Written Evidence Submitted by Professor Adrian Smith

(RRE0027)

I would like to provide my personal views on the subject, which are relevant to the following bullet points in the call:

• the issues in academia that have led to the reproducibility crisis;
• the role of the following in addressing the reproducibility crisis:

- research funders, including public funding bodies;

- research institutions and groups;

- individual researchers;

● what policies or schemes could have a positive impact on academia’s approach to reproducible research;

Disclosures
I am a British veterinarian (a graduate of Cambridge University). I have lived in Norway for the last 40 years and I have worked in the research animal community for all of that time[1]. I held the Chair in Laboratory Animal Science at the Norwegian School of Veterinary Science from 1988 until I left in 2011. During that time the Laboratory Animal Unit achieved full accreditation from AAALAC International[2], with whom I have no other connection.

I have been Secretary of Norecopa[3] from its foundation in 2007 until today. Norecopa provides over 8,000 webpages of global resources for implementation of the 3Rs (Replacement, Reduction and Refinement of animal research[4]). I issue a detailed newsletter of the latest advances in this field, 7-8 times a year[5]. These include numerous citations of papers designed to cast light on the reproducibility crisis and solutions for it. I have arranged over 50 courses in Laboratory Animal Science for researchers and animal care staff. I have also worked for the Norwegian regulatory authorities.

I am lead author of the PREPARE guidelines for planning animal experiments[6].

Due to shortage of time, the views expressed in this Evidence are mine alone and not necessarily those of Norecopa. Also for this reason, many of the citations are from my own publications or Norecopa's website.

Evidence

The Committee will no doubt receive extensive contributions from scientists who, rightly so, will emphasise the need for better experimental design and more detailed reporting in scientific journals. Their concerns have been well documented[7]. My main message concerns preclinical animal research and testing, and consists of four points:

1. Experimental design consists of far more than the more "mathematical" issues such as randomisation, blinding, group size, power levels and correct statistical analysis.
2. Reporting, however good, cannot improve an experiment which was incorrectly carried out. Scientists must be encouraged to improve their experiments from day one of planning, in close collaboration with the animal facility where these are to be carried out. A burnt cake cannot be improved by a better description.
3. Reporting guidelines are inadequate for planning: there are many issues, not least those related to the animals themselves and the facility in which they live, which do not need to be reported but which are critical to the validity, reproducibility and translatability of the data obtained from the animals - and not least to animal welfare and health & safety.
4. Planning guidelines should be used by funders, animal welfare and ethical review bodies, animal facilities and scientists, in collaboration, to work together for better research. Attempts to improve the situation at the publication stage are too late - unless the journal undertakes preregistration and peer review of protocols.

In other words, bespoke guidelines for planning animal experiments are needed, in addition to guidelines for reporting (which serve another purpose, although there will be some overlap). I and my co-authors published PREPARE, a set of general guidelines for planning animal experiments in 2017, based on our experience over a 30-year period, to satisfy this need[8]. Guidelines for specific research areas are also emerging, e.g. DEPART for osteoarthritis research[9].

In my experience, many scientists are unaware of the plethora of issues and details which can affect the quality of an animal experiment - or they assume that these have been considered by the animal facility (which is not always the case). Routine, and apparently innocuous, procedures such as housing, handling, injections and blood sampling can introduce variables into an experiment, and stress for the animals, which is not only a welfare problem but also a large potential cause of poor reproducibility. These effects are now being widely documented in the literature, as the website of the PREPARE guidelines demonstrates[10]. I have also described this in a recent paper[11].

Over 1,000 journals have endorsed the ARRIVE guidelines for reporting animal experiments, although compliance with these guidelines is not satisfactory[12]. In my opinion, the way ahead is to encourage scientists to seek guidance on planning animal research and testing, at the earliest possible stage. This should be done in a way which demonstrates its importance, without it being construed as yet another bureaucratic hurdle. The incentive is that they will then discover that they are in a far better position to tackle the questions which reviewers and editors will raise when they submit their manuscripts to a journal for publication:

"we have ARRIVEd because we were PREPAREd".

Therefore, I would encourage funders, animal welfare and ethical review bodies, and research animal facilities to promote planning guidelines, in the hope that scientists will perceive them as well-meant advice. Even experienced senior scientists will hopefully recognise them as a useful aide memoire, much in the same way as experienced pilots still value the checklists they use on routine flights. We are in the process of collecting endorsements for this approach to the PREPARE guidelines[13]. These currently include citation in the annotated UK Home Office Project Licence Application Template[14] and in a recent report from the Advisory Committee to the Director of the US National Institutes of Health[15].

(September 2021)