A guide to the perplexed: how to navigate conflicting research ethics policies.

• Author: Kimmelman, Jonathan

Genesis

In the beginning, there was the Nuremberg Code. But in 1964, the recently established World Medical Association saw that Code, that it was not good. So World Medical Association established its own code, the Declaration of Helsinki, and the Association declared to the rest of the world: lo, here is the standard against which all human experimentation should be held.

In time, Declaration of Helsinki begat a son, a new version of itself, which begat versions three on to eight. These also begat Council for International Organizations of Medical Sciences's guidelines (CIOMS). Now it came to pass that there was a distant land, seemingly untouched by atrocity. But this land too partook of forbidden research practices, and thus begat Belmont Report, which begat 45 CFR 46. And in the land of Canada, the Medical Research Council begat Guidelines on Ethics for Medical Research, which then begat the first version of Tri-Council Policy Statement (TCPS), which begat a second version (TCPS2). But the U.S. Food and Drug Administration declared World Medical Association a false prophet, and so laid with regulators in the European Union, Japan, as well as pharmaceutical manufacturer associations in these three jurisdictions. And this union was called International Conference on Harmonization (ICH), and from its loins sprang Good Clinical Practice (ICH-GCP), which smote Helsinki.

And it came to pass that two ethics committees encountered a research protocol. One declared its loyalty to the TCPS, implored investigators for more clinical equipoise. The other declared its loyalty for Health Canada, which dwelled in the land of ICH-GCP. It implored for more placebos. And soon the two found their speech confounded. And a darkness and confusion settled over the practice and policy of research ethics.

The Diversity of Research Ethics Policies

Despite 75 years since their first formalization, research ethics policies continue to show a remarkable degree of stability and consensus. Emanuel et al. identified seven core points of consensus that join the main, influential policies governing human research ethics, including: value, validity, and independent review. (1) Yet, upon examination at a finer level of resolution different substantive features such as: goals of the oversight, audience, level of detail, length, style, use of placebos and national or international mandates emerge as points of divergence. The policies also diverge on certain prescriptions. Some of these are detailed in Table I.

The International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use's Guideline for Good Clinical Practice (ICH-GCP) permits use of placebo controls except where permanent or serious harm is anticipated. (2) Instruments like the current version of Tri-Council Policy Statement (TCPS2) (3), as well as statements like the National Bioethics Advisory Commission's report on research involving human participants (4) take a far more restrictive stand. Another point of divergence is how issues of fairness are addressed. For example, the Belmont Report (5) contains little on international research (though it posits a principle of justice), whereas the Declaration of Helsinki, (6) the CIOMS guidelines (ClOMS), (7) and other policies address international research head on. Yet another example is that ICH-GCP contains practically no language on justice and fairness, whereas the Declaration of Helsinki, CIOMS, and TCPS2 do.

Allowable risk for persons lacking decisional competence is also contended. The Declaration of Helsinki restricts non-therapeutic research procedures to minimal risk, whereas the American regulations under 45 CFR 46 (8) entertain conditions allowing greater risk. Some policies also provide different definitions of minimal risk, whereas the Declaration of Helsinki provides none.

These differences present major navigational problems to those charged with planning, designing, reviewing, and conducting clinical research. Whose policies ought to be followed? What is the moral force of one policy against another? When studies are pursued across multiple jurisdictions, how should conflicting moral guidance be reconciled? Should decision-makers feel free to practice a la carte research ethics, drawing on preferred language in various instruments? If so, what meta-rules should govern which policies are sampled? To what extent should authorities strive to harmonize different instruments, or does diversity reflect an appropriate plurality of moral positions?

Policies as Tools

We are not prepared to offer a thoroughgoing response to the many moral and policy questions that stem from the existence of conflicting prescription in research ethics. Instead, we point to one way of engaging competing moral policies. One simplistic approach is to pick the instrument that is the law of the land. This policy won't get us too far, as jurisdictions often subscribe to several policies, research spans borders and actors often operate across multiple jurisdictions (e.g., international clinical trials which may come under standards set by home and host countries) or fall into cross-cutting categories (e.g., physician-researchers). Thus, an academically appointed physician-researcher conducting clinical trials in Canada may come under the aegis of a wide range of policies -ICH-GCP, of the Food e& Drugs Act Regulations (9) TCPS2, The Declaration of Helsinki and CIOMS. Moreover, regulatory policies establish a floor below which practices may not fall. However, nothing prevents one from abiding by more demanding standards. Indeed, one can imagine using such a mechanism as the Canadian General Standard Board's process to ratchet standard upwards. (10)

Another approach is to pick policies according to preferences of review committees, or to patch together a sort of composite policy. As previously indicated, here we would...