Fear and Loathing of Human Subjects
©2002, Tim Lightfoot

Notes on piece: This is a serious essay that attempted to express problems that I saw arising in the rush to further regulate data collection from human subjects. The hoops that researchers have to jump through just to ask people questions are amazing.....
Fear is not an emotion that is generally associated with Institutional Review Boards for Human Subjects Research (IRB) at most institutions. However, with the recent “emphasis” on consent forms from the U.S. Office for Protection from Research Risks (OPRR) that have led to shut-downs in research at UCLA, Duke, and the University of Illinois - Chicago, there is an atmosphere of fear that is starting to color the discussions and mandates of IRBs. OPRR is sending strong and clear messages to research institutions across the country that they need to be aware and make sure that the IRBs toe the line when it comes to human subjects research. In particular, it is apparent, at least in my institution and in recent articles, that a special emphasis is being placed on informed consent. It is also apparent that OPRR has sent word that an institution that ignores or does not respond to OPRR in a timely manner, will be punished swiftly and severely. With millions of dollars in research monies often riding on OPRR’s blessing, it is no wonder administrators at research universities are afraid.

What is being forgotten is that the end-effect of this enforcement campaign is the often sudden and complete halt and delay of many researchers’ work. As an example, I recently received the third “word-smithing” mandate from my institution’s human subjects Institutional Review Board (IRB) that demanded that I remove certain phrases from my consent form before giving me approval for a recent grant submission. The irony is that these phrases had been approved in consent forms as recently as three months ago. While the details of my recent struggles with my IRB are not relevant in this essay, the fierce debates that had and continue to go on between myself and the IRB, as well as similar horror stories from colleagues,led me to reflect upon the creeping fear and trepidation regarding human subjects research within Universities across the country.

It is interesting to me that NIH OPRR has chosen to make informed consent a topic of emphasis, when so little is known about informed consent. A little over two years ago, NIH put out a request for proposals to investigate several different aspects of informed consent. Seeing this proposal request, and with my 15 years of human subjects research and my five years of experience on an IRB (two of which where I was chair), I quickly started to pull all of the literature regarding informed consent with a special interest in determining how to increase comprehension of consent forms. At that time, and to this day, I still am amazed at how little is known about informed consent and how to insure its success. Therefore, I find it particularly amusing that OPRR is completely scaring IRBs across the country regarding a topic where so little is known.

I have always been taught and have taught that the purpose of informed consent is to give the subject/patient enough knowledge so that they can understand what they are consenting to do with the research. I strongly believe that it is an inherent responsibility of the researcher to help the subject understand the material to help increase the knowledge level of the subject before they sign the consent form. Therefore, it appears that the crux of the issue becomes how to present informed consent forms in a way that will maximize understanding by the subject/patient, so that they are truly informed enough to give consent. However, I now believe that this is an outdated notion; one that is virtually impossible on its face, and a notion that will never be accomplished due to the confusion, misunderstandings, and fear facing the IRBs.

If we are to ask someone for “informed consent”, we as researchers must also be informed. While that statement seems axiomatic, I’m not sure that it is truly possible. As researchers, we all look for some form of the truth. As such, because we are trying to discover “new” knowledge, we really do not know the outcome of our experiments before we do them. Therefore, many of our informed assertions about what will happen during an experiment, are merely our best guess based on previous literature. I am reminded of an experiment that I conducted with heart transplant patients several years ago. The literature suggested that these patients, because of the reduced nervous connections to their hearts, would not shows signs of fainting if physiologically provoked in a particular manner. However, to our surprise, and given the initial reviews of our subsequent manuscript, much to the surprise of our reviewers, we found that these patients reacted much the same way as did healthy patients. My point is, how can I assure a subject/patient that what I put on the informed consent as the probable outcomes is actually the probable outcome when that question is the object of the experiment?

Another issue that interferes with the notion of informed consent is probable versus possible outcomes. What do I put in my consent form? All of the probable risks and outcomes? Or do I put in every conceivable possible risk and outcome in the interest of the fullest disclosure possible? Taken to the absurd, do I also describe the possible risk that the subject/patient faces in driving to my lab and walking into my building? While this latter statement may sound foolish, the risk of a subject being in an accident on the way to my lab is often higher than the risk the subject faces once they reach my lab. However, some of the tests that we do in my lab could have possible serious consequences. All of the possible serious outcomes are not probable, or I would not ask subjects to do them. As an example, I often conduct maximal exercise tests on my subjects, most of which are young, healthy males and females that have been thoroughly screened. Even though the chance of a cardiac problem during an exercise test is relatively small (1/187,000 person hours), and in spite of our extensive cardiac monitoring during testing, I have recently been mandated to inform my subjects of every possible cardiac problem that might arise during treadmill testing. So, in spite of the chance of a cardiac problem during exercise testing being equivalent to only once if I tested every hour of every day for 21 years, I am required to list in detail every possible outcome and risk. In this case, the chance of my subjects having being a fatality in an automobile accident is much greater (chance is 1/3427 people) than the risk during an exercise test. As such, do I also list the risk of daily activities that the subjects will go through to get to our lab?

As scientists, we often judge the outcome of an experiment, treatment, or intervention based on a calculation of the probability of our results occurring in an appropriate population. Normally, if we see an event happen less than 5% of the time (in some fields, less than 1%), we judge this result to be generalizable to a whole population. Therefore, why do some IRBs insist on making principal investigators list every conceivable risk, no matter how far-fetched or improbable? Again, I believe that it is the atmosphere of fear that is being propagated by the OPRR. Whatever the genesis, the requirement to list every possible outcome in a consent form can actually hinder research.

If indeed we are required to list every possible outcome, we may scare subjects away by listing truly improbable outcomes. In some situations, the anxiety and fear induced by reading the consent form can be overcome with indications to subjects what in fact the chances are of these improbable events happening. However, our IRB will not allow an investigator to put such sentences in a consent form. In a recent experiment where we were going to use a drug that has been used for 30 years with no probable side-effects seen in healthy humans, we were required to remove a sentence that stated that in healthy humans this drug was considered safe (after we listed every possible outcome, of course). We were told that the use of the word “safe” was judgmental on our part. So it is now apparent that at least our IRB not only wants us to list every possible outcome, but also prohibits us from reassuring the subject about the safety of our procedures. Is this what OPRR wants?

Informed consent surely rests on comprehension of the form that the subject/patient is asked to read and sign. It is with this comprehension that the subject is supposed to be able to make an informed choice about whether to participate in a study or not. But comprehension is a slippery thing, not easy to measure and influenced by many factors, some controllable and some uncontrollable. Most IRBs like to encourage investigators to write consent forms at the 8th grade level or even lower. The advent of grade level calculations in most word processing packages make that task even easier. However, I have struggled, as I know many of my colleagues have, in trying to explain complex phenomena in low-level, lay language. Scientists often use a different vocabulary with each other to communicate complex ideas with accuracy and to insure understanding. When we take complex ideas and have to explain them at the 8th grade level, we lose accuracy and preciseness and as a result, we reduce the amount of information that the subject can understand and thus, be informed about. As I’ve raised this issue, I’m often told, “oh, you don’t have to explain the details”. But you try explain to a subject that you are going to collect their heart rate so you can get an indication of their parasympathetic nervous activity. There are any 8th grade level words that convey the essence of the last sentence. Invariably, reducing the grade level of the writing, reduces the possible informed state.

An often unconsidered side-effect of reducing the grade level of the writing in an informed consent is an increase in the length of a consent form. Lengthening the informed consent further compounds difficulties with comprehension and decreases the effectiveness of the informed consent. It is fairly well-known that individuals, especially with only 8th grade educations, have difficulty comprehending long documents. Thus, we are presented with another conundrum: when an informed consent form is written at the 8th grade level, the shorter a consent form, the less the information, and the longer the consent form, the less the comprehension. No matter what the IRB decides on, the Investigator has no choice but to put forward a flawed consent form that can not possibly completely inform the subject.

But what about comprehension? Should we be worried about consent form grade level and length and the impact on comprehension? Most researchers and all IRBs would say yes. However, I’m not sure we should be worried about factors that may effect comprehension if we do not all agree on what level of comprehension we expect. Do we expect 60% comprehension levels for subjects to be well informed, 75% comprehension, or 100% comprehension? There is no consensus on this topic. The literature reports that subjects exhibit a 70% level of comprehension even when videotapes, low level consent forms, and oral orientations are involved. Is this the best that can be done and is it adequate? I’m not sure we should expect high levels of comprehension, even if all other factors can be satisfied. Our own students are cases in point. No teacher, even with all the tutoring and study guides, expects all of their students to comprehend 100% of the material. Heck, I’m thrilled if the majority of my physiology classes, even with three weeks of study, understand (not memorize) 85% of the material. If we do not expect all of our students to comprehend all of the information put forward, even though they are actively studying, coming to class, and making an effort to comprehend the material, why do we expect subjects/patients, with no prior exposure, to read a consent form and comprehend a large percentage of that document? It seems that our expectations of informed consent may be unrealistic.

So is total informed consent impossible? I think that the answer to this question lies in our expectations of informed consent. If we have a naive view of informed consent and expect every subject to be completely informed and comprehend all of the possible and probable risks associated with an experiment before deciding to participate, I believe we have set an impossible task for informed consent. However, if we have a realistic view of informed consent as a process where we appreciate that there are trade-offs between possibilities and probabilities, grade level and length, and comprehension and human limits, IRBs, OPRR, and investigators can develop standards and expectations that allow subjects to be protected and science to progress.

Back to writing index