Monday, September 29, 2008

Risk - Part 2

In addition to identifying risks, IRBs must ensure that the risks are minimized, that the risks are reasonable in relation to the anticipated benefits, and that subjects are informed about reasonably foreseeable risks.

Minimizing risks does not mean eliminating risks. There is nothing in the Belmont Principles or the regulations that requires that research be risk free. Research sometimes involves risk to subjects. Research ethics demands, however, that the risks be the minimum necessary to obtain valid results. No researcher wants to subject people to unnecessary risks. Researchers, though, need to actively consider whether the risks in their research are actually minimized. How do we minimize risks? One way is to explore alternatives – is there another way to do this research that would be less risky? A second way to minimize risk is to add precautions to the research. These are procedures that reduce the likelihood that the harm will occur, such as additional screening or subject preparation prior to the research. Sometimes, however, the risk in the research is unavoidable. In this case, researchers need to build in contingencies that will reduce the impact of the harm when it occurs, such as increased monitoring.

Analyzing the risk/benefit relationship is one of the key components of research ethics. It is not ethical under any circumstances to inflict risk on subjects unless those risks are justified by the benefits that are expected to accrue from the research. Even the most zealous scientist would not engage in research that might harm subjects unless there was some good which could come from the research. The more risky the research, the more benefit it takes to justify that risk. Not all research, however, must involve direct benefits to subjects. Although the regulations talk about risk to subjects, benefits may be to subjects or others. Advancing scientific knowledge or improve social welfare are benefits; even training students is a benefit. Direct benefits to subjects, however, are stronger benefits than benefits to others and, therefore, more risks are justified.

Finally, it is required that subjects be informed about the risks and benefits of the research. These need to be described to subjects in a way that they can understand and make an informed decision as to whether these are risks they are willing to take. As I said in my first post on risk, simply listing the possible harms in a research study is not informing subjects about the risks.

To sum up, risk is a very misunderstood and misapplied term. We must make sure not to confuse risk and harm and remember that simply listing possible harms is not informing subjects about the risks in research. We must also make sure that the research involves the least amount of risk necessary to obtain valid results, that the risks are justified by the benefits of the research and that subjects are clearly informed about the risks and benefits of the research. Only then will we, researchers and IRBs, be fulfilling our ethical responsibility to the subjects of the research.

Tuesday, September 23, 2008

Risk - Part 1

Risk is one of the central concepts in human research protection and is one of the most misunderstood and misapplied terms in the field. Many IRBs and investigators treat risk as if it were the same as harm to subjects, which is why the IRB Forum had a recent discussion about the use of the phrase “possible risk” in consent forms. Risk is not harm; risk is an estimate of the magnitude and probability of the harm to subjects that might happen in research. “Your hair might fall out” or “you might go blind” are not risks, they are possible harms. Simply telling subjects that one or both of these might occur is not informing them of the risks.

In order to determine the risk of these events, you need to evaluate both the severity of the possible harm and the likelihood that the harm with occur. A very severe harm that is extremely unlikely to occur may be a minimal risk, while a less severe harm that is very likely to occur may be more than minimal risk. Clearly, the possibility of going blind is a very serious harm, but if the chances that it will happen are one in five million, then it may be a minimal risk that subjects are willing to take. Losing hair, for most people, is not a very serious risk, but if it is a sure thing, I might have second thoughts about it.

To inform subjects about risks, IRB and investigators must evaluate the risks and make sure that subjects are informed about the reasonably foreseeable risks in the research. This means that IRBs must identify the possible harms in the research, evaluate both the magnitude and probability of those harms, and make sure that subjects are given adequate information about what harms might occur and the likelihood that they will occur. Again, simply listing possible harms is not informing subjects about the risks in the research.

In fact, just providing subjects with a long list of possible harms not only doesn’t inform them about the risks, it is “disinformation”. When subjects see that list of harms, which in some cases goes on for pages, they tend to discount all of it and don’t pay attention to the real risks in the research that they need to focus on to decide whether to participate in the research. Consent forms that include such a “laundry list” of harms without indicating which are likely to occur and which are less likely, are not in compliance with the regulations and should not be approved by IRBs. If sponsors insist on using such consent forms, IRBs should require other mechanisms to explain the real risks to subjects, such as oral presentations and information sheets.

Informing subjects about risks is not the only responsibility with regard to risk that IRBs have; they must also ensure that the risks are minimized and that the risks are reasonable in relation to the anticipated benefits of the research. I’m going to discuss those two responsibilities in another posting.

Tuesday, September 09, 2008

Checklists

Checklists have gotten a bad rap!! I, too, had a negative attitude toward checklists. It seems like they work against real ethical decision-making by IRBs. But, as I have been working with institutions going through accreditation (AAHRPP is big on checklists), my attitude has changed. I now see that, when used appropriately, checklists serve an important function in the IRB review process. Checklists don't ensure good IRB review, but they do help ensure that things don't "slip through the cracks". In order to do a good review, the IRB must carefully look at each of the criteria for approval in Section 111 of the regulations. Even the best IRB reviewer can miss something and review checklists help make sure that the reviewer considers all of the essential aspects of the criteria when doing the review.

You might say (and many do), that this is just a bureaucratic exercise to comply with regulations and does not do anything to better protect subjects. However, the criteria for approval in Section 111 implement the Belmont Principles (I'm going to do another post on that), so making sure that the review addresses all of the criteria ensures that the ethical principles are adequately addressed.

So, I've become a fan of good checklists, but they should be used as guidance not bureaucracy. It's not about checking off boxes, but about focusing the reviewer's attention on the important factors in ethical review.