HCS Research Collaboratory Grand Rounds 4-15-16

NIH Collaboratory Grand Rounds moderator

for the Collaboratory Grand Rounds. And it's my pleasure to introduce John Lantos, who's going to present on a topic as part of our monthly regulatory and ethics series. So this topic actually has been really relevant over this past year for things as they emerge for pragmatic trials. And the issue is really understanding and considerations and evaluation and determination of minimal risk in research studies. something that seems such a simple thing, but has been quite a challenge for people to understand and implement. So, John, I want to thank you for joining us today. We'll pass it over to you.

Dr. John Lantos

All righty. I am in Kansas City. For those of you who haven't visited the Midwest, the circle here on the map is where we are. You've probably flown over us. This is my hospital, Children's Mercy Hospital. I'm a pediatrician. and run a bioethics center at Children's Mercy. Kansas City is a great place. You should all come visit. It has some of the best barbecue in the world and, of course, is home of the World Championship Kansas City Royals. We had a great celebration after the World Series. And this is my office here in the Bioethics Tower, so I had a nice view of that. My goals for today are to talk about

Dr. John Lantos

definitions of research, quality improvement, and comparative effectiveness research, and I'm going to present a few hypothetical studies, all of which seem to entail similar levels of risk, but by current regulations would trigger very different IRB responses, and we can speculate about why that might be or whether it's appropriate. Going to compare studies of minimal risk with what we know about practice variation, particularly idiosyncratic small area practice variation, which in my view is the baseline against which the risks trying to study such variation comparative effectiveness studies should be caged. Going to talk a little bit about the importance of minimal risk in the federal regs and the current debate about which risks should be attributable to a study as opposed to being risks of the therapies that are being studied and then speculate a little bit about what's really going on in the current debates about riskiness and still hopefully leave a little time for questions.

Dr. John Lantos

So definitions of research QI and comparative effectiveness research according to the common rule. Research is judged by its goals, not by the actual intervention. So research is a systematic investigation, presumably any systematic investigation, including research, development, testing, and evaluation that is designed to develop or contribute to generalizable knowledge. So if you're hoping to learn something, it's research. What is QI? Various definitions. It's not enshrined in federal regulation, but has been defined as a systematic data-guided activity designed to bring about immediate improvements in healthcare delivery in particular settings. So this seems to be a very similar sounding set of activities that often involve the same sorts of interventions And to a certain extent, they both seek generalizable knowledge, although there is a question about how generalizable knowledge has to be before a QI project becomes a research project. QI projects, in that definition, are thought to take place in particular settings, which may be the key buzzword there. Usually the knowledge that results from QI is most applicable to the local situation, but these authors go on to explain, insights from one setting ordinarily have some applicability to other settings. So do QI projects that are applicable to other settings become research? And if so, how should we think about the risks of those projects?

Dr. John Lantos

Comparative effectiveness research falls in somewhere along this spectrum. It's designed to inform healthcare decisions by providing evidence on the effectiveness, benefits, and harms of different treatment options, all of which are within the current standards of care. So let me present a few hypothetical studies with similar levels of risk, but that I think would be dealt with very differently by IRBs in order to highlight just how ambiguous these regulations can sometimes be. And for these, they're all sort of variations on a theme of the recent study that caused much controversy. The support study of oxygen saturation levels in premature babies,

Dr. John Lantos

which was an NICHD-sponsored study, prospective, randomized, clearly research. And the question was, or the controversy was about which risks should appropriately have been included in the consent form. But let's imagine a couple of other oxygen studies that say we're not designed like the support study. For example, what if doctors wanted to determine the optimum level of oxygen to give to premature babies? And to do that, they analyzed de-identified data from the electronic health records at 10 different neonatal intensive care units that were known to use different oxygen saturations. And the doctors knew that five of those 10 had a policy of targeting a higher oxygen saturation, say, 90 to 95%, and five targeted a slightly lower one, say 85 to 90%, and both picked their target saturation based on their own assessment of retrospective studies that appeared in the literature. So this would be an example of some practice variation. And when they did this electronic health record retrospective review, they were looking to see whether at these two NICUs there were different rates of mortality retinopathy of prematurity, and neurodevelopmental impairment among the babies in the NICUs. Of note, in this retrospective study, about half the babies would be treated in NICUs that targeted the higher saturation, about half in NICUs that targeted the lower one. The parents presumably would be unaware of these different treatment approaches, and certainly most parents don't choose their NICU since most premature births are not anticipatable in that way. If they were, they probably wouldn't choose it based on the oxygen saturations. So by current standards, this is research. It doesn't involve human subjects since it involves de-identified data, therefore has no risk to human subjects and could be carried out without parental consent or knowledge.

Dr. John Lantos

Let's imagine a different study. Doctors want to determine the optimum level of oxygen for premature babies, and they find five NICUs that target the lower oxygen saturation and five that target the higher saturation. And they go to the parents and say, we want to enroll your babies in an observational study. That is, we're not going to look at the medical records once you're gone. We're going to track you from the day you arrive and look at the same outcome measures, mortality, retinopathy, and neurodevelopmental impairment. I just got my slides out of order. So this is an observational study. It does require IRB approval and parental consent, but would likely be judged minimal risk because it is purely observational.

Dr. John Lantos

So let's imagine a third study. Doctors at a single NICU become convinced, based on retrospective studies, that lower oxygen saturation targets are safer So on a particular day, say January 1st, they decide to change their protocol so that instead of targeting what they used to target, 90 to 95%, they now target 85 to 90%. That is, up until December 31st, all the babies in the NICU got 90 to 95%. Starting January 1st, they all get 85 to 90. And then they retrospectively analyze data from the two time periods to compare rates of survival, retinopathy, and neurodevelopmental impairment. This would need IRB approval. It's research, but there would be no need for consent, I don't think, because it's using retrospective data and would certainly be judged as minimal risk, even though, presumably, if there were a similar number of babies admitted, the same number of babies last year as this year, half of the babies would get the higher oxygen level

Richard Platt

half would get the lower one.

Dr. John Lantos

At the same hospital, imagine the doctors notice that only 50% of the babies are kept within the targeted range of oxygen saturation, so they implement a QI program that includes education about the relationship between oxygen sats and outcomes, and the program leads to higher compliance with their new protocol. So now 75% of babies are within the target oxygen saturation range. This would be quality improvement, not research, with no need for consent or IRB approval, although it would clearly change the management of the particular babies, and depending on whether you think higher or lower oxygen is safer may have harms or benefits.

Richard Platt

Then there's

Dr. John Lantos

another NICU where doctors want to determine the optimum level of oxygen, so they prospectively randomize babies to targets of 85 to 90 or 90 to 95. They analyzed the relationship between oxygen saturations and the same outcomes. By current standards, this would be viewed as greater than minimal risk. Parental consent would be required. And as noted in the support study, anybody who did that would have to warn parents by enrolling their babies in the study. It could change the risk of death, retinopathy of prematurity, or neurodevelopmental impairment. So of note, doctors presumably do not know which level of oxygen is best in any of these studies. In all the studies, half the babies would get higher levels, half the babies would get lower levels, and the actual risk to the total population of babies in each of these attempts to solve the problem of figuring out the relationship between oxygen and outcomes seem pretty similar. but from a regulatory perspective, they would be viewed quite differently.

Richard Platt

It's also hard to know from any of them how the risks of being in any of the studies compared to the risk of what the babies would get outside the studies.

Dr. John Lantos

The

Richard Platt

situation

Dr. John Lantos

with regard to oxygen therapy in these hypothetical examples is quite similar to the current existing state of much medical practice today. And one of the interesting things about today's

Dr. John Lantos

research regulation, the Common Rule, OHRP, IRB guidelines is that they were all written before we really understood as much as we do today about idiosyncratic small area practice variation. because that wasn't really discovered until pioneering studies by Jack Wenberg in the 1970s and the 1980s, initially looking at particular surgical procedures in different counties in New England. I'll show you some data from those in a minute. But what Wenberg has shown, Wenberg and his colleagues now and many others, is that for almost any medical procedure, where there are options, clinical choices vary in dramatic, seemingly irrational, and totally unpredictable ways, and it's seen for almost anything you look at and almost anywhere you look. Here's a chart that I hope people can see from one of Wenberg's early papers that looked at different counties in Rhode Island, Maine, and Vermont, and at six different surgical procedures. And the x-axis here is the number of procedures per 10,000 people per year in each of those counties. And what you can see is that for things like tonsillectomy, it varied by about 700%

Dr. John Lantos

for hysterectomy by about 500%. And you can see the numbers for these other things, a little less variation for hernias and appendixes. But for many other things, there was significant variation in how many patients underwent these procedures. Wenberg looked at things like hospital discharges in relatively sophisticated medical markets like Boston and New Haven and found that per thousand beneficiaries, the number of times people were hospitalized varied by over 50%.

Dr. John Lantos

And the days per 1,000 beneficiaries varied by about the same amount. So for some reason, reasons that nobody could quite explain, patients in Boston were being hospitalized at much higher rates than patients in New Haven. There's more data on tonsillectomies just showing how they quantify this in these three counties in a tiny area of Vermont and New Hampshire. Some patients got tonsillectomies at a rate four times as high as in the county next door. It was true for chest X-rays. It was true for many other things.

Dr. John Lantos

So here's the question that comes up then in thinking about the risks of comparative effectiveness research. If we know that this sort of practice variation out there, that it exists out there, what are the risks to patients who are not in a research study and who are being treated with this known practice variation versus patients who would be enrolled in a study, say, if we imagine children in Vermont and New Hampshire, a study that randomized them to a more aggressive or a less aggressive approach to tonsillectomy, that is, Is it riskier to have your treatment assigned by randomization or to have your treatment chosen because you happen to live in Littleton and not in Burlington? Or what if we randomized the citizens of Boston and New Haven to different algorithms to decide on hospitalization? Would they face more risk than they currently face under whatever guidelines are leading to the 50% variation? in hospitalization. So it seems a key question here is which is riskier, undisclosed and unstudied idiosyncratic practice variation or deliberate formal randomization with careful monitoring and evaluation?

Dr. John Lantos

With that in mind, let's look at what the federal regulations say about minimal risk and how it should be assessed and then I'll turn to the current debate which is as I'll show a debate between what the common rule sees as the appropriate way to think about this versus the current leaders of OHRP.

Dr. John Lantos

First thing to note is that the determination of minimal risk matters in terms of what is permissible or what is not. If an IRB determines that a clinical trial entails only minimal risk, it may allow a waiver or alteration of the informed consent process, It may permit the study to be performed in certain vulnerable populations, for example, children. It may use an expedited review process to review the protocol, which may not be permissible if the study has more than minimal risk. But the definition of minimal risk, which I'll put up in a minute, is pretty non-specific, and in part because of that non-specificity, IRBs clearly vary in the way they evaluate identical protocols in determining whether they are minimal risk or greater than minimal risk. To the extent that they are risk averse and call things greater than minimal risk, which properly by the definition should be called minimal risk, it can restrict valuable research. And this is a particular problem with multicenter research of the type that most comparative effectiveness research studies would almost necessarily be. Here's the federal definition of minimal risk. Something should be considered minimal risk if the probability and the magnitude or discomfort anticipated in the research are not greater in and of themselves than those ordinarily encountered in daily life or during the performance of routine physical or psychological examinations or tests. commonsensical approach. It sort of takes as the baseline daily life and says whatever daily life is, whatever the risks of daily life are, those are the risks that we would permit

Dr. John Lantos

children or other vulnerable populations to undergo or maybe permit adults to undergo without consent. There are two major problems and many minor problems with this definition, however. Daily life is seen as one standard, but we all know different people's daily lives involve different sort of risks. So it's unclear whether the proper comparator in evaluating a study is the risk of each individual person's daily life, which might mean, for example, that somebody who lives in a neighborhood with a high homicide rate should be allowed to participate in riskier studies than somebody who lives in a safer neighborhood or somebody who has leukemia and is getting bone marrow transplants should be allowed to participate in riskier research than a healthy children,

Richard Platt

or

Dr. John Lantos

whether it should be something like an average person's daily life, whatever that might mean. And Dave Wendler, who was one of the co-authors on the paper we wrote as part of the specialist issue of clinical trials, has written a lot about the ambiguities of clinical life. In the context of clinical research, the routine physical and psychological examinations or tests that a potential research participant might ordinarily encounter are quite different from those that a healthy person might encounter. And so using one scale, the scale of the risks of daily life, squishes a lot of things into a measurement that is not really applicable to the wide variety of those things.

Dr. John Lantos

Minimal risk is used primarily in studies, I should have said without the intention of providing benefit to study participants, or what used to be called non-therapeutic studies. And those are things like observational studies, phase one studies, pharmacokinetic studies, studies that is in which the goal is simply to answer a scientific question rather than to provide any benefit to the study participants. According to federal regs, if the studies do have a goal of providing some benefit, such as most clinical trials, especially prospective randomized trials, then the minimal risk standard is not by itself the metric that's supposed to be used. Instead, IRBs are supposed to use a balancing task of whether the potential risks are outweighed by the potential benefits.

Dr. John Lantos

If there's no possibility of benefit, according to the federal regs, then IRBs have to make an absolute assessment of risk, and most people are probably familiar with these three gradations where a study can be determined to be minimal risk, which leads to some degree of oversight, a minor increase over minimal risk, which leads to a higher degree of restrictiveness. And occasionally there are studies that have a greater than a minor increase over minimal risk with no possibility of benefit that are permissible even in vulnerable populations such as children, but only if they are reviewed on a case-by-case basis by the federal government. If there's a possibility of benefit, as I noticed, then the proper... balancing act is not or the proper assessment is not whether the risk is minimal but whether the benefits and harms balance each other out. The current areas of controversy then are in those sorts of studies are research versus quality improvement or what I'd like to focus on next, the question of which risks ought properly to be attributed to the clinical trial itself as opposed to being thought of as the risks of the treatments that are being studied.

Dr. John Lantos

So the debate about attributable risk. Can a randomized controlled trial ever be classified as minimal risk? And just as a disclaimer, emergency research is sometimes permitted without consent, but It has a special category, that is, it's not allowed because it's minimal risk, but because it's deemed to be very important and infeasible or impossible to do without consent.

Dr. John Lantos

Randomized trials, though, generally are not judged to be of minimal risk. And here an interesting question comes up that feeds back to my questions about the relative risks of randomization. versus idiosyncratic practice variation. To the extent that randomization is viewed as more risky, it is because of the way the treatment is assigned. And people who view randomization as riskier than other forms of treatment assignment generally imagine that the alternative is a careful, individualized decision by the patient's doctor compared to treatment that's delivered by a standard protocol.

Dr. John Lantos

And then if individualized care choice of treatment or changes treatment would result in a better outcome then participation can be thought to increase risk. Although it's hard to know whether that's true unless we actually studied it because it seems at least theoretically possible if you think of those oxygen studies or studies of tonsillectomies in Vermont or of hospitalization in Boston or New Haven, that one group in the study would likely have higher risk, the other group would have lower risk, but how you would know which group was which and therefore which patients were more likely to benefit or be harmed would be impossible to know unless you did a study. So should the risks of treatment being studied be considered risks of a study? Or should they be considered

Dr. John Lantos

risks of the treatments that a patient would be exposed to whether they were in the study or not? The common rule says the risks of the treatments being studied are not a risk of the study. The recent draft guidance from the Office for Human Research Protections says that they should be. And let me just go through those quotes for you. The common rule says clearly, and this is a quote, in evaluating risks and benefits, the IRB should consider only those risks and benefits that may result from the research as distinguished from the risks and benefits of therapies subjects would receive, even if not participating in the research. The OHRP seems to disagree with that. in their recent draft guidance, they say, if a person in a research study is being asked to undergo procedures that involve reasonably foreseeable risks that they would not have otherwise been exposed to, then that person needs to be told about the risks. And the reasonably foreseeable risks of research include already identified risks of the standards of care being evaluated as a purpose of the research. So to try to give a couple of examples to illustrate how these two guidances might lead to different determinations of risk, let me imagine two prospective randomized trials, one of antibiotics for otitis media and one of stents for asymptomatic coronary artery disease.

Dr. John Lantos

So antibiotics for otitis media. Imagine, for example, that there are two FDA-approved antibiotics, antibiotic A and antibiotic B. Both are in widespread use. Some doctors prefer A, some doctors prefer B, and people have actually studied this and determined that there is practice variation among doctors so that some patients are more likely to get A and some are more likely to get B. Both antibiotics A and B have side effects that are well characterized. They both cause a little diarrhea. Babies, giving them antibiotics, can cause a diaper rash, and both, in very rare cases, can lead to anaphylactic reactions. So they are not without risks. So the question that OHRP and the Common Rule seem to disagree about is, If you're going to do a randomized controlled trial of A versus B, are the known problems with these drugs, the GI problems, the diaper rashes, and even the possibility of anaphylaxis a risk of being in the study? Or are those problems a risk which patients would face whether they were in the study or not? And the risk of the study would then be simply the risk of being assigned to one or the other of these antibiotics at random. rather than being assigned to one or the other based on the doctor or the patient's preferences.

Dr. John Lantos

That, it seems to me, is very much up in the air according to current regulations. To take another example, some cardiologists recommend a stent for patients with asymptomatic coronary artery disease. Others think that stents increase risk. and recommend medical pharmacological management but no invasive procedures. Somebody wants to do a study comparing stents with optimum medical management. Are the risks of getting a stent in those circumstances properly considered as risks of the study? Are the risks of getting a stent risks that most patients would be facing whether they were part of the study or not? Here's the thing about current research regulation.

Richard Platt

The current

Dr. John Lantos

system seems to focus on measurable risks, but as my oxygen studies examples show, that doesn't really seem to be the central concern. That is, if you describe studies in which

Dr. John Lantos

50% of patients would get one treatment and 50% of patients got the other, but they get it either by idiosyncratic practice variation or by a change in hospital policy, the risks seem to be exactly the same, but the regulation treats them very differently. The concern then seems to be not so much about some absolute measure of risk, whether it's minimal slightly greater than minimal or a minor increase over slightly greater than minimal. Instead, if you look at the way people think about the risk of prospective randomized controlled trials, particularly of comparative effectiveness studies where it seems that the risk is very comparable to the risks that people would face outside the study, the fear or the concern seems to be about something much deeper and more intangible than about measurable risks. So in the last part of this talk, I'm just going to speculate about what I think is really going on in current debates about the riskiness of research, which, as I suggested, is not, I think, about measurable risk. It is about the very nature of this thing that we call research and the people who do it. And that comes out in

Dr. John Lantos

OHRP's response to the support study in which after going through the debates about whether the patients in the study were actually exposed to more risk or

Richard Platt

not,

Dr. John Lantos

and reasonable people can come down on both sides of that question, OHRP said ultimately the issues, they don't say don't come down to risk, but they say come down to a fundamental difference between the obligations of clinicians and those of researchers. Doctors are required, even in the face of uncertainty, to do what they view as being best for their individual patients. Researchers do not have that same obligation. And this way of formulating what's at stake in oversight of clinical research, once you notice it, pops up over and over again anywhere you look. Ruth Macklin and Lois Shepherd use similar words. It's the doctors, not the researchers, who have a fiduciary obligation and longstanding ethic to pursue the patient's best interests above all other considerations. Or George Annis, a physician must be guided by a fiduciary obligation to the patient. A researcher has no such obligation. The real fear then is not the risks of the treatments themselves, but instead, I believe, It is the dark and conflicted heart of the medical researcher. The question that seems to motivate much of research regulation today is whether these people staring at their computer screens are angels trying to improve the quality of care or devils who would happily exploit their unsuspecting patients in order to pursue the goal of creating generalizable knowledge. And this is crucial because actual studies can be safer or riskier than conventional therapy or in many cases we don't know, that's why we're doing this study. But if the problem is the loyalty of the researcher to the patient, then any activity that is seen as research or designated as research leaves people unprotected by their fiduciarily bound physician and instead at the mercy of the whims of the thoroughly utilitarian researcher who would happily use them and harm them in pursuit of generalizable knowledge. Interestingly, there's no such concern if you call something quality improvement because that doesn't lead to the same fear as

Dr. John Lantos

fears about the researcher. Researchers are thought to ignore patient welfare. Steve Chaffee and Frank Miller say in the context of medical care, beneficence entails the health provider do what's best for patients. In clinical research, beneficence directs the investigators to promote social value by generating knowledge. Hellman and Hellman researchers are required to modify their ethical commitments, whereas Larry Churchill said the researcher The research subject relationship by its very nature compels or urges certain priorities and inclinations to perceive and act in certain ways. The researcher is seen as driven to pursue knowledge, committed to a utilitarian ethic, thus amoral and in need of constant oversight. They're addicts in their pursuit of knowledge. They can't control their moral impulses. Is this true? Many researchers completely reject this formulation of what they're about when they are doing studies and see themselves as exquisitely attentive to patients' interests and in fact as

Dr. John Lantos

honest questioners who are suspicious of the value of the clinician's individualized clinical judgment, that is, the individualized clinical judgment that is thought to be the manifestation of the clinician's fiduciary obligations to do what's best for their patient, when in fact, as many researchers point out, there's no good evidence to inform that individualized clinical judgment. So what you get is a sort of virtue as an empty shell, virtue that has no evidence upon which to base these supposedly virtuous decisions. Here are a couple of examples of researchers who have tried to formulate that view. Norm Faust, in talking about waiving consent for emergency research, says that what he'd like to say to his patients is it would not be responsible to give you an unstudied treatment in an uncontrolled way because neither you nor I nor future patients would ever know whether it helped or hurt. That is his virtuous physician, Fiduciary obligation does no good unless we know whether the treatments themselves are beneficial or harmful. Or in a similar way, Keith Barrington, a neonatologist, says, I have a fiduciary obligation to provide optimum treatment. I also have a moral obligation to know what the optimal treatment is. And I have a moral obligation to keep trying to find out what the best treatments might be.

Dr. John Lantos

Even as far back as Jay Katz, one of the pioneers of research ethics and informed consent, sees research and therapy as not separate but intertwined because the multiple purposes of medical practice include caring for patients, advancing science, improving the health of communities, natures, and nations and future generations, and cannot be clearly separated. So where do we go from here given what we know about practice variation, what we know about the difficulties of defining minimal risk, about what we know about the need for comparative effectiveness research, and what we know about some of the ambiguities in deciding whether something ought to be considered research quality improvement

Dr. John Lantos

or something else. I think there is a crisis in research regulation today that shows up in the way that people are thinking and writing about it. There was a whole issue of the American Journal of Bioethics devoted to debates about minimal risk. The Hastings Center Report ran a special issue in January of 2015, I believe, on the learning health care system, which I'll come back to in a minute. PCORI has a special journal issue on the ethics and regulation of comparative effectiveness research and pragmatic trials. And there's this clash of principles between ethics and QI. I think all these rumblings suggest the need for a new framework to at least update the current Bell-Mont framework, which after all was developed in the 60s and 70s looking backwards towards Tuskegee and the Nazi experiments. The current problems we face are different and need a different sort of analysis. One of the best of such recent analyses was written by Ruth Faden, Nancy Cass, and some colleagues

Dr. John Lantos

in the Hastings Center report where they talked about the ethics of a learning healthcare system. And one of the things they said was that just healthcare requires a constantly updated body of evidence about the effectiveness and value of interventions and of alternative ways to deliver and finance healthcare. And they outlined seven ethical obligations in learning healthcare systems. Four of them look very familiar from current clinical ethics and research regulation. But numbers five, six, and seven depart from traditional research ethics. And they say in addition to, for example, respecting the rights and dignity of patients and providing optimal care to each patient, we also have an ethical obligation to reduce health inequalities among populations, an ethical obligation to conduct responsible activities that foster learning from clinical care and clinical information. and an obligation to contribute to the common purpose of improving the quality and value of clinical care and healthcare systems. They conclude that traditional presumptions need to change, that health professionals and organizations have an obligation to learn, and this is probably their most radical claim, that patients have an obligation to contribute to, participate in, and otherwise facilitate learning, they make clear that what they have in mind is a group of activities that would be minimal or arguably no risk for patients, which feeds back to debates about how to make that assessment of what sorts of risks ought to be considered so minimal that people have an obligation to participate in such research. A learning healthcare system in this view would link the obligation to respect the rights and dignity of patients with the obligation to contribute and improve the quality of care.

Dr. John Lantos

So the

Richard Platt

new framework

Dr. John Lantos

would see an obligation to improve quality. It would see justice concerns with opting out. It would acknowledge the riskiness of idiosyncratic practice variation as the baseline against which to measure the risk of a new research paradigm that doesn't then starkly dichotomize research and clinical practice, but sees the two linked in much the same way as Jay Katz did almost 50 years ago. One conclusion from this is that the central dogma of research regulation today, the dogma that clinical investigators cannot be trusted to make moral judgments about their own research, needs to be at least seriously questioned if not overturned and thrown out, and instead to imagine a dogma in which clinical investigation is seen as both an obligation for

Dr. John Lantos

doctors, virtuous doctors, and an obligation for patients who desire the best care for themselves as well as the best health systems for society. The change we need is to acknowledge that research can often benefit current patients, not just future ones, if only by protecting them from the harms of unstudied practice, and that clinical researchers are no more conflicted than practicing physicians. Both should be expected to balance their moral obligations to their patients with other conflicting obligations.

Richard Platt

I think I will stop

Dr. John Lantos

there since it looks like at least a couple of questions are starting to come in and open us up for discussion.

NIH Collaboratory Grand Rounds moderator

Hey, thanks, John. So that's great. So why don't we go ahead and open the line to Rich. He has some experience in some of these issues with some of the cluster trials and always a challenge.

Richard Platt

Yeah, so John, that was terrific. Could you tell us how you think IRB should think about the oversight of cluster randomized trials? I'm talking particularly about the kinds of trials that involve the higher level groupings of, that involve practices that are typically applied to groups of patients. So the couple of examples I gave you are setting staffing ratios in a hospital unit or deciding which soap an ICU uses to bathe its patients. These are the kinds of questions that are essentially never discussed by clinicians and their patients and that individual clinicians really have no control over. As I read the Notice of Proposed Rulemaking for the Common Rule, it basically ignored that. that area, and it seems to be a big gap.

Dr. John Lantos

Yeah, I absolutely agree it's a big gap. I remember when they had the hearings about the support trial, somebody asked Jerry Manikoff a question about that, and he talked it and said, well, that's a whole different topic. We haven't gotten to that. I mean, to my mind, they would fit into the same sort of framework that I'm proposing here. That is,

Dr. John Lantos

if we compare the risks of whatever intervention is being tested by a cluster randomized trial to the risk of whatever standard practice might be and judge that participation in the cluster randomized trial is, we don't know whether it's gonna increase risk, decrease risk, or have no effect compared to the standard practice. And it seems those should be judged as minimal risk and therefore as something that wouldn't necessarily require individual patient

Richard Platt

consent. It's not,

Dr. John Lantos

at least clear and it may even be clear that the opposite is true in terms of how an IRB would think about it. That is if the framework is based on this notion that any attempt to develop generalizable knowledge makes something research and once something becomes research it is by definition riskier than whatever exists outside of research. then you'd end up in a situation where the only reason a cluster randomized trial could ever be approved would be if consent was deemed to be impracticable or impossible or unfeasible.

Dr. John Lantos

That seems like a bad way to go to me, but I think that's where current regulation would be.

Richard Platt

I don't want to hog all the discussion time, Do you have an opinion about substituted consent that is having designating certain individuals to give consent for a cluster? Because it's almost never possible to get consent of the individuals in a cluster. A hospital will have a certain number of nurses on a unit per shift or it won't. I mean, I think,

Dr. John Lantos

you know, imagining that whatever a designated individual would do, is what we think of as informed consent is, I mean, I think it's not informed consent. Whether it ought to substitute for informed consent and therefore allow such a project to go forward,

Dr. John Lantos

I guess it would depend who you would imagine being that person. I mean, if we all lived in communities where we deferred to elders, then perhaps the elders would be the ones. But in America today, it seems like that's as symbolic as the idea of community notification for emergency research, which, again, seems like a nice idea, but probably not actually fulfilling the function that informed consent is thought to fulfill.

Richard Platt

Okay, my parting thought is if the people who normally make a decision about stamping ratios are hospital VPs, maybe disinterested hospital VPs would be the ones who would make the consent decision. Okay. I'm handing the microphone back. Thank you.

NIH Collaboratory Grand Rounds moderator

Well, and I just, why don't we go ahead and open the line to

NIH Collaboratory Grand Rounds moderator

Carolyn Miner, and then just a note, in the chat, Jeremy Sugarman did note that there is some guidance regarding the oversight of clutch randomized trials, and put the link there so we can also look at that. But Carolyn, you have an interesting question here.

Audience member 1

Well, no, I was just wondering, is it truly a dogma Or is it just a reflection of public perception? I mean, in popular movies, having the mad scientist as the villain is not uncommon. So I don't think it's just as easy as changing the perceptions of the regulators.

Dr. John Lantos

I think the movies are a little more complicated than that, and public perception. That is certainly one. one portrayal, that the mad scientist is out to destroy us all. But the other is that scientists are too cautious. And here I'm thinking of things like Lorenzo's Oil or Dallas Buyers Club for

Dr. John Lantos

public reaction against overly cautious science, right to try laws, where people say,

Dr. John Lantos

Let me decide which risks I want to take.

Dr. John Lantos

Matthew McConaughey's best line, I think, in Dallas Buyers Club was, screw the FDA. I'm going to be DOA. I think the public wants it both ways. Everybody wants medical progress. Nobody wants to be a guinea pig. Everybody wants responsible research. unless they themselves are in need of some innovative treatment. And then they don't want oversight or anybody telling them what to do. But in that sense, you're right. It's probably dogma is not the right word. And maybe a better way to frame that would be that there is a real tension in public perceptions and desires about how they want science to be. regulated and overseen.

NIH Collaboratory Grand Rounds moderator

Thanks. Let me open the line to Jeremy and get a few comments from him as well in here. But one thing that's really struck me is that the variation and interpretation of minimal risk is quite striking. And so, you know, some have advocated, well, if we only had, you know, one single IRB, maybe that would address the issue. But on the other hand, there is this concern about who does interpret things as minimal risk. So maybe in the last few minutes, if you could comment, and then Jeremy, if you have other comments or what are the final answers here?

Jeremy Sugarman

Do you want John to respond to you first or have me chime in?

NIH Collaboratory Grand Rounds moderator

Once you chime in here, Jeremy, then we'll get John. You can have the last few words.

Jeremy Sugarman

Okay, great. So, John, great presentation and a really nice review of the issues. And this whole topic, it seems so familiar that every few years in research ethics we've been talking about the same thing, but we haven't

Richard Platt

really cracked it.

Jeremy Sugarman

And I admire the idea about moving forward and trying to rethink. what the right ethical framework might be so that we could then hopefully have the regs follow. My one concern about using the Faden and Cass approach that they described in the New England Journal of Medicine is that their whole system is predicated on a fully enabled learning health system. And we're not there. And the three areas, the three sort of principles or guidelines that you pointed out, don't really have any moral weight in a system in which there's not the reciprocity associated with a learning health system. So I'm wondering, before we achieve a learning health system, if we ever do, what's the right ethical framework that we need to apply that recognizes that there are limitations on things like justice and reciprocity? Because I think that's the way we might want to be thinking going forward.

Dr. John Lantos

Great question.

Dr. John Lantos

Whether

Richard Platt

a

Dr. John Lantos

learning healthcare system could be fully achieved before it was tried out I think is a sub-question there. That is, I think it may be the only way to get to a learning health system is to start trying to become one.

Dr. John Lantos

Or I guess you could say it's acceptable in Norway but not here or something.

Dr. John Lantos

I mean, I think

Dr. John Lantos

it could be done on a hospital-by-hospital basis if at the time of admission you were told, we are a learning healthcare system. That means we, you know, if you get your care here, we use the data that we generate as part of your routine care to try to improve the quality of care for you and other patients. You could imagine that being done as an opt-in, as an opt-out, or as a condition of treatment.

Dr. John Lantos

My understanding is the Mayo Clinic has some sort of statement like that, that all patients who come there sign, and if they don't want to sign it, then they're sent into St. Paul, I guess. I don't know what happens. But your main point that

Dr. John Lantos

the Faden group is suggesting a modification of the regs with a promise that it will have a payoff in improved quality and justice is only as good as the promise is good.

Jeremy Sugarman

Yeah. And also, it's a future, it's an aspirational system. And so I don't think we can just take it wholesale and use it either because it's not designed to fit with the current medical world and research world as we know it. But second, because we don't know empirically whether some of the assumptions are valid. So I agree with you that the regs have been problematic and the guidance documents are sometimes conflicting and hard to follow and that there's diversity on views. But I think what we may need to do is reinterpret sort of Belmont and the approach and come up with how we specify the principles we use for both research and practice in ways that accommodate the changing world in which we're trying to get things right so that we can best assure that we do really needed research and not undermine the rights and welfare of patients and the general public.

Dr. John Lantos

Yeah. I mean, I think

Dr. John Lantos

Ruth Baden and Tom Beecham, who wrote that, would see what they're doing as a reinterpretation of Belmont rather than overthrowing it. The question would be how much reinterpretation is permissible without destroying the core of respect for persons. One quick question that somebody wrote in just to clarify, a QI project needs IRB approval if it results in a publication, correct? After the debate about the Keystone project, which was a QI project that was published in the New England Journal, OHRP said no. That is not the criterion. QI projects can result in publications and still not be considered research. And if you want to find out how they then go on to split those hairs, you could look up their statement in response to that project. But it's not as simple as publication means research.

NIH Collaboratory Grand Rounds moderator

John and everyone else, I want to thank you all for joining today's Collaboratory Grand Rounds. As Jeremy noted, this is a topic that continues to come up. I imagine it's actually not going to just cycle up and down, that this will continue to escalate as people drive towards learning healthcare systems and what are the challenges to address this issue. So, John, thanks again. And I want to remind everyone, next week we have Grand Rounds to discuss challenges and opportunities around common PRO measures and ER. So look forward to everyone next week. Thanks, everyone.