Episode 10: What is Human Subjects Research in Law?

[00:00:00] Speaker A: Imagine a justice system built on rigorous evidence, not gut instincts or educated guesses about what works and what doesn't. More people could access the civil justice they deserve. The criminal justice system could be smaller, more effective and more humane. The Access to Justice Lab here at Harvard Law School is producing that needed evidence. And this podcast is about the challenge of transforming law into an evidence based field. I'm your host, Jim Griner, and this is proof over precedent. [00:00:35] Speaker B: Okay, so we're back with Shannon Seward. Thank you so much, Shannon, once again for your patience and your perseverance and pursuing the podcast. We are talking some more about IRBs and the common rule and ethical review of the technical term is human subjects research. I always found that a little ghoulish. I like human participant research. [00:00:56] Speaker C: I agree. [00:00:57] Speaker B: I just think it's a more friendly term because that suggests that there's a role for the folks who are involved in the research and we should treat them as participants. So anyway, the technical term is human subjects research. And especially going to begin to talk today about how the nuts and bolts of how IRBs work and try to demystify the process for folks who have not engaged in IRB governed or IRB regulated research before. Try to talk about what some of the common rule terms are, the most important common rule terms are, and what those mean. And then we'll see if there is time to try to try to tackle some of the problems that these terms and that the process poses for research in the law or and more generally, research that is outside the medical context. So in our previous podcast, Shannon, you clarified, it's pretty clear if anybody looks at the regulations, these regulations in the IRB process, process are constructed with one type of study in mind, which is a traditional medical clinical trial. [00:02:04] Speaker C: That's correct, yeah. [00:02:06] Speaker B: And so the idea is we, there's no trouble there, for example, of figuring out who the participants are. They are the people who are going to be taking the drug or the old drug or the new drug or whatever it is. We know we're going to need consent from them because in fact, you have to get consent to undergo the treatment anyway, much less research into the treatment. And that's a key principle is we don't allow someone else to administer something to someone's body without the recipient's consent. And so we would need consent anyway. And there's no trouble figuring out whether we're going to need consent. None of these easy, none of these calls are easy in the law. And so this is what we're hoping to get Into. Okay, terrific. Super. So let's just go through real nuts and bolts here. There's an irb. Suppose you're at an institution and you're contemplating human subjects research. Or actually, let's start there, what might be human subjects research. And we need to figure out, first of all whether it is human subjects research and therefore whether you need to even apply for IRB clearance. What thought process should I go through in order to figure out whether something actually is human subjects research? [00:03:23] Speaker C: Well, I think the actual process is very prescriptive, although you will see different interpretations depending on which IRB that you're working with. And that was really, I think, the impetus behind getting IRBs accredited so that it's more an even playing field regardless of what IRB you're going to. But there are basically two buckets that the IRBs look at. And please let me know if I'm not defining something, because I, I will. I speak irb, Jim. And sometimes that doesn't translate to English too well. So basically there's two buckets. And we deal with determinations and approvals and determinations. Are those things such as, is it research or is it not research? Is it research involving human subjects or is it not? And if it does involve research with human subjects, then you go up the review ladder and we can get more into that. But insofar as determining, you know, whether you're dealing with something that is research and whether something that involves human subjects, there is a very specific definition in the regulations. First, before you even get there, you have to really conceptualize what type of research am I doing? Because depending on that type of research, you're either dealing with one set of regulations or another set of regulations, specifically HHS common rule or FDA regulations. And those two regulations differ greatly at that level of what's research versus what's a clinical investigation as well as what's a human subject, they're wildly different. So if you know that you're not looking at a device, drug or biologic to study safety or effectiveness, then we'll focus in on the common rule. And the common rule has a very specific definition of research. You know, systematic investigation. Is it going to be designed to contribute to generalizable knowledge? Right. So clear, so specific. [00:05:42] Speaker B: Actually, let's go through that again then. So let's assume we're dealing with a common rule. Right. Which is going to be the thing that's going to be most applicable to the audiences for proof over precedent. Because I don't think we manufacture too many medical devices in the Law or those sorts of things. But let's go through that one more time. Could you say that, Just say that one more time. What are the key pieces that all these boxes have to be checked before you need to go to IRBs to begin with? What are those key pieces again? [00:06:14] Speaker C: Systematic investigation designed to contribute to generalizable knowledge. [00:06:21] Speaker B: Got it. Okay, so let's talk. What are some examples of things outside medicine say, that are, that are in or out for each of those pieces? [00:06:32] Speaker C: It's actually much easier with the biomedical realm. [00:06:36] Speaker B: Sure. All right, well, we'll start there. We can start there. Go ahead. [00:06:40] Speaker C: I think that really things that are not research things that it really lends itself to that generalizable knowledge piece. Right. Are you looking to improve how you teach? Right. So you're just evaluating your own teaching methods. It's not going to be generalized to other situations. It's just basically to improve your performance. So is it encapsulated within the environment of which the activity is taking place? [00:07:12] Speaker B: So let me give you an example there. Suppose I Happen to teach 165 civil procedure students per year. Some people say I inflict 165% students per year with my teaching. And if I want to improve my teaching and improve their performance, I can randomize, say over three years, I can randomize half of them to one form of study guide and the other half to say, a different form of study guide, and then watch them and compare, I don't know, their grades at the end of each of the three years and I can find out that Study Guide A is superior to Study Guide B. And then from now on I'm going to assign Study Guide A in my civil procedure teaching. Am I doing human subjects research according. [00:08:08] Speaker C: To a strict interpretation of the regulations, as long as it's contained within how you're improving your own specific teaching, that would not be considered research subject to IRB review. [00:08:20] Speaker B: Now then, if I, from the beginning, if I had planned to publish the results of that of Study Guide A versus Study Guide B, a study of Study Guide A versus Study Guide B in a journal that, say the journal you know, of higher education or the, or a journal of law teaching or something like that, then that would be human subjects research. Is that correct? [00:08:44] Speaker C: Not necessarily. [00:08:45] Speaker B: Okay, tell me how, how would, what would go into that determination? [00:08:48] Speaker C: This is where it gets a little bit gray. So if you look at the definition itself, there are many instances where it could be a very subjective interpretation. Right. Within institutions, what they've done is they've added these additional ways to interpret that. So a lot of institutions will say, oh, are you publishing? Oh, you need IRB review. When publication in and of itself is not necessarily a criterion to make whether you need IRB review or not. Some institutions will say, oh, you're randomizing. Oh, randomized randomization requires IRB oversight. Or are you conducting an activity that's especially risky? Oh, you need to go to the irb. But none of those, I guess, you know, none of those criteria are found in that definition, but institutions have adopted them to make making that decision a little bit easier. So I've tried not to do that. [00:09:56] Speaker B: Bless you for doing that, by the way. [00:09:58] Speaker C: And really look at what do the regulations say that requires IRB oversight? For example, you might do a quality improvement activity that might introduce some form of risk. However, it may not contribute to generalizable knowledge whatsoever. It's just part of understanding whether, you know, A is better than B for the institution's interest. That wouldn't mean that IRB review is required. There may be some other mechanism at the institution that oversees this type of activity, but that doesn't necessarily mean that the IRB has to be involved. [00:10:36] Speaker B: Well, actually. Go ahead, go ahead. Sorry. No, I'm just going to say, what, what do you think then? If you were, if you were in charge of the universe, what would you say? And applying the regulations, what would you say to my. My hypothetical here of a study guide A versus study guide B, where I am thinking that I'll publish the results at the end of the study because I think it will be of interest to civil procedure professors and law professors more generally? [00:11:06] Speaker C: I would say, you know, I'd probably ask you a few more questions about how you're going about it. Do you have. Are you working off any previous, you know, work that was done, you know, what are you doing or work cited? How are you conceptualizing how this might be generalizable? You know, if you think about case studies, right, you're just describing a person's condition or you're describing a situation. It's informative, yes, but is it contributing to generalizable knowledge? And I think this has been a huge argument among social behavioral researchers where, you know, tell me why this requires IRB oversight, especially with ethnographers, right, because they're describing a culture or, you know, situation, etc. It's not necessarily following scientific method per se. And so this has been a real challenge for social, behavioral IRBs is, well, what criteria are you using to determine whether something needs IRB oversight or not? [00:12:24] Speaker B: And Let me just point out here, this is obviously not your problem, but a problem that researchers and IRBs generally face is how easy it was for us to run into interpretive difficulty. I mean, we just barely got started here, right? And we're already in trouble in terms of figuring out what is and what is it. Right? Because again, I think your first reaction was instructive is the same as mine, is that if I have no publication plans and all I'm trying to do is to, is to run and run an RCT to figure out how I can improve my teaching and how, and how I can make my students learn better with the same amount of investment just because is one study guide superior to another one and I have no publication plans. It's an entirely self contained within my class. Your immediate reaction, same as mine, is no, that's not human subjects research. Now notice as we, as you pointed out, I'm randomizing. So if the trigger were randomizing, then I'm doing human subjects research. But that's not, as you said, that's not what the regulations say. And so both of us agree, no, that's not human subjects research. Entirely self contained. Soon as I introduce a concept as simple as publication, all of a sudden we're in trouble. And then I haven't even my next nuance, which we didn't even need to get to, to create trouble, was what if I, what if I plan just to keep it self contained, see the results and say, oh my word, those results are really startling now I should publish them because the rest of the world should know about this, but it's too late to go back and get IRB approval ahead of time. Right. [00:13:58] Speaker C: You know, and, and I think that's, that's an excellent point. And you know, a lot of times what I'll ask people is when they're like, you know, I, I want to publish this and I'm like well, are you going to publish it regardless of whether the outcome is favorable or not? [00:14:13] Speaker B: And for me the answer is always yes. Because I think null results are fundamental and underappreciated. Right. In science. Right. [00:14:21] Speaker C: And, and that's, that's a very important aspect when you're, when you're conducting research, right. Is you know, whether or not the results are favorable. It's, it's important to the community to know that. And so a lot of times when you're, you know, encountering an individual that's doing an activity that might just be an evaluation of teaching methods or what have you, that's not necessarily Generalizable. I'm going to publish it. [00:14:46] Speaker B: Right? [00:14:46] Speaker C: Well, okay, great. Is that only if you find great results, right? Or not. And so people will walk back. Sometimes they're like, oh yeah, only if I find great results. So in, in those cases, what they might be doing is an activity that isn't considered regulated research. However, if they do find, you know, results that are favorable, they may want to expand on it. And at that point in time that may transition over into something where it is. The intent is to develop something that is generalizable. [00:15:25] Speaker B: Well, if, and, and, and again, I don't want to get too far into the, into the specifics of this hypothetical. Although again, our. I hope we will be able to get into some specifics going forward because this, the difficulties of the regulations. Is there a provision within the regulatory structure and IRB custom and practice, if activity has already been completed and all of a sudden a scholar says, goodness, I should publish that. But the activity was originally the scholar didn't go get IRB review to begin with because the scholar said, I don't think this is human subjects research. I don't think I should have to do it. So my hypothetical again is I do the ab. The, you know, the, the AB testing for the, for the study guide. I find out that there's an astronomical difference between, you know, A versus B. Something's going on. There's magic somewhere in one of these study guides. I don't even know what it is. I just know all of a sudden one of them works really well. What I need to do is to get the result out in the world and have somebody else figure out what the magic is. Some behavioralist, some ling. Some linguist or somebody else. I need to publish this because this is a big deal. But I haven't gotten IRB clearance to begin with. When I did it, can I publish? [00:16:41] Speaker C: That happens frequently, actually, where you look at, you know, what was that activity? Did you design that activity? Was it a systematic investigation? Designed. Right. So that's that key word, designed. [00:16:55] Speaker B: Right. [00:16:56] Speaker C: Was it your intent to design this to contribute to generalizable knowledge? Not just like, oh, I found a cool aspect, I want to share it with people. That still doesn't change what the initial intent of that activity was. And I know you'll have other IRB scholars that will like, intent is not. [00:17:11] Speaker B: In the definition, but designed seems to correct. Yeah. [00:17:18] Speaker C: Yeah. [00:17:18] Speaker B: Okay, go on. Sorry. [00:17:21] Speaker C: No, that's. And there's a lot of people that will, you know, you can get into philosophical debates with, with many people in the IRB world, especially at their national conference about this very topic. So if, you know, randomization doesn't mean that it's generalizable knowledge or risk isn't part of it, then what is, what isn't? So you'll find a lot of IRBs do hang on to that as far as, you know, randomization, risk and other criteria to give them sort of an easier way to make the determination between what's not research and what is research. [00:18:01] Speaker B: Got it. And again, the odd thing from an outsider might be one I. One thing this way, one thing might be that's odd to an outsider is from the experience of. Again, what I like to say is the participant, the subject, and the language of the regulations, none of this matters. The same thing happens to them or with them either way, in some instances, and in some sense it might be IRB regulated. In other, other situations, it might not. Even though they are experience, they're having the same experience. Is that accurate? [00:18:35] Speaker C: Well, that would be ideal. Right, but sometimes you say it might. [00:18:40] Speaker B: Be ideal in the sense that. In what sense, sorry, that the participant. [00:18:44] Speaker C: Would have the same experience. [00:18:46] Speaker B: Right. [00:18:46] Speaker C: There might be other situations in which they don't. For example, someone may not, you know, think that they should tell people why they're doing this activity or what their participation entails or that, you know, maybe they're deceiving someone and there's no, you know, disclosure of that at the end of the activity, et cetera. So there may be variability in the participant experience depending on whether it's IRB overseen or not. [00:19:14] Speaker B: Got it. Got it. And by the way, this, this language that we're discussing here and already have gotten hung up on is 45 CFR 46.102 D. And so, and so that's for, for the law nerds among us that actually want to go figure this out and be able to try to interpret it in the future. That's. There's. There's your citation. Then we get to the next part, which is. Is going to be extraordinarily problematic in the law. And I'm going to bring up some hypotheticals to suggest that it is, it is quite difficult and intensely frustrating. I'll just find that this is one of the things I've found intensely frustrating is what is a human. What is a human subject? And the problem is not so much, luckily to define the term human as it is to figure out who is the subject of the research. And again, in medicine, that's typically screamingly obvious. It's the person into whom's body, the different interventions are going to go. And again, since one of the origins, and in fact, I think the entire origin of this, of this insistence that we have about consent comes from the insistence that we have about cons, about people's consenting to what happens to their physical selves, to their bodies. We have no trouble insisting that we're going to need to have consent. On the other hand, with legal studies, it is often quite difficult to figure out who is a subject of the research, who counts as a subject. So to begin with, let's go through this definition. This is now at 45 CFR 46.102 F. And I've got that. So. But in. And why don't you, why don't you just summarize the definition for us, Shannon? Because it's a full couple, a full paragraph long. [00:21:09] Speaker C: It is. So they have, actually, with the revised requirements, they've doubled down a little bit on the definition of what a human subject is. And there's kind of two parts to this. One involves where you have an actual intervention interaction. Right. And the definition now includes biospecimens. But IRB professionals always know that you treat biospecimens and data the same way. But now they've included the definition. [00:21:42] Speaker B: Got it. [00:21:43] Speaker C: So you. [00:21:44] Speaker B: Luckily we don't have to deal with that very much. That's true, yes. At least we escape that. But anyway, go ahead. [00:21:50] Speaker C: So basically it's intervention or interaction with a living person. Okay. And you're getting something from that, whether that be material or that be data. Okay. And it's, it's about that individual, so it's their opinion, it's, you know, their thoughts, etc. Etc. [00:22:07] Speaker B: Okay. [00:22:08] Speaker C: The second part is really in regard to private identifiable information. Okay. Stuff from a medical record, stuff from an employment record, educational, etc. Something that someone is, would guard as private and not open to public disclosure. So there's really, there's really two pieces to it. And I think that a lot of times those two things get conflated in the definition. So for example, if you're, you know, conducting survey, the man on the street, you're, you know, stop. And you, you ask people questions, you know, how are you feeling today? What do you think about this, etc. Etc. Etc. But you're not collecting any identifiable information that's considered a human subject, that intervention interaction piece, Right? Or if you're, that is just to. [00:23:07] Speaker B: Clarify, that is a human subject, correct? Yes, exactly. Yes. Okay. [00:23:12] Speaker C: If you are, you know, getting information from a, some Sort of record that is, you know, is private. That too. Even if you're recording it in a non identified, excuse me, non identifiable way, the mere fact of seeing it counts it as a human subject. [00:23:33] Speaker B: So again, that is something that law, that law professors, legal scholars do frequently. So let's clarify that for a moment. It is still human subjects research if someone else de. Identifies private information and gives it to you. So the fact that you cannot re identify it, is that correct? [00:23:53] Speaker C: That's not correct. [00:23:54] Speaker B: Okay, so tell. So unpackage that for me. [00:23:57] Speaker C: So if you're using. A lot of times someone will use someone that's not part of the study team or maybe someone that is the holder of the records. If they go in the records and they extract information that would not be considered readily identifiable to you, the researcher, that would not be a human subject. [00:24:20] Speaker B: Got it. So then if I have this right, then you need two key pieces in order to say you're not in human subjects land. One is you need the data. Data to be retrospective. And the other is you need it to be not re identifiable. And if you have, if it's both retrospective and not re identifiable, then you are not. Then it is not human subjects research. Is that correct? [00:24:42] Speaker C: They changed that. [00:24:43] Speaker B: Okay. Oh dear. [00:24:45] Speaker C: So if you're getting, if you're just getting data that's not identifiable, regardless perspective, it's not considered a human subject because it's not. You don't know who the person is at all. However, you know the control function there is that that person that's giving you that information can't be like part of your study team. [00:25:09] Speaker B: Right. [00:25:10] Speaker C: Or considered a collaborator, where there could be some lack of firewall between identity and the data you're getting. [00:25:19] Speaker B: So you have to have a screen or a firewall in front of the research team. Not just you, the entire research team. [00:25:25] Speaker C: Exactly. For it to move into that not human subject area. So for example, if you're the person that's actually going into the record, see all this stuff, even if you de. Identify it when you record it, that would still be considered a human subject. [00:25:43] Speaker B: Got it. Okay. And so this brings up. That's sort of the data part again. We're barely even scratching the surface here. Already hitting these issues. Right. Okay. That, that deals that. That sort of addresses one set of things that, that legal scholars sometimes do and economists sometimes do, sociologists, which is they, they obtain information, say private information from government records or from private information aggregators. And then they match it and they try to match it across data sets and answer Interesting questions, right? That that sort of addresses one situation. Another situation is this word or another situation arises from this term private information. And the key term private here. I think we, at least as a researcher at the Access to Justice Lab, have run into trouble persuading people that certain types of information are not private. And the most common example is court records, which are presumptively in the United States, unlike in other countries, criminal records especially are presumptively public. You know, in the United States, this is. This is considered a key part of our accountability mechanism for the citizenry vis a vis the government is that the citizenry is allowed to see what the government is doing in the criminal justice system in a quite transparent way. Many people think that has negative consequences for the individuals who are criminal justice involved downstream. That's an area of research. But nevertheless, the information has to be private so that, for example, if it. If the information is available, readily available to the public upon request, it is not considered private. Is that correct? [00:27:32] Speaker C: Correct. That is. That is my interpretation. [00:27:35] Speaker B: When you say it's your interpretation, are you suggesting that others don't hold it? [00:27:39] Speaker C: I think people, some people may disagree. And it's. I think that there is a lot of area, especially within the research area where you live is that there is a lot of the general public, including IRB staff, IRB members, that are not aware of these aspects. You know, and I think also technology is really blurring the line between what's private and public. Not, not to deviate too much from. From court records. But I think it's been a huge challenge in the IRB world. You can find out a lot of stuff about someone on the, the Internet. Yes, it's readily available. It's. It's all out there. You know, an interesting situation that I was involved in was in regard to the Dark Web. Interesting. But is there, you know, you have to ask yourself, is there. Would one consider that private, even though that stolen data has been made public right over Facebook and the hordes of information that people share about themselves, you know, where does that, that line between private and public lie? [00:28:56] Speaker B: Do the regulations answer that question, where that line lies? [00:29:01] Speaker C: No, they don't. So it's basic. It's only what we have in the definition is, you know, basically what one can expect to be reasonably private. [00:29:15] Speaker B: Yeah, but I'm looking at the definition here. Private information includes information about behavior that occurs in a context in which an individual can reasonably expect that no observation or recording is taking place. And it includes information which is provided for a specific purpose. By an individual and which the individual can reasonably expect will not be made public. For example, medical records. So unlike, by the way, in certain legal contexts, there's a rule that basically says if you give the information to a third party, it's no longer considered private. That's an assertion seizure context in fourth amendment constitutional law. That definition definitely doesn't apply here. You can give the information to someone as long as you do so with the expectation that it will be with a particular purpose and with. And that it will be kept private. And then you have. And that's still considered private for IRB purposes. Is that correct? [00:30:13] Speaker C: Correct. That's correct. [00:30:15] Speaker B: Although on the other hand, again, it doesn't include something like a court record, which, again, legal scholars have a lot of interest in. I have, just to share a quick anecdote, I have had the experience with an IRB officer, someone who used to work for you. I won't say who. Who didn't believe me on this. And so the court records were public. And so I actually logged on to the Massachusetts court website and searched their. I won't say. I won't say his or her, but their name and pulled up their. Their court records and shared them with them. Shared the court records with them. And you know, in the context of a video call, which was a moment, you know, in which it was persuasive, put it that way. [00:31:00] Speaker C: Oh, I'm sure it was incredibly persuasive. [00:31:05] Speaker B: So anyway, there. So this again gets into this difficulty. One, One question I have for you that I think a research scholar might ask, put it this way, is to the extent that IRBs are, or IRB officials are repulsed by the amount of information that is on, that is in the public domain in the sense that it is, I shouldn't say in the public domain, it's available online or it's available on the dark web or something like that, is. Is interpreting private information in particular ways in the context of research a good way to address that problem? You understand my question. In other words, one could take the position, look, I'm an IRB officer. I can't regulate information access through my position. I just can't do that. Right. There have to be other institutions involved in the government and in the world that are going to regulate information access. Once something is out there on the web, especially if it's accessible via Google, I just have to throw in the towel and say that this is no longer private information. Is that an attitude that you find appealing or that you think that other people would share? [00:32:20] Speaker C: I I think it's reality of, of where we're at. I mean there's actually businesses that will what, that will clean your profile? [00:32:30] Speaker B: Sure, sure. [00:32:32] Speaker C: Because of that and depending on your locale, there may be government protections that from happening, such as in the eu. [00:32:43] Speaker B: In the eu they have the right to be forgotten. Right. And, but we don't have that in the United States. Might be a researcher's point. [00:32:48] Speaker C: Right, we do. I think, you know, one should be careful about what they're sharing. I think, you know, the Internet is the friend and foe and a lot of people have put a lot of information out there without thinking that this may come back in some way. And from the IRB perspective, you know, our role in the larger scheme of things is to act as a safeguard for those who choose to volunteer in research to ensure that it's conducted ethically, to ensure that no, you know, undue harm happens to those that, that volunteer their time, et cetera. Where that, you know, comes into question is specifically in this regard for a lot of research studies that are conducted without consent. And that does involve, you know, the use of information that may not be public, but rather may be private. And IRBs do consider, you know, what is, would this have any risk or harm to the rights and welfare. Is there a way to do the research in another way, will people be informed of this after the fact, etc. So there are some circumstances where, you know, it, this research will lead to the greater good, but it will involve using what might be considered private information and a human subject without that individual's knowledge. That's where it gets a little bit more challenging from an ethical perspective. You know, you can even think something as benign as like a chart review. [00:34:33] Speaker B: Right. Which is the, which is the closest medical analogy to this problem. But does still doesn't present many of the, many of the same problems because the charts themselves are governed, are presumptively private in most cases and on top of that have stringent federal law protection through hipaa. Correct, Right. So therefore there is a reasonable expectation that they will be kept private unless of course, the individual consents to make them public. [00:35:03] Speaker C: Exactly. And so, you know, a corollary to this is with artificial intelligence. [00:35:08] Speaker B: Right, right. [00:35:09] Speaker C: Which just eats huge amounts of data. [00:35:11] Speaker B: Sure. [00:35:12] Speaker C: You know, when does it become a human subject? [00:35:15] Speaker B: Right. And there we can go to movies like Blade Runner and, and, and There you go, 2001, et cetera. Yes, exactly, exactly, exactly. [00:35:27] Speaker C: Now, you know, with, with the revisions that, that came into play, there is a little carve out in There that the feds are supposed to reevaluate what's considered identifiable at specific time frames. That has not happened. And it. Who knows? [00:35:49] Speaker B: I. I was, I was gonna. I was waiting, wondering whether we would ever fit this in. But at least as of the time that we're reporting this, it doesn't seem like it's likely to happen in the immediate future. Is that right? Because is the. The staff that are. That are. Well, why don't you tell me? There's been. There's been some dojification in this area. [00:36:09] Speaker C: Yeah. Correct. And it's. It's really kind of uncertain what's going on. You almost feel like you're listening to gossip in a way. [00:36:17] Speaker B: Right. [00:36:17] Speaker C: But. [00:36:18] Speaker B: Yeah, what do we know? We know that certain people have been. That certain offices have been cleaned out and that the staff are no longer there. And the federal government. I mean, is that. Is that correct? [00:36:27] Speaker C: Yeah. And especially with office Human Research Protection. And I know that the think tank for ohrp, what was known as Sacchar Secretary's Advisory Committee on Human Research Protection, that's been officially disbanded. [00:36:40] Speaker B: All this is within hhs. Right. Which again, just to review, is the lead agency on the common rule, which we talked about in our last conversation. The common rule is HHS leads the effort to revise and to maintain the common rule. [00:36:57] Speaker C: Exactly, exactly. And, and again, getting back to what's considered identifiable or private, for that matter, a lot of IRBs have, you know, set up, you know, criteria that helps them make that. That decision. So, you know, it's. One can think of, you know, this is where the regulations are the floor and every institution can always add on. So, you know, you may be very accustomed to working within one irb. You go to another institution and like, whoa, this is very different. [00:37:32] Speaker B: And we, and we have had that experience. You and I have had that experience. I want to share the anecdote, one in particular that's so startling. I'm still getting over it. In a minute. Let me just to close the loop on the, on the hhs, the staff are no longer there. That review may not occur in terms of updating what is private and what is considered public. Is that right? Yeah, correct. We're just not sure when that'll happen. [00:37:59] Speaker C: We're not sure about a lot of things. Right. [00:38:01] Speaker B: And then to go on to your thought about this is going to be a theme that I'm hoping to explore specifically later, which is variation across IRBs, that. Which will become more and more of a problem. It already is becoming a problem. Including for legal scholars and social science scholars, as inter institutional cooperation and joint research projects become more common. Right. So there was some recent provision in the 2019amendments to the regulations to try to typically set up a situation in which one IRB was enough and the others could arrange to say, okay, you'll be the lead IRB if you, if you, in a multi institutional study, if you approve, fine. But that doesn't really, that doesn't solve all the problems. It does. It's a big step forward. But if the IRBs are interpreting the regulations, not just building on top of, but interpreting the regulations to mean quite different things, we still have issues potentially. [00:39:00] Speaker C: Correct. You know, even though one IRB is doing the IRB review, the other institutions involved, their IRBs still have a dog in the fight because they have to review for institutional policies which may go above and beyond as well as local context issues. Maybe there is, you know, state differences and how they define, you know, who's a minor or not, etc. So that review is still happening locally and that may be a contentious issue between IRBs. [00:39:36] Speaker B: Right. And this by the way, presumes that there is an IRB review. And we might as well just go ahead and share this antidote because, because I'm laughing so hard about it that, that this particular study, which occurred several years ago, this, which began, put it this way, several years ago, involved recruitment of 17 year olds while they were incarcerated, typically for enrollment in a study that would randomize them either to prosecution for a low level felony charges as usual, or to a diversion program. Typically they'd been arrested again as young as 17. Since the study occurred in a state that defined adults to be 17 year olds under the, under the regulations, they were adults, they were not child children because as something we'll talk about, but provisions, the regulations involving children define children as children under state law. So if they're a child, if they're an adult under state law, they were an adult. So we didn't trigger the children, but nevertheless we had 17 year olds and we were randomizing them either to business as usual, prosecution as usual, or to a diversion program. And they were going to be able to talk to their lawyers, but they typically hadn't slept for 24 hours. Because one of the lovely things that we do when we arrest people is deprive them of sleep. It's almost like going in a hospital in that sense. And so, and we were, since it was a multi institution study, we were trying to arrange to where the location that was closer to. And in the state where the research would take place, that institution's IRB would be the lead irb, that they would be the one. And I remember working with you on this and calling you on the phone when I got an email saying that the local cited IRB was going to take the position that this was not human subjects research. And that put us in the position of saying, number one, what planet am I on? And I remember laughing harder than I've laughed in a long time when I got this email saying, you got to be kidding me. I've got 17 year olds, they're incarcerated. Which triggers a whole other set of the regulations we're going to talk about. Right. And they're arrested and they haven't slept, they're highly vulnerable. And we're going to randomize them to prospective treatment and interact with them to go, you know, to, you know, to Helen and, you know, etc. And we're, and we couldn't even if you, you informed me, am I right about this? You could. That Harvard's IRB couldn't even defer to that determination even if we wanted to, which we didn't, because it wasn't actually an IRB review, it was a declining to go through IRB review. [00:42:31] Speaker C: Correct? Correct. Yeah. In order, in order to have a reliance between IRBs, it has to be considered non exempt. [00:42:39] Speaker B: Right. Which we'll talk about. But they were going to consider this exempt. Right. And therefore they couldn't. Right. Which actually is different from not human subjects research. But anyway, they were going to say we don't have to review it. Right. And so the variation across IRBs can be startling. [00:42:56] Speaker C: Yeah, that one was surprising. Very surprising. [00:43:00] Speaker B: I still remember thinking, my goodness, you know, number one, I desperately want IRB review on this study. Right. You know, to go back to a theme. It's not like I don't want there to be regulations here. I desperately want IRB review on that study. And number two, I can't believe, again, what planet am I on here if this is not human subjects research, what is? [00:43:21] Speaker C: Yeah, it's very interesting. And, and you know, I think I learn a lot about different things that go on at other IRBs, not only through the reliance process because you get to see their, you know, approval documents and how they approved a particular study, which is always interesting. But I think that also the ohrp, when they do an audit, either for cause or not for cause, those are all posted on their website. So those are also very interesting to look through to see how people, not only what were they not doing correctly, but how they did or did not rectify the situation. So those are always interesting tidbits. [00:44:07] Speaker B: And OHRP, again, is the office within HHS that oversees IRBs generally. [00:44:12] Speaker C: Correct? Yeah, correct. [00:44:13] Speaker B: Got it. Super. And still on the definition of human subjects, of a human subject, believe it or not, we're still going to have other sources of problems, which is the definitions reference to intervention or interaction with the individual. So one way you can trigger human subjects research is by obtaining data through intervention or interaction with an individual. And let's focus on that for a moment because again, in the medical context that's typically quite straightforward, but in the legal context it is not. And so let me give you a hypothetical which is not in any way a hypothetical. [00:44:55] Speaker C: It's a real study being tested. [00:44:57] Speaker B: Absolutely. Here. And by the way, for folks who are interested in. One of the reasons I'm so interested this way, in this is that this is an active area of research. [00:45:09] Speaker C: And. [00:45:09] Speaker B: I co authors and I are drafting a law review article, believe it or not, as you know, not as I speak, but this week we're drafting some more. And this is one of our hypotheticals that we're using in the law review article. Because of course, it's not a hypothetical, this is an actual research study. So the research involves randomization of offers of full legal representation, a traditional attorney client relationship by a civil legal services provider. Free civil legal services provider. So legal aid, what people think of when they think of legal aid. Randomization of offers of traditional attorney client relationships versus self help materials. So the two treatment conditions are the enhanced services group, where the enhanced service is the traditional attorney client relationship versus the business as usual group, which is the self help materials and the legal, and this is legal aid. And so as is the case in almost every legal aid setting, there are more people that are seeking the legal assistance than the provider has the capacity to provide the enhanced services to. And so we know that we're going to need to get consent from the potential clients. You know, that's easy because we are clearly interacting or intervening in their environment. We are, you know, they've come to the legal aid office, legal aid office could turn them away saying, sorry, we can't help you. We don't handle that kind of case, whatever it is, no research yet then. But when they, when they create a partnership with a scholar such as the, or an outfit such as the Access to Justice Lab, randomize those two, we're going to follow them. Even with publicly available information, even if we just did court records, we have interacted or intervened in their environment, they are clearly a participant. The question that arose in that context, though, is suppose the client here has a summary eviction problem where there's a lawsuit, they're facing a lawsuit to evict them from their housing unit and their tenants, they rent their apartments. Are the landlords human subjects? And the reason why that's important is because if they are human subjects, we might face an obligation under the regulations to obtain their consent. Is that correct? If they are human subjects, we might need to get their consent, maybe, or we could go through a waiver process. [00:47:38] Speaker C: Correct? [00:47:39] Speaker B: Right, right, right, right. But to go. To walk through the way the regulations work, if they're human subjects, then we trigger either a consent requirement or a. [00:47:47] Speaker C: Waiver analysis, if that's the level of review. [00:47:52] Speaker B: Right, right. If that's right. Exactly right. And so, and so the question, though, I think the question is, is that are they humans? Are they human subjects? [00:48:03] Speaker C: Well, I would ask how would they be human subjects? Are you obtaining private, identifiable information about these individuals? [00:48:12] Speaker B: No. [00:48:13] Speaker C: Intervening or interacting with them? [00:48:15] Speaker B: That's the, that's the problem. That's the problem is, are we intervening or interacting with them? And that's what I would. And I'm. And if your answer is no, then I'm going to say I agree with you. But this is, you know, but this is. This has come up, right? This is, this has come up. And the, the argument has been that, that we are. That the research intervenes or interacts in their environment by fundamentally changing the way in which something in which they are intensely interested proceeds. Because it changes arguably. And this is why we're doing the research. We don't know, but it arguably changes the probability of a court victory for them. In other words, they may not be able to regain possession of their house, of the housing unit that they pro. They think they have the right to regain possession of because of the legal assistance that is being provided pursuant to the study. So what do you think? [00:49:11] Speaker C: I think it's an overreach. [00:49:13] Speaker B: So tell. Unpackage that for me because I agree with you. But. But unpackage that. [00:49:17] Speaker C: Well, I mean, if you think about, you know, an analogous situation, let's say, you know, you give someone a survey that makes them feel sad and therefore they miss their dinner reservation that evening, and therefore the restaurant has a loss of, of a sale. Like how. How far do you go? And. [00:49:46] Speaker B: Yep, go ahead, go ahead. [00:49:48] Speaker C: I mean, right. So, you know, it's. You have to be careful about your boundaries. And there's actually, I don't even know if it's even around anymore. But way back in the day when I first started IRB work, there used to be a guide that was put out by. It was called oprr. That was the, the pre. Name of ohrp. [00:50:12] Speaker B: Right. Again, the, the. The sub office within hhs. [00:50:16] Speaker C: Correct. [00:50:17] Speaker B: That oversees this area. Yes, go ahead. [00:50:19] Speaker C: Correct. And they did a really nice job of, you know, really providing guidelines of what is the scope of the IRB's review, because any intervention and interaction may have an effect on ancillary aspects. But is that the IRB's job? And so if I find, if I, if it still exists, I'll share it because they did a really nice job of conceptualizing that. But you could imagine with that same logic how far out you could go with any particular research study then, you know, I mean, it's, it's almost, you know, that person and that person and that person. So my, my independent assessment would be, yeah, it seems to be a bit of an overreach. [00:51:10] Speaker A: Proof over precedent is a production of the Access to Justice Lab at Harvard Law School. Views expressed in student podcasts are not necessarily those of the A J Lab. Thanks for listening. If we piqued your interest, please subscribe wherever you get your podcasts. Even better, leave us a rating or share an episode with a friend or on social media. Here's a sneak preview of what we'll bring you next week. [00:51:34] Speaker D: For centuries, both medicine and law operated primarily based on traditions and anecdotes and. And sort of popular wisdom. But while medicine has transformed itself through empirical research and scientific methods, the legal profession remains largely stuck in its traditional ways. So the question is, what can law learn from medicine's journey? And what might a data driven renaissance in the legal field look like?