and more professional) consultation, including 116 sets of comments from a wide range of reviewers. It was published in 2010 in The BMJ (one of the journals that had rejected the original GPP). GPP3 has again benefited from logistical support from ISMPP, including making the draft available for comment by 1,630 ISMPP members and about 1400 investigators and journal editors involved with the Medical Publishing Insights & Practices (MPIP) initiative. The author group for GPP3 was selected from over 200 applicants and comprised 15 people from 7 countries working in drug and biotech companies, communication agencies, publishers or as freelance publication professionals – very different from the three of us who authored the original GPP (with input from a working group of about 8 people at various times).
Having been involved from the start, I am delighted that GPP has been so widely accepted and grateful that organizations such as ISMPP and many individual volunteers have helped put it onto a professional footing from its small beginnings. The guideline is freely available from the Annals of Internal Medicine and ISMPP websites.