A Survey to Assess Journal and Reviewer Requests for Evidence in Animals is doing the rounds in the research community around Replacement of animal research. This survey addresses an important question in the discussion of how the research community self-regulates (or not) the use of animals in research. Colleagues who study biological mechanisms such as gene regulation or the role of a given molecule in a certain context have repeatedly commented that trying to publish a paper with only work done in non-animal or at least non-vertebrate systems is often met with requests for experiments in animals. People who are not normally using animals in their research are driven to do so to meet these requests and get their work published in the desired journal. Of course, such anecdotal evidence is only relevant to point out that there is an issue to investigate. We do indeed need systematic research to understand how often it happens, in which contexts and to what extent this phenomenon actually drives the use of animals in research.
But is the present survey a credible systematic research effort? Unfortunately, I find two serious reasons to question that Physicians Committee for Responsible Medicine is a credibly author of research into drivers for the use of laboratory animals.
Before I develop my reasoning further, I need to make a reservation. I know – because I just did so! – that if you google “Physicians Committee for Responsible Medicine Criticism” you will find harsh statements against the NGO. I’m not echoing this criticism here, in particular because it comes from sources which are in themselves just as questionable or perhaps even more so.
But PCRM is an advocacy organisation, with an agenda:
Would you trust a study into the impact of meat consumption run by an entity whose mission was to promote beef or pork? I wouldn’t. I apply the same skepticism when it comes to studies run by entities with the opposite agenda. PCRM lobbies against the use of animals in research. Which is a perfectly valid mission for an advocacy organisation. But research and advocacy are not easy bedfellows. And that’s my first reason not to think PCRM is the right entity to run this survey.
To develop my second reason, I will put my peer reviewer hat on. As a senior researcher, I serve or have served on panels which assess applications for research funding for publicly funded research councils in several European countries. Any given call that opens will get many more applications than its budget allows it to fund, and it is important that the proposals are assessed by experts to make sure that the public money is spent on research projects which can be expected to deliver reliable research results. One of the critical questions any reviewer of a project proposal will ask is: Does this team have the relevant expertise and experience to carry out the proposed research? Previous publications by the research team presents the most important proof that they have enough experience and expertise in the field to be trusted with money for a research project.
Assessed on basis of the publications they present, PCRM wouldn’t score high in a funding review. This is what we find under “Our research”: one paper on diabetic neuropathy, six papers on diabetes, seven papers on diet interventions, and 30 “reviews, editorials and additional research”, all on nutrition. Not a single paper on research models, and as far as I can see not a single paper using survey methodology. Would a research council entrust this team public money to use survey methodology to investigate issues affecting model choice? I don’t think so.
Having said all the above, one thing would make me reassess my skepticism: preregistration of the study. No researcher is completely neutral and any researcher is prone to wishful interpretation and even to analysis methods that supports such interpretations. That’s why the concept of preregistration was born. By declaring up front what a study aims to do and which methods it will use, researchers are prevented from ‘sliding’ interpretations of the kind that worry me here. If PCRM had preregistered their study, with a clear declaration of the research aim, the method for data collection and the planned analysis, I would be much less worried. Both because I would know that the authors would be committed to their registered approach, and if the registration was public I could also assess the quality myself.
Unfortunately, I don’t think they have (but I would be happy to stand corrected!) and therefore I’m not taking the survey. But we need the research.