Liz Wager: GPP3—thoughts on becoming a guideline grandmother


The third version of the Good Publication Practice guidelines (GPP3) were published last week which makes me a guideline grandmother. Like any grandparent I am excited and proud of my “offspring,” slightly nervous about how the world will treat it, but hopeful that it will grow up to do something useful. And, like a grandmother, I can’t resist telling you some family history.

The birth of the original GPP wasn’t easy. It was conceived in 1998 at a meeting of journal editors, clinical trial investigators, and drug company publication people. This meeting to discuss publication concerns was the brainchild of Leni Grossman (who worked for Merck) and was organized by the Council of Biology Editors (now Council of Science Editors). After that meeting, a few people from drug companies got together and drafted some guidelines. We were a self-appointed group (although happy to welcome other volunteers). We had no funding but our employers didn’t object to the project so we could do some of it in office time (at the start of the project, I was working as a medical writer for Janssen Cilag, but later moved to head the international publications team at Glaxo Wellcome). We discussed drafts with a working group of about a dozen people (and those who had attended the CBE meeting) and got helpful feedback. We attempted wider consultation by sending the draft to about 70 drug companies but this seemed to generate more hostility than support. An “informer” sent us worrying evidence that some companies’ legal departments were attempting to discredit and even block our work by suggesting that companies endorsing the guidelines might be viewed as an illegal cabal. We thought this was ludicrous (since when has it been illegal for companies to get together to raise practice and do something ethical?) but it was still scary. We had suggested to companies that doctors might trust them more if they signed up to a guideline that committed them to publish results of all clinical trials of their marketed products. Had we been naive, or had our tactic backfired horribly? In the end, we persuaded nine companies to endorse GPP. Other companies didn’t want anything to do with the guideline—so much so that we had to publish with only three authors because the fourth person, who had made a substantial contribution, was forbidden to be named by his company.

Drummond Rennie had taken part in the CBE meeting, was supportive of our efforts, and encouraged us to submit the guideline to JAMA. We took his advice but were rejected after negative comments from a reviewer who felt that the guideline offered nothing new, didn’t go far enough, and should probably be published in an “industry journal.” We then submitted the manuscript to The BMJ who sent it to the same reviewer who, unsurprisingly, said the same things, so it was rejected again. Given companies’ resistance to our proposal that they should publish results of all their trials we howled in anguish to be told our guideline didn’t add anything. Irritated but undaunted, we sent the guidelines to Current Medical Research & Opinion (CMRO). This may have been what the reviewer meant by an “industry journal”, as it did cater largely to drug companies and levied page charges (this was long before the days when Open Access made article processing charges respectable, and such “pay journals” were scorned by many, but this one was, and is, properly peer-reviewed and indexed on Medline). CMRO not only published the guideline (after more constructive peer review) but helped us by waiving the page charge and producing lots of reprints (at no charge) which were distributed at conferences.

But it still wasn’t plain sailing. During the delays caused by the rejection from JAMA and The BMJ, the US industry association, PhRMA, had developed its own guidelines. Annoyingly, these watered down our key requirement that results of all trials should be published. The PhRMA guideline only required companies to publish “meaningful” results leaving a big loophole for inconclusive or negative findings and an unanswered question of who would determine whether findings were meaningful or not. However, given a choice, it was obvious that many companies would prefer the less demanding PhRMA guideline and they then gave this as a reason for not endorsing GPP. On a personal note, by the time GPP was published in 2003, Leni Grossman had retired from Merck, and Betts Field and I had been made redundant following the GlaxoWellcome merger with SmithKlineBeecham. This meant that all three authors lost the little corporate support we’d had earlier and were now, effectively “on our own.”

Despite our lack of status or corporate backing (or funding) and frank opposition from some companies, GPP was published and gradually people started noticing it. Despite declining to publish it, BMJ was one of the first journal publishers (along with BioMed Central) to include a link to the guidelines from their instructions and encourage authors of industry-funded studies to follow them. These endorsements were encouraging and helped make researchers aware of the guidelines. Although drug and device companies were still reluctant to endorse them, medical communications agencies were more enthusiastic and started to spread the word via their websites and in their work on company-sponsored publications. Even without the benefits of social media (neither Twitter nor Facebook existed in 2003) and with no budget or paid time to spend on the project, with the help of friends and some notable supporters, such as Iain Chalmers, we nudged forward.

By the late 2000s, the good publication practice (GPP) guidelines were widely known but were starting to feel a bit dated. Although we had called for all trials to be published, we hadn’t demanded that clinical trials should be registered, but, following the ICMJE requirement in 2005, and US legislation in 2007, this was now the norm. While we were the first guideline to outline the role of professional medical writers, we had said little about other publication roles, such as planning. A new organization, the International Society for Medical Publication Professionals (ISMPP) had been set up in 2005 with the aim of raising professional standards and sharing good practice. I was delighted when ISMPP suggested it would coordinate an update of GPP. Although I was tempted to stay involved, I was busy running my own company and heavily involved with the Committee on Publication Ethics (COPE) as a volunteer. I also felt it was important that the guidelines maintained their roots in the pharmaceutical industry but I was now mainly doing training rather than medical writing. I also knew that founding fathers (or mothers) can sometimes be a nuisance when initiatives or organizations need to move into a new phase, so I decided to cheer from the sidelines.

ISMPP recruited a strong team, including one of the original authors (Betts Field), but with much broader representation from publishers, device companies, and communication agencies. However, all but two of the authors were based in the US (the other two were from the UK). Ably led by Chris Graf (from Wiley), the development of GPP2, and in particular the consultation and reviewer input, was much more professional and thorough than our original efforts. The team sent a draft to 193 people and got 116 sets of comments which they painstakingly reviewed and considered.

The BMJ saw the error of their earlier ways and published GPP2 in late 2009, so it was good to know it wasn’t only fit for an “industry journal” despite the support we’d had from CMRO. The title also changed from GPP for pharmaceutical companies to GPP for reporting company sponsored research, reflecting the fact that these guidelines were for investigators as well as companies (with hindsight, perhaps we should have realised and used a bolder title from the start).

GPP2 was considerably longer than the original guideline. New sections were added, with more guidance on publication planning, authors’ access to data, and documentation. The guideline supported trial registration but, to my disappointment, the requirement to publish all studies got relegated to a checklist and didn’t seem so clear. But GPP2 had an impact. Companies used it for audits (something we had never envisaged when we first had the idea for GPP) and it won acceptance as “the” industry standard. A survey of 469 publication professionals in 2013 found that over 90% consulted GPP2 regularly — that felt like an achievement! We had originally picked the name, almost cheekily, to mirror industry codes such as Good Clinical Practice and Good Manufacturing Practice but, unlike these, which were based on laws and regulators’ requirements, GPP was a voluntary guideline with no “teeth,” but evidence suggested it was taken seriously.

And now, less than six years later, we have GPP3. This time, I couldn’t resist the urge to jump back in and be a nuisance. It was hard but rewarding work. The steering group had members not only from all over the US, but also Europe and the Asia-Pacific region. This meant late-night or very early morning phone conferences for some of us. Getting nearly 20 professional writers and editors to agree on both content and wording (not to mention punctuation) is tougher than herding cats, but Wendy Battisti (of Janssen-Cilag) accepted the challenge and coordinated us with efficiency and good humour.

We have added sections on data sharing and plagiarism and more detail about determining authorship. Another new feature of GPP3 is a series of principles which we hope will be helpful for training, will counteract the problem that guidelines tend to grow as they get older, and should help address new problems that arise in future. While still a voluntary guideline, the tone of GPP3 is more confident than that of GPP2 — many of the statements that were prefaced by “we recommend” have been replaced with assertions about what companies, investigators and writers “should,” or sometimes even “must,” do. (This may have something to do with the fact that the stroppy grandmother did two rounds of editing after we had agreed the content)

Everybody says that grandparenthood is more fun than parenthood and I do feel more relaxed about GPP3 and perhaps less defensive than I did when the original GPP was published (although I now realise that if I was the mother of the original GPP guidelines then GPP3 makes me a guideline great-grandmother, but that sounds too elderly, so I’ll stick to granny). However, us oldsters must realise that, however much we dote on our intellectual offspring no guideline is perfect and documents must keep up to date so we should look forward to future generations of GPPs. I also hope that I will have the sense and humility to retire from their rearing before I get too ancient and out of touch myself. But for the moment, I’m enjoying being a guideline granny!

You can judge the result yourself.

Liz Wager PhD is a freelance medical writer, editor, and trainer. She was the chair of the Committee on Publication Ethics (COPE) (2009-2012).

Competing interests: The author has no further relevant interests to declare.