NIH’s 7 principles relating to clinical research ethics

Use these benchmarks during review process

The NIH has created a framework and benchmarks program for evaluating research ethics. Christine Grady, PhD, RN, head of the section on human subjects research in the department of clinical bioethics at NIH, provides these insights into the seven principles outlined in the framework:

1. Value or social value: "The simplest way of understanding it is to ask whether this research question has any value either socially, scientifically, or clinically, and whether it’s worth answering," Grady says.

For example, a research ethics expert once told Grady about a nightmare he had in which he was in the hallway of an institution where a colleague described having successfully transplanted an appendix into a research participant, Grady recalls.

The question is whether this transplantation was an ethical thing to do, Grady says.

The colleague answers that question by saying he received informed consent, but that’s not proof of research value, Grady notes.

"Unless there is a good valid reason to understand what a transplanted appendix would do then this is not an ethical study," Grady says. "There is no value to this intervention unless you have a hypothesis about it."

Another example of research that might be put into the questionable value category is me-too studies in which there already has been conclusive investigation into the research question, she says.

"If you already have answers to a question, how valuable is it to answer the question again or to answer it in a slightly different way?" Grady asks.

2. Validity: Once an IRB has established that the research has value, then the next question is whether the proposed study has an acceptable scientific design, statistical methodology, implementation strategy, and feasibility so that the research question can be answered and interpreted, Grady explains.

"If the question is worth asking you have to ask it in a way that you can interpret the answer," Grady says. "If you can’t design the project in a way that will get you an understandable answer, then the question is not worth asking anyone to assign any kind of risk in order to answer it."

For example, in the 1990s, there was a great deal of discussion about the effectiveness of giving antiretroviral treatment to health care workers who had been stuck with a needle that might have been contaminated with HIV. Post-exposure prophylaxis was the term used for this type of treatment.

"It’s an important question to answer, but the design of such a study seemed to be impossible because the incidence of transmission was so low that you would need an enormous number of people to answer the question," Grady says. "And you would never have enough of a sample to get the power needed to answer the question in a vigorous way."

So a randomized control trial was never conducted, and instead public health officials relied on data from case reports, she adds.

3. Fair subject selection: "This is basically the notion that people should be selected to participate based on a scientific question and not for other reasons, such as favor or disfavor or ease of manipulation," Grady says.

"Once scientifically appropriate participants are identified there still needs to be consideration of vulnerability and risk and benefit, and so you balance the need to fairly select people with the need to minimize risks with the people who are selected," Grady says.

An obvious example of an unfair subject selection would be the research work done historically in orphanages, Grady says. While an orphanage has an easy-to-contain population, it’s not a good population for any study, she says.

A more contemporary research subject selection issue involves who can be justifiably excluded from a study, Grady says.

"I think that’s a very interesting issue," Grady says. "I have a colleague who talks about how everyone should have a fair chance to participate in research, and the onus should be on how well you can justify your reasons to exclude people based on whether it’s inappropriate."

For instance, some populations or people might be placed at a risk in a particular study and there wouldn’t be any means for alleviating, overcoming, or compensating for that risk, Grady explains.

"They’re so vulnerable that you can’t compensate for that vulnerability in some way," Grady says.

4. Risk/benefit: This principle is more complicated than its label, Grady says.

"The notion is that even after you’ve got a valuable question and a rigorous design that will answer it and a fair subject selection process and a goal, there still is more that can be done to minimize risk and to maximize benefits to the participants in this study," Grady says.

"Minimizing risks and maximizing benefits should be done all the time," Grady says. "But there’s an interesting reality that maximizing benefits does not mean the people participating in the study are going to benefit from the study."

Some studies will provide no benefits to subjects, and these can be perfectly ethical, Grady notes. "However there are lot of things people can do within the context of the study to minimize the risks," Grady says.

For example, some straightforward changes could be to make sure that the number of procedures required of participants are precisely what are necessary to obtain useful data, Grady says.

So if there are four MRIs proposed, but three MRIs would be adequate for obtaining the necessary data, then the study should have three MRIs instead of the four, she explains.

"Similarly, minimizing risks means making sure that interventions, procedures are done when clinically indicated," Grady says.

For instance, when a research study requires a tube of blood for a research assay, it’s better to have that blood drawn at the same time the patient is providing blood for clinical care so the person doesn’t have to undergo two blood draws, Grady says.

"Some of these changes are very detailed, but they can make a huge difference in terms of the risks people are exposed to," Grady says.

"Certainly the IRB makes the judgment about whether the risks you can’t take out and can’t minimize are justified by the value of doing the study."

5. Independent review: While this is a procedural requirement, it is one that has other benefits, such as public accountability, Grady notes. "It’s a way to check both enthusiasm and potential conflicts on the part of investigators," she says.

"Understandably and actually in a very positive way, people who do research are often enthusiastic about the research they’re doing," Grady explains. "Enthusiasm without careful methodology could lead one to their own conclusions, so an independent review of the proposal and methodology and statistical analysis plan serves to check that kind of enthusiasm that could lead to the wrong conclusion."

Independent review is a hot button issue right now, Grady notes. "The IRB system has been widely adopted not only in our country, but in many places around the world," Grady says. "It makes a lot of sense on one level, but people have argued, and there is a case to be made that it was based on research that was done at an institutional level."

IRBs worked most effectively for smaller, institutional research, but the complexity of the multi-site and multi-national research of today has raised questions about whether the current system is the correct one, Grady says.

"One example that people talk about all the time is the multi-site study that has 64 different sites, for instance," Grady says. "Under the current system many of those studies will have to go through 64 IRBs, and the time table for each might be different, and the exact things an IRB wants changed or altered might not coincide across the 64, and it creates a huge lag in doing the research."

While the FDA allows for a central review with one IRB of record as long as all sites accept that IRB, this isn’t done as often as it probably should be done, Grady says.

6. Informed consent: Some investigators and IRB members might question why informed consent is in the sixth position instead of being placed earlier in the framework since it’s typically considered one of the most important ethical issues, Grady notes.

"But if we’re right about this framework, you would never get to the informed consent unless other things are satisfied," Grady says. "If you don’t justify the risks and benefits then you never ask people to participate, so while informed consent is an important part of doing research and a universal requirement, it’s not where you start."

Informed consent is based on people’s ability to make their own decisions, and they need the right amount of information to make this decision, Grady says.

"I guess one point I would make to reviewers of research proposals is there seems to be a lot of attention paid to the details of the information written into a consent document," Grady says. "Whereas, I think it’s pretty clear that the process around giving information and helping people understand it is so much more important."

A study could have a pristine consent document, but people might not read it, so the focus should be on what people are told and the context and slant of this discussion, Grady says.

"That’s much more important to the ethical evaluation of the study than what the words say on the paper," Grady says.

7. Respect for enrolled participants: "This category reflects the notion that even after you have carefully designed a study and individuals have consented to participate voluntarily, there still are things one needs to do to show respect for the individuals who have chosen to participate," Grady says.

"It includes monitoring the well-being of the individuals in the study, protecting their confidentiality, making relevant information available that people would need to know to make a relevant decision about continuing participation, and they need to be reminded they have the right to withdrawal if they want to," Grady says. "There are a host of things one needs to do to show respect to individuals in a study."

An infamous example of research that lacked the ethical respect for participants was the Tuskegee study in which African American men were enrolled as part of a study of the symptoms of syphilis.

When the study began antibiotics were not a proven treatment for syphilis, but as participation in the study continued over decades, the research participants could have received antibiotic treatment but were not informed of its availability.

"I think Tuskegee is a good example of a lack of respect for enrolled subjects because as information became available over the course of many years, it was relevant to their well being, but it wasn’t shared with the participants," Grady says.