Bf3R Seminar with Professor Malcolm Macleod

The Bf3R seminars address issues related to the use of animal experiments and alternative methods in basic research and toxicology, as well as the 3Rs.

Improving the ethical position of animal research by increasing benefit: Rationale for and features of a research improvement strategy

All research has costs and potential benefits. In research involving animals the costs include the ethical burden of causing harms to animals for the purposes of human benefit. There is general (but not universal) consensus that this ethical burden is acceptable to society, if it is as small as possible and if the prospects for human benefit are sufficiently large. Much attention has been paid to the minimisation of the ethical burden through refinement, reduction and replacement; but until recently there has been less focus on opportunities to improve the ethical balance by increasing benefit.

We and others have shown that the majority of reports of animal research do not tell the research user whether the study was randomised or blinded, reported a sample size calculation, was guided by an a priori study protocol, was selective in the reporting of outcomes measured, used their first choice of statistical test, reported data for all animals included in the study and even whether the hypothesis presented was the one they started with, or evolved in the light of the data presented in the paper in support of that “improved” hypothesis.

We have shown that studies at such risks of bias are associated with larger effect sizes, and that there is a substantial publication bias in most animal literatures. We have shown that journal-based improvement activity has only minor impact unless substantial resources are devoted to this.

Taking a lead from health care improvement activity, we are developing tools to allow the quality of reporting of animal research to be evaluated in real time across a range of stakeholder domains including by institution, funder, regulator, journal or nation. This measurement will provide data for research improvement activity, where for instance an institution might measure performance, implement strategies intended to improve performance, and then measure what impact has been made. Such information will allow much better targeting of, and greater efficiency in, research improvement activities. And by improving the credibility and validity of research findings, the ethical balance between harms and benefits will be shifted towards benefit. I will sketch out the rudiments of an institutional research improvement strategy.