1 DEPARTMENT OF HEALTH AND HUMAN SERVICES SECRETARY'S ADVISORY COMMITTEE ON HUMAN RESEARCH PROTECTION MEETING TUESDAY, OCTOBER 5, 2004 The Advisory Committee met in the Jefferson Ballroom in the Radisson Hotel Old Town Alexandria, 901 North Fairfax Street, Alexandria, Virginia, at 8:30 a.m., Ernest Prentice, Ph.D., Chairman, presiding. PRESENT: ERNEST PRENTICE, Ph.D. Chair BERNARD A SCHWETZ, D.V.M., Ph.D. Ex. Secretary CATHERINE SLATINSHEK, M.A. Ex. Director THOMAS L. ADAMS, CAE Member MARK BARNES, J.D., L.L.M. Member CELIA B. FISHER, Ph.D. Member ROBERT G. HAUSER, M.D. Member NANCY L. JONES, Ph.D. Member FELIX A. KHIN-MUANG-GYI, Pharm.D. Member SUSAN KORNETSKY, M.P.H. Member MARY L. POLAN, M.D., Ph.D., M.P.H. Member SUSAN L. WEINER, Ph.D. Member EX OFFICIO MEMBERS: HOWARD L. BRADLEY Social Security Admin. KATHRYN LYNN CATES U.S. Department of Veterans Affairs ROGER CORTESI U.S. Environmental Protection Agency HELENE DERAMOND U.S. Dept. of Education SUZANNE GAYNOR U.S. Dept. of Housing & Urban Development DAVID KLURFELD U.S. Dept. of Agriculture DAVID LEPAY, M.D.,Ph.D. Food & Drug Administration AMY PATTERSON National Institutes of Health MICHAEL VIOLA U.S. Dept. of Energy 2 AGENDA ITEM PAGE CALL TO ORDER/OPENING REMARKS: Ernest Prentice . . . . . . . . . . . . . . . . . .3 SUBPART A ISSUES PANEL: Ada Sue Selwitz . . . . . . . . . . . . . . . . . .5 David Forster . . . . . . . . . . . . . . . . . . 21 Daniel Nelson . . . . . . . . . . . . . . . . . . 39 Steve Peckman . . . . . . . . . . . . . . . . . . 58 QUESTIONS FROM SACHRP:. . . . . . . . . . . . . . 86 PUBLIC COMMENTS TO SACHRP: Ivor Pritchard. . . . . . . . . . . . . . . . . .122 Mary Ann Cabo . . . . . . . . . . . . . . . . . .123 John Mather . . . . . . . . . . . . . . . . . . .126 SACHRP MEMBERS DISCUSSION:. . . . . . . . . . . .131 MOTION TO ESTABLISH SUBCOMMITTEE: . . . . . . . .142 AMENDMENT TO MOTION:. . . . . . . . . . . . .144/147 VOTE TO APPROVE SUBCOMMITTEE: . . . . . . . . . .152 DEFINITION OF RESEARCH V. NON-RESEARCH ISSUES PANEL: Michael Carome. . . . . . . . . . . . . . . . . .153 James Hodge . . . . . . . . . . . . . . . . . . .161 Mary Ann Baily. . . . . . . . . . . . . . . . . .185 Peggy O'Kane. . . . . . . . . . . . . . . . . . .203 QUESTIONS FROM SACHRP:. . . . . . . . . . . . . .219 PUBLIC COMMENTS TO SACHRP: Robert Levine . . . . . . . . . . . . . . . . . .250 MODELS OF IRB REVIEW: Michael Carome. . . . . . . . . . . . . . . . . .255 David LePay . . . . . . . . . . . . . . . . . . .259 Susan Kornetsky . . . . . . . . . . . . . . . . .268 Robert Levine . . . . . . . . . . . . . . . . . .280 David Forster . . . . . . . . . . . . . . . . . .293 Lowell Schnipper. . . . . . . . . . . . . . . . .316 DISCUSSION OF COMMITTEE BUSINESS: . . . . . . . .331 MOTION BY COMMITTEE:. . . . . . . . . . . . . . .365 AMENDMENT TO MOTION:. . . . . . . . . . . . . . .367 VOTE TO APPROVE MOTION: . . . . . . . . . . . . .368 ADJOURN: Ernest Prentice . . . . . . . . . . . . . . . . .368 3 1 P-R-O-C-E-E-D-I-N-G-S 2 8:38 a.m. 3 CHAIR PRENTICE: Good morning, everybody. 4 Welcome to the second day of the SACHRP. Welcome 5 Members of the Committee and Members of the Panel. 6 I'm going to briefly overview the agenda for today. 7 I would, however, encourage you to turn off your cell 8 phones. We have a rather in tact program today. 9 We're going to be starting with a Subpart A Issue 10 Panel, Ada Sue Selwitz, David Forster, Daniel Nelson 11 and Steve Peckman, who will go in that order. 12 The way these Panels are going to work, we 13 would like each presenter to come up to, I suppose, 14 here, if you are comfortable. You can change your own 15 slides or we'll get somebody else to change them for 16 you, so you would speak from here and then at the end 17 of the Panel's presentations, all four, we'll have all 18 of you come up and sit up here and respond to 19 questions from SACHRP. 20 I want to encourage everyone to stay on 21 time. You already know the time allocations, 15 22 minutes plus or minus, maybe a minute and a half or 4 1 so. You got that Ada Sue? 2 MS. SELWITZ: Yes, she said I could use 3 20. 4 CHAIR PRENTICE: 20? See, now, you're 5 going to really start a problem here. We're going to 6 have a break after that, then it's public comment 7 period, lunch, and then we go into our second Panel, 8 which is Definition of Research versus Non-Research. 9 There seems to be a lot of confusion these days as to 10 what exactly is research. It's interesting, I'm 11 leaving for Singapore tomorrow and they've asked me to 12 talk about exactly the same thing. They haven't 13 figured out what research is, either, so I guess they 14 are in as bad a shape as we are. 15 Then we will have a break. Then we move 16 into our Panel on Central IRBs at 2:45 to 4:45. Mike 17 Carome, David LePay, Susan Kornetsky, Bob Levine, 18 David Forster and Lowell Schnipper. We'll have a very 19 short period of time for discussion of Committee 20 business and then hopefully we are going to be able to 21 adjourn on time. 22 Now, these are the future SACHRP meeting 5 1 dates. I would encourage you to write those down. We 2 would like to invite you to attend January 31st and 3 February 1st, April 18th and 19th, August 1st and 2nd and 4 November 1st and 2nd in the year 2005. 5 Okay. I think it's time for the first 6 Panel to begin and I'm going to move over to the 7 table. All right. I am going to introduce each 8 speaker, but in the interest of time, I am not going 9 to read their bios. Their bios are available to you 10 and the speakers, I think, are known to almost 11 everyone here. So the first speaker will be Ada Sue 12 Selwitz from the University of Kentucky. Ada Sue, 13 welcome to SACHRP. 14 MS. SELWITZ: Thank you. I'm thrilled to 15 be here. I have to tell you all I was truly honored 16 to be invited to talk to you. I have followed your 17 deliberations. I think what you are doing has the 18 potential to truly strengthen human subjects' 19 protections here in the United States. I think you 20 guys are doing a great job. And I have to tell you I 21 was also really thrilled to be invited, because I was 22 asked to talk about Subpart A. 6 1 You know, honestly, I have sort of grown 2 up professionally with Subpart A. I mean, when I 3 started I was very young. I'm not so young now. I 4 grew up applying Subpart A, and this may sound 5 melodramatic, but it's true, Subpart A has been a part 6 of my life. You know, I don't know about you all, but 7 it's like some family members. There are parts of 8 Subpart A that I love and there are some parts of 9 Subpart A I would really like to change. So I was 10 thrilled to have a chance and my purpose today is to 11 discuss whether 45 CFR Subpart A should be revised, 12 either to strengthen human subjects' protections or to 13 reduce the unnecessary regulatory burden. 14 I want to acknowledge, I talked to lots of 15 people about this presentation. I got lots of ideas 16 from individuals and incorporated some of the ideas 17 into my remarks and I really struggled. I thought how 18 do I do this? How do I in 20 minutes, Ernie, actually 19 -- I did, I talked to Kelly on it, in 20 minutes, how 20 do I layout what I like and what I dislike about 21 Subpart A? And I decided the best way to do it was to 22 do a risk benefit analysis. Okay. 7 1 What are the risks of revision and what 2 are the benefits of revision? So let's talk with the 3 benefits. The benefit is what provisions need to be 4 changed? It's about first on my hit list, the 5 assurance requirement. I'm sure you all are familiar 6 with it, but, essentially, this requirement that we 7 have assigned assurance agreement and what kinds of 8 things need to go in the assurance agreement. 9 Now, if you ask me what do I think about 10 that, based on my years of experience, I don't think 11 it does not strengthen protections. It's an 12 unnecessary regulatory burden. And I think it's 13 especially problematic with those sites that are not 14 legally tied to your institution, but they are engaged 15 in research. This is where the problems I found lie. 16 Let me give you an example of what I'm talking about. 17 We had a researcher that was NIH funded 18 and she was doing research at a variety of non- 19 traditional research sites, including Girl Scout 20 Troops, and if you apply the engagement memo, we had 21 to get assurances throughout this country at places 22 like Girl Scout Troops. But honestly, I'm not 100 8 1 percent sure what the problem is. Okay. I mean, I'm 2 not sure whether the problem is the assurance 3 statement. 4 There is also -- I almost put on my slide 5 that provision 46.113 on cooperative research, because 6 if you look at the FDA version, it doesn't talk about 7 written agreements, but you look at 45 CFR 46, it does 8 talk about written agreements. But I have so much 9 problem with this, that I'm thinking I'm going to have 10 to get a full-time person. A full-time person to 11 manage assurance and other site agreements. And I 12 have to say OHRP, I know you have tried to streamline 13 this process using a variety -- I mean, you really 14 exerted a lot of energy to try to do it. But the 15 truth is it's a problem and it's still a problem that 16 I don't think adds to protection. 17 Next on my hit list continuing review. 18 Now, I think it's important, perhaps, that I explain 19 to you guys where I stand philosophically on 20 continuing review. I really believe if anything we 21 should be doing more at looking at what happens after 22 a protocol is approved. I mean, I think that's where 9 1 we can really strengthen protections. So 2 philosophically, I love continuing review. But from 3 a practical standpoint, I don't think it's working. 4 There is a lot of unnecessary bureaucracy. 5 So I asked myself questions. Do less than 6 minimal risk studies or studies in which no subjects 7 are enrolled need re-review on a yearly basis? I 8 think the answer is no. And so maybe what the problem 9 is is how, if I go back, this is written, but not less 10 than once per year. Because remember, I 11 philosophically support continuing review, but I just 12 don't think it works. There is a lot of paper 13 flowing. 14 Another question is are there more 15 effective models for assessing implementation of 16 approved protocols and I think there are. And maybe 17 in the Q&A we can talk about that, but more than just 18 a yearly review, based on a model where we have to 19 send X kind of material to every IRB member. You 20 know, is the problem 46109(e) or is it interpretation? 21 I want to give you answers today and I don't have 22 answers. I don't know what the problem is, but I do 10 1 think there are more effective ways to strengthen 2 protection through continuing review, and I would like 3 to see continuing review interpreted in a way, so that 4 we could use innovative approaches to try to approach, 5 particularly, for greater the minimal risk protocols. 6 Next on my hit list waiving informed 7 consent. Now, I got to tell you guys, I like the 8 framework and I like we got to have a way to waive 9 informed consent, particularly for social science 10 research. But in terms of applying it over the last 11 too many years, I found some problems. An example, 12 this criterion, you know, you can waive informed 13 consent if an IRB can make a determination that it 14 meets these for criterion. 15 Look at that second criteria. "The rights 16 and welfare of the subject is not adversely affected." 17 Well, it's the word rights that bothers me, because if 18 you have, say, a right as a human subject to give 19 informed consent, then technically I don't see how you 20 could ever waive informed consent. So it's the word 21 rights that I would remove. I think in practice what 22 IRBs really do is they focus on the welfare part of 11 1 this, practicably. 2 You know, I looked at my slides this 3 morning and saw what everybody else is going to say 4 and you're going to hear this again, and I bet if I 5 took each one of you that are on the Committee and 6 took you separately and said okay, Felix, tell me what 7 practicably means. Mary, you tell me what practicably 8 means. If you could even give me an answer, it would 9 probably be different. And so I think there is a 10 problem with this word. 11 On the other hand, I have to tell you 12 truthfully I think in applying this criteria, when you 13 put the word practicably within the context of a 14 research protocol, when you look at a protocol, 15 sometimes it does make sense and it's not as difficult 16 to apply. I mean, really for me, I think the rights 17 is more of a problem than practicably. 18 Next hit list reporting. And again, I'm 19 not sure the problem is this provision, Subpart A. 20 The problem, I think, has to do with the entire 21 system. I know that you are well-aware of it. It has 22 lots to do with other regulations and how they 12 1 interface with this. I won't spend any time on this, 2 because I know that you are already spending a lot of 3 time. 4 Definitions. What are my hit list? What 5 are things I don't like? I can't stand legally 6 authorized representative. I think it's hard to 7 apply. I think it's problematic from the standpoint 8 of state law. I actually think its too stringent in 9 many cases. Human subjects, this is going to be 10 talked about later. I won't spend a lot of time, 11 except living and dead, now, we have HIPAA, right? 12 And, I think, so now we have disharmony between HIPAA 13 and Subpart A. 14 About whom? I think this is going to be 15 talked about later, so I won't spend time here. So 16 maybe I can get it down to 18 minutes. But that I 17 think is a problem. The issue of what is research, 18 you're going to be looking at it. Quite frankly, on 19 what is research? I believe guidance could solve some 20 of those problems, I genuinely do, but I don't go 21 through the day rarely that I don't get a question 22 about is this research? 13 1 Exemption categories, I put them on my hit 2 list, because I think there is a lot of confusion. 3 And I could pick this to death, guys. If we had time 4 today, I could take you through and show you this word 5 and show you that word. So I think there is a lot of 6 confusion about exemptions. In truth though, I think 7 a lot of the confusion can be addressed through 8 guidance. I don't necessarily attribute it to how it 9 is written. I'm not sure my other colleagues will 10 agree with me on that. 11 Another part about exemptions though is 12 from a pure regulatory perspective. I think it would 13 have been good if in the framers of Subpart A they had 14 made as a requirement that it be minimal risk and fit 15 in the categories. I would have liked to have seen 16 that in the regulations. Now, in terms of practice 17 and applying it, I haven't found a real problem with 18 that. I mean, I have not seen exemptions that were 19 approved that were greater than minimal risk. But 20 likewise, if you ask me what would I change, that's 21 one of the items that I would change. 22 Is Subpart A outdated? You know, I hear 14 1 a lot about this. And I sort of agree with it that's 2 why I included it, but Subpart A, we didn't have 3 professional administrators. I remember I went to my 4 first PRIM&R meeting and everybody asked me now, what 5 do you do, are you chair? I said no. Are you the 6 secretary? No. I'm professional staff. And nobody 7 had ever heard of that in 1979. Okay. You know, like 8 there's that part in the reg that the IRB Chair must 9 appoint the expedited reviewers, well, in truth, the 10 IRB professional appoints the expedited reviewers. 11 I hear a lot. I included this last one, 12 because my faculty, my social science faculty would 13 probably shoot me if I didn't say that Subpart A was 14 based on an NIH model and doesn't work as well for 15 social science. But, you know, having said that, and 16 Ernie asked us to talk about where -- I don't blame 17 Subpart A as much as I blame sometimes how IRBs are 18 applying Subpart A, in truth. I think there are ways 19 to apply it in a social science setting, but IRBs 20 simply are afraid or I don't know why they aren't 21 doing it. 22 Okay. Let's switch to risks. One 15 1 analysis is doing this risk benefit assessment. What 2 are the risks of revising Subpart A? What works well 3 in Subpart A? And, you know, I know you can say to me 4 well, Ada Sue, if we revise it, that doesn't mean we 5 will get rid of the good stuff. We'll keep the good 6 stuff. Well, you know, I sort of believe in adages 7 like "be careful what you wish for" or "don't throw 8 out the baby with the bath-water." I heard one 9 recently and I have to write it down. I wish I could 10 tell you who I heard it from, but I don't know. But 11 I can relate to it. "Most problems we face today were 12 created by solutions we implemented yesterday." 13 Okay. And so I think that if we're truly 14 -- if I'm going to lay this out as a risk benefit 15 assessment, you need to think about the risks of 16 opening up Subpart A. So what do I like about it? 17 Well, I have to ask the core question. Would revising 18 Subpart A make ethical decision making more effective? 19 I mean, that's what we're really about, isn't it? 20 Whether it is sound ethical decision making, and my 21 reaction is no, it wouldn't. 22 I think, though, we are all guilty, myself 16 1 included, of always looking to OHRP and IRBs are 2 looking for OHRP and the FDA, we're not going to let 3 the FDA off the hook, and say if you just give us more 4 guidance or if you tell me what this word means, we 5 could do it right. Okay. And so I think there's a 6 tendency for IRBs to look to the regulations for 7 assurances that it's not possible or perhaps even 8 appropriate for regulations to provide. 9 I'm convinced after watching IRBs and 10 Animal Care Committees for over 20 years that it's not 11 possible to eliminate the "grey" associated with sound 12 ethical decision making or to anticipate new questions 13 or ethical challenges that can surface as science 14 progresses. Okay. I'm really convinced of this. You 15 say well, Ada Sue, where are you going? Where is the 16 strength then? What do you like about Subpart A? I 17 really believe it has provided an excellent framework. 18 An excellent framework for ethical decision making. 19 As a matter of fact, I'm amazed at how 20 well it has stood up over time. Truly amazed at how 21 well it stood up over time. And what do I like about 22 it? Some of the very things I told you a minute ago, 17 1 I would change. All right. I like the exemptions. 2 I like the expedited review. I think there is 3 flexibility there without diminishing protections, 4 criteria for approval. The criteria for approval, 5 there's almost a one-on-one match. If I had had time 6 today, I could have illustrated between criteria for 7 approval and the ethical principles. Okay. I think 8 it's excellent. 9 Basic and additional elements of informed 10 consent, yes, I could pick all of these to give. If 11 you give it to me, I can show you words I don't 12 understand and I can show you picky little problems, 13 but overall, I think, it's a great framework, 14 including the fact that we can waive informed consent 15 and documentation of informed consent. I hate the 16 short form. I'm sure some of you all will talk about 17 it. But, you know, overall, I think, it has been an 18 excellent framework that has held up over time. 19 Next issue, let me go back. I better 20 explain this before I say it, because I'm going to 21 make people mad. I really think from a pure 22 regulatory standpoint, pure regulatory here, not human 18 1 subjects' protection, but regulatory, I believe the 2 current system is set up to fail. I think there is so 3 much detail that IRBs and investigators you can't 4 comply. And I hate to say it, but I'm not in the 5 field and I'm telling you there is just so much. And 6 I don't think it's a matter of just how much money my 7 institution puts into my system. There is a point at 8 which there's just so much detail you can't do it any 9 more. 10 Okay. So what does that have to do with 11 why I like Subpart A? Well, I think, Subpart A avoids 12 prescriptive detail and is straightforward compared to 13 many of our other regulations. I mean, I can read 14 Subpart A and understand it. I don't have to hire 15 Mark Barnes, even though he would tell me I should, 16 okay. Actually, we did. 17 MR. BARNES: You need to. 18 MS. SELWITZ: And actually, we did hire 19 Mark Barnes, but to tell me about Subpart A. I liked 20 that it's prescriptive, it's straightforward. I 21 compare it to more recent regulation like HIPAA. You 22 know, this is, I think, a real positive of Subpart A. 19 1 Next issue, one of the most significant 2 regulatory challenges, I think, that face IRBs, and 3 I'm not telling you guys anything you don't know, 4 okay, but one of the most significant is that the 5 research community is the lack of harmony between the 6 federal funding and regulatory agencies with respect 7 to Subpart A and how it is interpreted. Now, I could 8 blame it on Subpart A and I could say well, if it was 9 written different, there wouldn't be any of this lack 10 of harmony. 11 But, you know, I'll be honest with you 12 guys. What I believe is it doesn't matter how it is 13 written. When you have different agencies, you're 14 going to get different interpretations. Okay. I 15 mean, I do. So where am I going with this? What do 16 I think is a prime strength of Subpart A? Subpart A 17 is based on the Common Rule. The Common Rule is the 18 basis of Subpart A. Okay. And I believe that if HHS 19 revises Subpart A that there is no guarantee that the 20 other agencies will revise them accordingly. 21 I think there is just simply no guarantee. 22 And I believe that having uniform regulations provides 20 1 one of the only avenues we have for harmonizing 2 requirements. We maybe are never going to make it, 3 but at least it gives us something, a common uniform 4 set to work off of. I'm afraid that if Subpart A is 5 revised, it could intensify this lack of harmony that 6 we are all struggling with. 7 My final issue, I apologize, this final 8 issue is none of my business. And, in fact, I bet 9 it's not the Committee's business in making your 10 deliberations. But I was on a roll and I couldn't 11 help myself. Okay. And so my final is a practical 12 consideration. There are other parts of 45 CFR that 13 are truly broke and you all have identified those. 14 The work you are doing is fabulous. They need to be 15 fixed and I think any agency has limited resources in 16 correcting. So, I think, if you are going to do a 17 risk benefit analysis, you have to consider, you know, 18 what really needs fixed versus what could perhaps be 19 improved, but doesn't really need to be fixed. 20 So in conclusion, my purpose today was to 21 discuss whether Subpart A should be revised and I 22 tried to address that by outlining for you the risks 21 1 and benefits. Thank you very much. 2 (Applause) 3 CHAIR PRENTICE: Thank you, Ada Sue. You 4 have another 15 seconds. Our next speaker would be 5 David Forster, who is a lawyer and the Assistant Vice 6 President, Office of Compliance at Western IRB, 7 located, as most people know, in Olympia, Washington. 8 Welcome to SACHRP, David. We're looking forward to 9 your presentation. 10 MR. FORSTER: Great. Thank you. And I 11 have way too many slides, so I'm just going to talk 12 really, really fast and go all the way through. I'm 13 going to cover a lot of the same points as Ada Sue, 14 actually, so I can go pretty fast through some of 15 those. But I structured mine a little different. 16 Basic questions, does Subpart A protect subjects? 17 Does it have too much regulatory burden? Those are 18 the questions we were asked. And I just wanted to 19 start off with something this Committee knows very 20 well. It's legislation. 21 Legislation, all legislation suffers from 22 the problem that it's very hard to draft so that it 22 1 applies correctly to all of the cases to which it 2 applies. It's the problem with legislation. So it's 3 not unique to these regs here. It's a problem all 4 through. So do I think Subpart A is broken from a 5 human subject protection standpoint? No. Could it be 6 tweaked? Sure. 7 Do I think there is extra regulatory 8 burden? I have a hard time thinking where to cut 9 back, you know. I see one out of 20 protocols that 10 needs to be stopped, but the rest they just kind of go 11 through and get a bunch of consent form tweaks and, 12 you know, there is regulatory burden there, but I 13 don't know where to cut back on it. And I tried to 14 think should I think outside the box? Do we really 15 need individual review of each protocol? I couldn't 16 get myself away from that either, so I'm kind of stuck 17 within we need something like Subpart A. 18 So what I put as far as the rating scale 19 is do we need to change the regs? I have a couple of 20 those where I think we should, but mostly I think 21 guidance could do it. So I talked to Charlie McCarthy 22 on something else and he said boy, if I was going back 23 1 to look at Subpart A, I would change the exemption 2 criteria. We wrote them in one night, because the 3 Secretary of DHEW rejected them. I had three days 4 before the Carter Administration ended, so in one 5 night we put them under there. 6 They are confusing. The structure is 7 awkward. There is all sorts of unidentified terms and 8 there's this footnote that's a page away that says you 9 can't apply children, prisoners, etcetera in various 10 ones. It's just confusing. Do you need to revise it? 11 You know, we can live with it. Now, that I know that 12 it was written in one night, I think it's really 13 pretty well done, you know. 14 Here is one we do need to change, the 15 definition of LAR. This is a nightmare. It kicks off 16 to state and local law, most state and local law 17 doesn't address it, with a couple of exceptions. OHRP 18 has put out some guidance that's very difficult to 19 interpret. Not guidance, but citation letters. I 20 would like to see a definition put forth in the 21 Federal Regulations that stricter state laws could be 22 held up against, but that would be there for us all to 24 1 rely on. Because right now, I think there is a lot of 2 confusion and there's a lot of legal liability in 3 this. 4 We're going to talk about this later. I'm 5 not going to spend any time on it. The definition of 6 research is difficult. I was kind of playing around 7 trying to come up with a better one. I don't know if 8 I could. I think guidance might be the way to 9 approach that. Another problem with the definition of 10 research, of course, as Ada Sue mentioned, is the 11 different agencies interpret it differently. CDC is 12 just off on a roll with this one, and I kind of get 13 into arguments with them about that. 14 Okay. Definition of human subjects. They 15 came up with -- OHRP really started this idea of 16 secondary subjects when they shot a -- put a shot over 17 the bow at VCU at Mid-Atlantic Twin Registry and then 18 they really haven't said much since and IRBs are 19 incredibly inconsistent on this issue of whether 20 secondary subjects exist at all and what you do with 21 them if you do think they exist. So there's a lot of 22 inconsistency across the country. 25 1 Minimal risk there is lots of discussion 2 on this Panel. The one thing I thought I would 3 mention is I think it needs a better definition. It 4 needs an objective tied down definition. The one 5 thing I hadn't seen in the newspaper reports, etcetera 6 on SACHRP or the various reports is whether you guys 7 had thought about this as a differential risk. And so 8 as I give an example here, to study investigation or 9 new formulation of an approved drug for asthma is 10 minimal risk, because the risk of the intervention is 11 the same as for the approved drug for asthma. 12 I see this reasoning all the time popping 13 up. And here is a classic case, there is this NIH 14 memo out there and some of you may have seen this, 15 this protocol floats around a lot of institutions. 16 It's this drug that's used for intrathecal 17 methotrexate overdose. You have to give it almost 18 immediately. You don't have time for consent. And 19 there's this memo out there from NCI that says the 20 difference is minimal between the risk of an 21 undesirable outcome when standard accepted therapy is 22 used and the risk of undesirable outcome with the 26 1 experimental therapy. 2 Okay. I can use that kind of logic and I 3 can get anything to be minimal risk, right, in way of 4 consent, why not? So I really think that you, as a 5 Committee, when looking at minimal risk should be 6 careful of this differential risk type of argument 7 that is out there. It just comes up in a lot of 8 different formats and it's a hard one to get yourself 9 through in a logical manner. 10 Okay. A very small point. The IRB regs 11 only say that the IRB has the authority to prove or 12 require modifications that are disapproved, but we all 13 have table to defer approval in principle in all these 14 other things. We all use them differently. I think 15 some guidance there might be nice, but it's kind of a 16 small point. 17 Continuing review, like Ada Sue, I 18 wondered do we really need annual continuing review 19 for minimal risk research? And I think maybe not, but 20 I wouldn't want to see that lessened until minimal 21 risk was really tied down as a definition. So it 22 would take a change of the regulations to do it. I'm 27 1 not sure if it's worth cracking open Subpart A for 2 that. 3 Expedited review. There is some 4 regulatory burden created by the fact that the 5 expedited reviewers cannot disapprove research. It 6 has to get kicked up to full board and, you know, that 7 applies to the minor changes in research too, so you 8 have an ad that comes in after research is approved. 9 You can't disapprove it, you have to kick it up to 10 full board. So a simple saying no to an advertisement 11 has to go to full board and it takes a long time to 12 do. I think you could -- that would get rid of some 13 burden and it would take a change to the regulations. 14 Here is an interesting thing, criteria for 15 IRB approval, 111. Here is really what IRBs are 16 supposed to do. This is the kickoff for the ethical 17 decision making. Now, here is the Belmont Report. 18 And where is it? It's after expedited review, right? 19 It's just another little checklist bearing all this 20 procedure stuff and then your substantive material is 21 buried in the middle and it's very short and 22 nondescript. Now, that's both good and bad. And I 28 1 kind of want to bring that point out. 2 First of all, (a)(2) starts off "Risks to 3 subjects are reasonable in relation to anticipate it 4 benefits any of the subjects, and the importance of 5 the knowledge that may be reasonable to be expected to 6 result." Does this require or does this include 7 scientific validity and value to society? I think it 8 does. But there are IRBs out there that have decided 9 it doesn't, and I know that there was a lot of big 10 flap about this recently with the OHRP when they went 11 to look at the ventilator study. 12 A lot of people were saying that they had 13 overstepped, that the IRBs and the OHRP shouldn't be 14 looking at scientific benefit. I don't think you can 15 do a risk benefit analysis without looking at these 16 things. But it's not clear from the regulation that 17 IRBs should. 18 Here is an interesting one. "The IRB 19 should not consider possible long-range effects of 20 applying the knowledge gained in the research." Well, 21 how do we look at things like sex election techniques 22 for embryos? And how about public registries? If the 29 1 IRBs aren't supposed to be looking that far out on how 2 the data is used, well, how far is this boundary that 3 we're not supposed to go beyond? It's in the regs. 4 I don't know where that boundary is. Some guidance 5 would be useful. 6 Here is where I think we have had a lot of 7 discussion yesterday and it's a point that vulnerable 8 subjects is not well-defined. It's confusing and I 9 think there could be a lot of guidance on this to try 10 and clarify. I know my own IRB members get confused 11 all the time. What's vulnerable? And they think that 12 people who can be hurt are vulnerable, rather than 13 thinking of it as a limitation on autonomy. 14 Now, the word vulnerable isn't even in the 15 Belmont Report. It's not in the National Commission 16 thinking, although apparently was underlying it, but, 17 you know, it's not in there. In the Belmont Report 18 they talk about people with limitations on autonomy. 19 So I think it would be nice to tie that back more 20 strongly in the public knowledge. The information is 21 all there. It's just that a lot of people on IRBs 22 don't know that that's the idea of vulnerable, at 30 1 least that's what I believe the idea of vulnerable is. 2 I think guidance is enough here. 3 Defining additional safeguards might be 4 nice. I don't know what they are. I tried to comb 5 over the list once. You get into that problem if you 6 write a list and then something is not on it, you 7 know, it can be additional safeguard stuff issue. So 8 46.111 for me is the crux of Subpart A. It's where 9 the ethical decision making gets its kickoff. It is 10 very broadly written, which is nice, because you can 11 use it very broadly, like the Belmont Report with the 12 principles approach. It's broad. 13 Being broad and non-specific also means 14 that people can fail to use it correctly and you can 15 come up with a pretty skimpy interpretation. That's 16 a strength and weakness. I like the Belmont Report. 17 I think it provides a good framework, but you can 18 misuse the Belmont Report. So I don't know whether we 19 would want to go in and crack the subpart, you know, 20 46.111 and try to make it better. We could make it 21 far worse. 22 Suspension or termination. This is kind 31 1 of a funny thing. Technically, under the regs the IRB 2 only can suspend the research if the research isn't 3 conducted in compliance with the IRB's requirements or 4 there is unexpected serious harm to subject. Well, 5 what if there is just the potential for harm? We do 6 that all the time. We have written our SOP so that we 7 have far more reasons we can suspend or terminate. 8 But quite frankly, I'm always worried some 9 investigator is going to turn around and sue us and 10 say hey, you didn't have the regulatory authority to 11 do that. So I don't know, I wouldn't crack the egg 12 for that one, either. 13 IRB records. This is a tough one. One of 14 the things that came up for us is, for instance, with 15 the minutes. Do you really have to have the basis for 16 requiring changes in our disapproving research in the 17 minutes? Can't it just be over in this -- over here 18 in your computer system? The reason I bring this up 19 is for us, we take our consent forms, we redline them 20 in, and when we make changes we use a legend code. 21 This for regular compliance. This is for state law, 22 etcetera. So we have to -- and we just used to put 32 1 our legend code on our minutes and say that's why we 2 changed our minutes. Not good enough. 3 So now, we put our redline consent forms 4 in our minutes. Luckily our minutes are electronic 5 now, but if you print them out, they are about 400 6 pages, because of all the consent forms attached, or 7 longer. And then, of course, like say under the State 8 of Maryland Law where IRBs have to turn over their 9 minutes, then you have to go through an expurgate all 10 of that stuff in the minutes. It gets to be quite a 11 burden when there are these interpretations that are 12 so rigid as to what the minutes have to contain. It's 13 a difficult issue. 14 This is something that's getting to be a 15 problem for us, just as a record keeping matter, 16 correspondence has to be kept. Email chains, you 17 know, you can get 20 people commenting on some of 18 these things. They go for pages. You have to get it 19 all in your IRB record somehow, and I would just like 20 OHRP to be aware of these electronic record keeping 21 possibilities that are out there and have their 22 guidance keep up with it. I think that a lot of the 33 1 guidance is still very paper-based. So just to be 2 aware that there is a lot of change out there in the 3 record keeping capabilities that exist. 4 Informed consent. You know, people always 5 complain, content there is a process and why do people 6 not pay attention to that? Well, there's only one 7 sentence mentioning the process. The rest all focuses 8 on documentation and the elements. This is a tough 9 one. The part of informed consent regulations, it 10 says you can't waive any rights and the language is a 11 little more complex than that. It very clearly 12 prohibits waivers of negligence. I think it's getting 13 settled on the property rights issue that you can't 14 waive property rights. 15 Here is the one I'm facing now. What 16 about ownership of the intellectual property that 17 results from the research? I think under IP law they 18 probably do own it, the sponsors, Government and 19 private, but there still appears to be a waiver if you 20 put that in there. We're taking it out. So, you 21 know, I'm like Ada Sue, you know, I like their point 22 that maybe more guidance isn't the solution, but I was 34 1 just thinking change the regs or values, so I put 2 guidance down. 3 I thought about should we change this list 4 of the required elements? What could we take out as 5 a regulatory burden? What really doesn't help 6 subjects? You know, the list isn't bad. I could come 7 up with some minor additions, but the IRB already has 8 the right to put those in under the regulations. They 9 can put in additional important information. I 10 couldn't think of anything to take out. I like the 11 elements. 12 Waiver of informed consent. I think all 13 four of us will probably talk about this. Oh, wait, 14 I'm sorry, I jumped ahead. This is one where, this is 15 a waiver of informed consent under 116(c) for state 16 projects where they are looking at various types of 17 benefits programs. You can waive consent for it if 18 it's a state or local Government. But guess what, if 19 it's a federal agency doing it, then it's exempt. 20 Exact same language, right? So I think it's a little 21 hubris from here inside the beltway that somehow the 22 federal agencies are safer and think better of human 35 1 subjects than the state ones. I think we should make 2 it consistent, so that they are both either exempt or 3 you both need IRB review for the waiver of consent. 4 Okay. 116(d), the waiver of informed 5 consent. There are all sorts of problems with this. 6 One is I have highlighted here that "The IRB may 7 approve a consent procedure which does not include or 8 which alters some or all of the elements of informed 9 consent. If, one, the research involves no more than 10 minimal risk." Well, then you can only alter the 11 elements of consent for minimal risk research. 12 Anything above minimal risk you can't alter the 13 elements. 14 Now, there have been, you know, people 15 saying that you can, but five years from when there is 16 all sorts of different people at OHRP and, you know, 17 your compliance always comes when they are looking at 18 something that is three or five years old, you know, 19 just ignoring that black and white reading may get us 20 in trouble. The waiver alteration will not adversely 21 effect the rights and welfare of the subjects, just 22 like Ada Sue was saying. 36 1 What about dignity rights? What about the 2 fact that people don't want their records looked at by 3 other people? And there's a violation if that 4 happens. And, you know, I don't know if you have seen 5 the Research America, there is a poll that comes out 6 once a year, very pro research organization. And one 7 of the questions they ask is assuming there is no way 8 anyone will have access to your identity, would you be 9 willing to release your health information, so that, 10 you know, research can be done? 11 29 percent say no. 29 percent of people 12 don't want their records looked at. But IRBs are 13 waiving consent for records review research all the 14 time. So we kind of tend to put a consequential as 15 harm on this definition instead of a dense logical 16 view. Practicably, I think, Ada Sue did a great job. 17 I don't know what it means. 18 Here is one I do feel really strongly 19 about, just get rid of the short form. The only time 20 the short form pops up is when somebody wants to do 21 consent in the back of an ambulance on stroke or heart 22 attack patients. Then they say, oh, we're going to 37 1 use the short form, you know. We'll get -- you know, 2 that's the only time it comes up. The thing is a 3 nightmare. 4 It's all sorts of extra signing 5 requirements, so I would just get rid of it. And 6 while we're at it, I would get rid of the OHRP 7 guidance on using the short form for non-English 8 speaking subjects. I think it was well-intended at 9 the time, but you can't short change consent to try 10 and satisfy justice. 11 A funny little thing, waiver of 12 documentation of consent. The IRB can waive 13 documentation of consent if it thinks that the main 14 risk of research is a breach of confidentiality 15 occurring because there is a signature on the consent 16 form, right? But then it says at the end, but you 17 have to ask the subject if they want to sign anyway. 18 So imagine this, you're doing an interview at a Needle 19 Exchange Program, you know, in some busy urban center, 20 and you say okay, we want to ask you these questions. 21 We're not going to have you sign a consent. I've 22 described the purpose to you. Now, do you want to 38 1 sign after all, you know? 2 It's just very awkward. It makes, you 3 know, in some of those situations people start to 4 think, as we've had reported to us, that the 5 investigator is really an undercover cop, because all 6 they're saying really is sign it. So I think it's a 7 good idea at times to respect autonomy in that way and 8 the documentation requirements for record keeping, but 9 it's not always a good idea and it's a requirement as 10 the reg is written. 11 And then there's the funding catch-22 that 12 even though just in time funding helped a lot or just 13 in time review, excuse me, there is still this 14 situation that occurs where an investigator wants to 15 finish their research proposal and want to get a grant 16 to do that. They can't get the grant until they get 17 IRB review. They don't have a finished project to 18 give the IRB. And so all the IRBs have this kind of 19 approval and principle or some funny thing we do. 20 But, you know, we always, I always have 21 this problem where I write this letter saying you're 22 approved in principle, but it's not approved, right. 39 1 And then they turn it into the institute and the 2 various institutes of health. Some of them take it, 3 some don't, some call up and say that's not an 4 approval, and so we kind of back off and take out 5 sentences until they say okay, that's an approval, 6 we'll give the grant. But there's hopefully still 7 enough in the letter left to make sure that it looks 8 clear, it's clear to the investigator, they have to 9 bring it back to us. 10 So 20 minutes. Was it close? 11 CHAIR PRENTICE: Only 13. 12 MR. FORSTER: You're kidding. I couldn't 13 do it that fast in my room last night. All right. 14 That's it. 15 (Applause) 16 CHAIR PRENTICE: Thank you, David. All 17 right. Next up is Dan Nelson. Dan Nelson is a social 18 professor of social medicine and pediatrics and 19 Director of the Office of Human Research Ethics at the 20 University of North Carolina Chapel Hill. Welcome, 21 Dan, to SACHRP. 22 MR. NELSON: Thank you. Thank you, Ernie, 40 1 and thank you to the audience and to the Members of 2 SACHRP and I, too, am honored and appreciate the 3 opportunity to talk to you today. My colleagues at 4 UNC, I think, had this mental picture of klieg lights 5 on a Congressional testimony with lots of angry 6 Congressmen facing you across the way and I assured 7 them no, this is a friendly group. I maybe even 8 helped nominate some of them to get on in the first 9 place and we'll just have a discussion among friends, 10 and so remember that when I say something you don't 11 like. But I appreciate the chance to give a critical 12 assessment of Subpart A. 13 The outline is easy. I'm going to split 14 into half general observations on Subpart A as a whole 15 and then some specific problem spots. I often have a 16 problem using too many slides, and so I'm delighted to 17 look at the stack and see that I have fewer slides 18 that any of my colleagues, so I think I can get 19 through this. 20 So beginning with some general 21 observations. The first message, if it ain't broke, 22 don't fix it. And I don't think Subpart A is broke as 41 1 a whole. I think it has served us well for the most 2 part. There are some parts in need of fixing, perhaps 3 in need of guidance and I will get to that in just a 4 second, but I don't think it's in need of total 5 overhaul. Problems where there are, and there are 6 problems, I think, lie more with implementation, 7 application, interpretation and the confusion that 8 comes from misapplication, over-interpretation and 9 poor implementation. 10 Well, if you accept that, then what we 11 really need is a lot of guidance and interpretation 12 from you all as to how to interpret and apply the regs 13 that are vague in parts and purposefully so. So it's 14 great to have this guidance. Indeed, I would say it's 15 critical to have this guidance and much good guidance 16 exists thanks to OHRP and others, but I have to say 17 that it's difficult for the working, the people out in 18 the field to assimilate this and integrate it into 19 their daily practice. 20 And here is an example of that and this is 21 not to slam OHRP. In fact, I could at least find all 22 these in relatively one spot on the OHRP web, but 42 1 there's a lot out there and, again, I think it's 2 difficult to tie all this together. I have talked to 3 people over the last couple of weeks who have been in 4 the field for 20 years and didn't know half of this 5 existed, so you can bet that they are not integrating 6 that into their daily lives and their daily routines 7 in working with IRBs. 8 So a few examples. We have internal memos 9 from division chiefs within OHRP to their own division 10 that may or may not be interpreted by the IRB 11 community as something they should even be looking at 12 or if they can find it and know about it. One of the 13 most useful things out there, engagement of 14 institutions in research, not everybody even knows 15 where that is, even people with a great deal of 16 experience. 17 A lot of dear colleague letters, OPRR 18 reports by the dozens. We have got determination 19 letters to institutions by the hundreds, 20 unfortunately, and great places to learn, but good 21 luck taking those hundreds of letters. And we have 22 staff who love to start their day by reading through 43 1 the latest letters and they learn a lot, but then to 2 take that and put it into the daily practice again is 3 an issue for us. 4 You have actually done some of that hard 5 work by compiling into the OHRP common findings, but 6 again I think to tie all of this together is the point 7 of this slide. I guess, you can find opinions that 8 have been given to individual groups, the Oral History 9 Association being one recent example. The IRB 10 Guidebook was an early attempt to tie some of this 11 together, but I would say as more of an academic 12 exercise than a practical, functional application 13 exercise, but we could debate that later. 14 And there are other things that are just 15 out there that people don't know about unless they 16 happen to have attended a workshop at PRIM&R or happen 17 to get together with you guys on a site visit and work 18 through some of these issues, the Human Subjects 19 Research Decision Charts as an example of that. 20 So again, a lot of good guidance out 21 there, difficult to tie together, and I would say what 22 we would really need is a user's guide to the regs 44 1 that somehow takes your guidance, your 2 interpretations, ties all this together into a 3 working, living, evolving document or source of 4 guidance beyond just the website that is consolidated, 5 updated, integrated, validated and, unfortunately, 6 complicated. I think somehow, and I would be happy to 7 talk about this more during the Q&A period, how can we 8 take all this good guidance that we desperately need, 9 keep it integrated, current and really apply it to our 10 daily routines? 11 Yes, times have changed. I won't say that 12 it's completely out of date. I think it has actually 13 weathered quite well and join Ada Sue in that 14 sentiment, but on the other hand I think we do have 15 what I call single site regulations in a multi-site 16 world. We have regs that were drafted 30 some years 17 ago, 20 at least to govern research that really has 18 evolved and doesn't look like it did 20, 30 years ago. 19 And so we have internal discrepancies, I 20 think, between the importance and relevance that we 21 all give to local review and we need to maintain. 22 There is an internal conflict there with the realities 45 1 and demands of the modern research environment. When 2 we have multi-site trials, we need to be thinking and 3 working more on a common understanding, cooperative 4 review mechanisms. International research has 5 introduced a whole new level of complexity into these 6 cooperative reviews. 7 I'll throw this out just as devil's 8 advocate. Ada Sue already beat me to it, but I have 9 often heard it said that we have inappropriately 10 forced social and behavioral research into a 11 biomedical model. I will throw a shot across the bow 12 and say no, I don't think we have inappropriately 13 forced it in. I think if you look back at the charge 14 of the National and the First Presidential Commission, 15 which had behavioral right in its name, they were told 16 to identify the basic principles, which should 17 underlie the conduct of biomedical and behavioral 18 research, and I think they did a pretty darn good job. 19 I will agree again with Ada Sue that there may well be 20 misapplications, but I don't think it's Subpart A 21 that's at fault for this. 22 Preaching to the choir and the need for 46 1 more education, and OHRP and others have already done 2 a lot of work in this regard. I guess I put this up 3 mainly for the sake of that last item or the last 4 target audience, the funding agencies, and Ada Sue 5 also touched on this. I think if I had just stayed 6 fourth in line, all of these guys would make all my 7 points and I could have made this even shorter, but we 8 haven't compared notes beforehand, so it's going to be 9 interesting to see how this all ties together. 10 I agree, too, there is some discord 11 between the regulatory agencies' interpretations, IRB 12 interpretations and the interpretations of funding 13 agencies. Program officers at NIH are, unfortunately, 14 some of the least informed people I have run into in 15 terms of applying the regs and integrating them and we 16 have many tugs of war not with OHRP or with other IRBs 17 in applying and interpreting, but with funding 18 agencies. 19 So I will close the general observations 20 with a few rhetorical questions. Can compliance and 21 common sense coexist? Is our overemphasis on dotting 22 the Is and crossing the Ts undermining the underlying 47 1 intent of the regs? I like the regs. They are my 2 friend. I guess that's one of the reasons I'm here, 3 but I do fear that we have swung the pendulum so far 4 in the direction of slavish attention to dotting Is 5 and crossing Ts that we're in danger at least of 6 starting to undermine the intent that they are there 7 to serve. 8 Are we forcing institutions to take the 9 easy way out? For example, I hear more and more 10 people talking about limiting their FWA to apply only 11 where federal funding really obligates them to do 12 that, so that they don't have to get caught up in some 13 of the dotting Is and crossing Ts when push may come 14 to shove. 15 So shifting to some specific problem 16 spots, and I will cover these in the order of 17 numerical citation in Subpart A. Exemptions come 18 first, 101. You know the regulatory citation. These 19 are research activities in which the only involvement 20 of human subjects will be in one or more of the 21 following categories. These can be exempt, but I 22 think there is a real catch-22 in the area of 48 1 exemptions. They are technically at least exempt from 2 the policy and that includes IRB review, but of course 3 we all review that to make that determination, so 4 there's a circular argument there, and by the time you 5 make that determination, you're really talking about 6 exempt from continuing review, which is okay. 7 These activities are technically eligible 8 for exemption, yet the issues are still there, the 9 issues we all care about, informed consent, 10 confidentiality, overall protection of subjects, 11 continuing review of amendments that may come up. 12 Well, if you have told somebody to, essentially, go 13 away with an exemption, they may or may not be coming 14 back to you when they are making amendments that may 15 shift it out of an exempt category. 16 I think probably the underlying intent is 17 good. We still exempt several hundred studies a year, 18 so it's not like we have gotten rid of this in our own 19 setting. On the other hand, it creates kind of 20 uncomfortable mishmash, what I would say is an awkward 21 no-man's land somewhere between not human subjects 22 research and what really constitutes expedited review. 49 1 David already touched on this, internal 2 discord. I don't necessarily view it as hubris, but 3 I do see it as a technical discrepancy within the regs 4 themselves at least as I read it. The exemptions cite 5 demonstration projects that are conducted by or are 6 subject to the approval of department or agency heads 7 and I have inserted the word federal there, because 8 that's my understanding of conventional wisdom and I 9 believe that's backed up by OHRP guidance. 10 However, if you look a bit further on, as 11 David already noted, under the waiver of consent 12 criteria, there it specifically states a demonstration 13 project conducted by or subject to the approval of 14 state or local Government officials, and I'm not sure 15 people quite know how to apply that. I'm not sure 16 this is a big ticket item in terms of the work today 17 world. We don't see a lot of demonstration projects, 18 but there is some internal discrepancy there. 19 This is a much bigger deal. You have got 20 a whole Panel this afternoon. I'm glad to see that. 21 The definitions of research, systematic investigation, 22 the application of that and the definition of human 50 1 subject. The problem is, as you well know, that there 2 are many activities that are not quite human subjects 3 research, but carry the same risk for people who are 4 not quite human subjects. We have lots of student 5 projects. 6 It might be interesting to debate at least 7 at a philosophical level the difference between 8 information that is generalizable versus that which 9 will be generalized. You can have a student going out 10 to do the same kind of project that carries the same 11 risk for the person who is not quite a human subject 12 as defined. Why should we be handling that or do we 13 handle it differently than a faculty member who is 14 going to publish that data, therefore, generalizing? 15 You can see where I'm going with that. 16 Studies of organizations. Any unit of 17 interest larger than one individual you can 18 technically rule is not human subjects research. We 19 don't have individual human subjects but, again, we 20 can have the same risks for people involved in that 21 research and the whole area of case studies, an n of 22 1, but the same, certainly, confidentiality issues 51 1 even if not as generalizable as a larger controlled 2 study. 3 I won't spend a lot of time on this. The 4 minimal risk definition, I will just say and I think 5 others have already said it, that many IRBs have 6 trouble applying this. Don't ask me how to fix it, 7 but I can tell you it's a problem in the work today 8 world for IRBs to make this determination fitting in 9 with this current existing definition. 10 This is a much bigger problem, Ada Sue 11 touched on this, the assurance mechanism. I think, 12 again, tying it back to the general observation I have 13 already made about the modern research collaborative 14 environment for multi-site collaborative studies, I 15 think IRBs need guidance and I would say, in fact, 16 encouragement to ask them at least, not to force them, 17 to take a look at ways to share in the review process. 18 I'm not sure that 300 individual IRBs reviewing a 300 19 site study necessarily contributes to the global 20 protection of subjects if that's what we're all about, 21 but yet we all do that. Legal issues, institutional 22 interests force us in that direction. 52 1 I think we need a lot more work and 2 thought and I, again, join Ada Sue in thanking you for 3 the work OHRP has already done toward improving the 4 assurance process, but I think still more needs to be 5 done. The notion of a designated IRB within that FWA, 6 we have simply untenable situations that we're being 7 forced into or guided into in the setting of 8 cooperative, collaborative multi-site studies. IRB 9 Authorization Agreements, unaffiliated investigator 10 agreements. When do you use these? How do you use 11 them? These are all issues. 12 Written procedures. This one is not 13 necessarily a big ticket item, but I will throw it in 14 there just as a small problem spot. The regulations 15 say that we should have written procedures for 16 determining which projects need verification from 17 sources other than investigators that no material 18 changes have occurred since previous review. 19 Perhaps just as the devil's advocate I 20 will ask how often does this really happen? How many 21 IRBs are even aware of this requirement? I view this 22 as kind of one of the trick questions in the regs. If 53 1 you want to ding somebody, go out and survey this 2 audience or go out into the 3,000, 5,000, 10,000 IRBs 3 out there and see how many actually even have the 4 procedures that they need to satisfy the regs, much 5 less are actually applying it. 6 IRB membership, so there's a couple 7 sections in there. 107, as you well know, each IRB 8 shall include at least one member whose primary 9 concern is a nonscientific area and also one who is 10 not otherwise affiliated with the institution. 11 The whole notion of community 12 representation, this has been discussed by many 13 commissions, perhaps by you already, but I would say 14 that the n of 1, that's the latter of the regulation, 15 doesn't necessarily jive with the n of 1 of 5, the 16 notion of a 20 percent representation and I would stay 17 away from setting a percent on that, but I think when 18 IRBs get to have 20, 25, 30 members as many do these 19 days to provide the breadth of expertise, that 20 community voice can tend to get lost and I would urge 21 just to at least be aware of that and I think many 22 IRBs do. Out of our 14 members, we try to have two or 54 1 three who are from outside the institution or a non- 2 scientist, but I don't know that that necessarily 3 translates into a consistent approach across IRBs. 4 This is a favorite whipping boy of ours, 5 convene meeting review in which contingencies are 6 required to go back to a full board. So there are a 7 couple sections that command us to review research at 8 a convened meeting, except when expedited review is 9 used and that we should notify investigators of 10 modifications required to secure IRB approval. There 11 is a problem in returning minor contingencies to a 12 convened meeting. I will give you an example. If the 13 convened group looks at a consent form and says well, 14 the purpose statement is a little too complex, they 15 need to simplify that or clarify that. 16 We have been instructed and advised that 17 we need to have more specific stipulations as to 18 exactly the verbatim language that an IRB would be 19 expecting from investigators. Otherwise, it has to 20 come back to the full board. I will say this as a 21 source of regulatory burden. When you have whole 22 studies and whole consent forms that can be handled 55 1 through expedited review, why is this one needing to 2 go back with IRBs that may be seeing 20 or 30 3 protocols at a meeting to bring back things that truly 4 are minor contingencies, I think, is something I would 5 like to talk about more during the Q&A if we have a 6 chance. 7 Expedited review, I agree with others. In 8 general, I think this works pretty well. There is a 9 small point of confusion in here that I see more and 10 more IRBs at least scratching their heads about. The 11 regulations specify that reviewers may not disapprove 12 the research, but in the next sentence, a research 13 activity may be disapproved only after review in 14 accordance with non-expedited procedures. 15 I see some confusion in applying minor 16 amendments. Does that research or research activity 17 mean the study as a whole that can't be disapproved or 18 if somebody comes up with a hairball amendment to 19 recruit subjects with an ad that the chair can already 20 see needs to be thrown out, can they disapprove that? 21 Does that need to be taken back to the full board, so 22 they can say no, that's a hairball amendment. Let's 56 1 not accept it. That must be a southern phrase, Ada 2 Sue, hairball amendment. I think there is some minor 3 confusion in there, not a big point. 4 Waiver of consent. I like the waiver of 5 consent. We use it every day. I think the practical 6 section is problematic in its application, and I would 7 say that some interpret this waiver as applying only 8 to retrospective reviews. It creates a dilemma where 9 some data exist where you, basically, have a mix of 10 retrospective and prospective data being gathered. 11 They look just identical. 12 I think we're at risk of providing 13 incentive at least for PIs to gain the system in a way 14 or for IRBs to help them gain it. In other words, if 15 you're finding yourself in a situation where you're 16 going to say well, gee, if you were just coming to me 17 a year from now and that data was already on the 18 shelf, I would be waiving this lock, stock and barrel, 19 but now we have got this odd mix of prospective and 20 retrospective, I can waive some, I can't waive the 21 other. I think IRBs can find their way through it, 22 but I think some guidance and thought might be 57 1 helpful. 2 Another element in waiver of consent, 3 whenever appropriate, the subjects will be provided 4 with additional pertinent information after 5 participation. Perhaps turnabout is fair play for the 6 social-behavioral folks. I think this is a behavioral 7 research criterion that was really put in originally 8 to address issues of deception in research or 9 withholding of certain information. I think we often 10 have trouble applying this in the biomedical setting 11 to the retrospective chart reviews, the use of 12 existing specimens. What does it really mean to be 13 coming back to somebody with additional pertinent 14 information? Well, we interpret that and apply it as 15 coming back with clinical information that might be 16 relevant to their individual medical situation. That 17 rarely happens, truth be told. This doesn't 18 necessarily fit. 19 The last slide, and I guess this probably 20 could just as well have been grouped up with the 21 general observations, because I don't know that it 22 ties to a specific citation as much as the overall 58 1 focus, the overall audience for the regulations. I 2 just borrowed this slide from Jeff Cooper who nicely 3 diagramed the sort of patchwork approach that you all 4 recognize. I think we do have some real issues with 5 harmonization across all the different groups and, of 6 course, a whole area of unregulated research. You and 7 others are discussing that and I encourage further 8 looks in that regard. Thank you for your attention. 9 (Applause) 10 CHAIR PRENTICE: Thank you very much, Dan. 11 You may have had the fewest slides, but you had the 12 longest time, but you're still on time, because it was 13 19:30, so thank you for keeping it on schedule. Our 14 next speaker is Steve Peckman and Steve is the 15 Associate Director for Human Subjects Research at the 16 University of California, Los Angeles, UCLA, and I 17 won't go into his bio any further. Welcome to SACHRP, 18 Steve. 19 MR. PECKMAN: Thank you very much for 20 inviting me to speak this morning. I hope to use 21 David's five minutes and Ada Sue's 15 seconds to add 22 on to my time. I was a little shocked to be asked to 59 1 come and speak to this issue just because to discuss 2 whether Subpart A needs modification or not to me at 3 least was a fairly overwhelming task. I haven't been 4 working with Subpart A as long as Ada Sue has, but 5 long enough to know that when I started I didn't have 6 any gray in my beard. 7 I think that Subpart A actually is written 8 really well. Every time I have spoken with Charlie 9 McCarthy about it, like David, I spoke with him about 10 it, I'm impressed at the short amount of time they 11 spent writing it and its enduring quality, and I think 12 what we're going to see this morning and hear this 13 morning is that we all think that flexibility is the 14 key in terms of using Subpart A and probably the most 15 important thing is to -- wow, that was different, is 16 to look at how we can get more guidance in applying 17 Subpart A. 18 So is Subpart A in need of modification? 19 I'm going to talk about the purpose of the regulations 20 as I see it, responsibilities, obligations and 21 institutional culture. I think this is probably going 22 to be the most important part of what I'm going to 60 1 contribute this morning in that we haven't heard 2 anything about the institutional culture and the 3 responsibility of the institution. 4 And in reality, if you read Subpart A very 5 carefully, it describes the responsibilities of two 6 entities, the institution and by extension of the 7 institution, the IRB. Nowhere really in the Subpart 8 A does it explicitly outline the responsibilities of 9 the principal investigator like it does in many 10 aspects of the FDA regulations, and we're not going to 11 pick on FDA this morning, but I hope to attend another 12 meeting where that occurs. 13 Flexibility. The biomedical lens and the 14 ability to define and address all human research. 15 Ernie didn't say this in my introduction, but I'm 16 responsible for three IRBs at UCLA, three biomedical 17 and two social-behavioral. And if I didn't address 18 the issues regarding social-behavioral research, I 19 would be strung up when I returned to Los Angeles. 20 And specifically, I'm the administrator where the 21 rubber meets the administrator at UCLA. And I can 22 tell you that social-behavior researchers, if you 61 1 haven't heard from them already, are really upset 2 about some of the wording in Subpart A and how it 3 inhibits their ability to do research, and I'm going 4 to speak specifically to some of those elements. And 5 then furthermore, how some of the wording and 6 interpretation seems to reflect the biomedical model 7 as you have heard earlier. And then finally, the 8 difference between guidance and regulation. 9 So the purpose of the regulations. Well, 10 the purpose of the regulations, as I see it, is to 11 define in a general outline institutional and IRB 12 responsibilities and obligations. If you read Subpart 13 A carefully, that's all it keeps talking about. So it 14 also talks about the administrative process and 15 documentation requirements. And we have heard a 16 little bit about that, whether these administrative 17 processes and documentation requirements are actually 18 a burden to conducting review of research and the 19 research itself, because the extension of creating a 20 burden on IRB and IRB staff is the biggest concern, is 21 then it delays the implementation of research that is 22 ultimately approved. And actually, in the biomedical 62 1 world, that actually delays the ability of subjects to 2 be accrued just for research and in treatment 3 research, it delays the ability of subject patients to 4 access what wouldn't be normally available treatments. 5 It's a template for addressing ethical 6 issues. The Federal Regulations in Subpart A offer 7 the IRB and the institution a template in which to 8 review research and to address the ethical issues that 9 come before it, and this is where I think the 10 documentation aspects of Subpart A are very important 11 in that what it does is it requires the IRB to step 12 back and say these are the things we must do in order 13 to carry out our obligations as a board that is 14 reviewing the ethics of the conduct of research and if 15 we don't check these off on our list, then we're not 16 actually doing our job. I think that's a pretty sound 17 idea to say let's give them a checklist to review and 18 make sure they document it. 19 And finally, the regulations are a means 20 to an end and not an end in itself, and this has been 21 repeated three times already this morning, which is 22 the regulations are there to help us ensure that 63 1 ethical research is conducted in that manner, and 2 that's why we have the checklist, and though through 3 the OHRP letters that we have heard a bit about this 4 morning, and I will talk a little bit about it as 5 well, we have been instructed on how to use this 6 document as a means to our end. 7 Now, the question becomes whether letters 8 to institutions is an appropriate guidance to the 9 research community and to IRBs as to how to do their 10 job, but ultimately we have to recognize that this is 11 a means to an end to ensure the ethical conduct of 12 research and to ensure the ethical conduct of the 13 institution who is hosting the research. 14 So the responsibilities, obligations and 15 institutional culture, which I feel is the most 16 important part of ensuring an ethical program for the 17 conduct of human research. And this is outlined 18 briefly in a few sections in the document. In 103(a) 19 we talk about the institutional assurance of 20 compliance, which I'm going to go out on a limb this 21 morning as the outlier, I think this is a really 22 important document. This holds the institution's feet 64 1 to the fire. This outlines what the institution's 2 obligations and responsibilities are if they want to 3 be involved in human subjects research. Whether it's 4 an adequate document to that task or not is a 5 different question. 6 103(a) also defines who the institutional 7 officials should be, and I'm going to spend a little 8 bit of time talking about that. 103(b)(2), the 9 responsibility of the institution to support the IRB. 10 And if any of you ever monitor or participate in 11 MCWIRB, now known as the IRB Forum, you will find that 12 at least once a month there are several inquiries 13 about whether there is adequate institutional support 14 of their IRB and what are other institutions doing to 15 support their IRB. 16 And finally, what happens if the 17 institution doesn't discharge its responsibilities 18 effectively? Well, in the little discussed 123 there 19 is an outline of material failure to discharge 20 responsibilities as an institution and what the 21 Federal Government can do to you as a result. 22 So what about these obligations? Without 65 1 an appropriate institutional ethos, regulations and 2 guidance will not successfully protect human subjects 3 period. If the institution that is responsible for 4 the research doesn't have an ethical ethos in how they 5 are going to carry out their charge of responsibility 6 of protection of human subjects, it won't matter how 7 many regulations we have, how much guidance we have 8 and how it's interpreted. It's just not going to 9 happen. 10 So institutional support of the IRB in 11 their decisions builds an ethical institutional 12 culture, and that is what we need to look to in terms 13 of OHRP, in terms of guidance to carry out this charge 14 articulated in 103. What are the agencies' 15 expectations of the institutional official regarding 16 the building of this institutional culture? What is 17 the required definition of an acceptable qualification 18 for an institutional official? Who should that person 19 be in the chain of command? How should they carry out 20 their authority? 21 In 107(e) it describes conflict of 22 interest requirements for IRB members, but there is no 66 1 guidance in the regulation on how conflict of interest 2 requirements should be thought of in terms of an 3 institutional official. Shouldn't the person who is 4 responsible for bringing in, whose sole responsibility 5 at the institution is for bringing in research dollars 6 be the institutional official? If we ask ourselves 7 that question, then we're actually questioning the 8 majority of institutional officials in the country. 9 103(b)(2), institution must constitute an 10 IRB with provisions for meeting space and sufficient 11 support staff to support the IRB's review and record 12 keeping duties. What constitutes sufficient and how 13 should this be assessed? We know from at least one 14 citation letter from OHRP they claim that there was 15 insufficient support of the IRB given by the 16 institution. How was this determined? 300 protocols 17 per staff person? How do you count the protocols? Do 18 adverse events count as a submission? These are the 19 kinds of issues that institutions need as guidance to 20 help them understand what constitutes sufficient 21 support. 22 IRB composition. We have heard a little 67 1 bit about IRB composition. I'm going to spend a 2 little bit more time on this. 46.107 says you have to 3 have a minimum of five members. At most academic 4 medical centers, you will find that their IRBs consist 5 of a lot more than five members, at least one 6 scientist who is an expert in areas of concern, at 7 least one member not affiliated with the institution, 8 at least one non-scientist. So what does this say? 9 IRBs are not obligated to privilege scientists over 10 non-scientists on the IRB. We're going to get to 11 statistics in a minute. Local review and sensitivity 12 to community attitudes are also required by the 13 regulations. 14 So what is the intent of the National 15 Commission? Well, the Commission believes that the 16 rights of subjects should be protected by local review 17 committees operating pursuant to federal regulations 18 and located in institutions where research involving 19 human subjects is conducted. As we have heard 20 already, most of the research now is not conducted at 21 just one institution. It's multi-site institutions in 22 both social-behavioral and biomedical research. 68 1 Compared to the possible alternatives of 2 a regional or national review process, local 3 committees have the advantage of greater familiarity 4 with the actual conditions surrounding the conduct of 5 the research. The National Commission went on to 6 state that you must balance scientific, individual and 7 community concerns on IRBs. You need to guard against 8 scientific self-interest and demonstrate an awareness 9 and appreciation of the various qualities, values and 10 needs of the diverse elements of the community served 11 by the institution, and you have to have a diverse 12 membership. 13 So how does this work out in real life? 14 Well, what's the face of an IRB? The 1995 survey 66 15 percent male, 90 percent white, non-Hispanic. Follow- 16 up survey in '98, little change in the composition of 17 IRB membership, 92 percent white, 58 percent male 18 institutionally affiliated members comprised 83 19 percent of IRB membership. Most non-affiliated 20 members are non-scientists and account for 10 percent 21 of total members. The remainder are institutionally- 22 based non-scientists, such as lawyers, contracting 69 1 grant officers, etcetera. So these are our IRBs that 2 are representing community values. 3 NBAC advised and strongly suggested that 4 IRBs should include members who represent the 5 perspective of the subjects unaffiliated with the 6 institution whose primary concern is in non-scientific 7 areas. It should collectively represent at least 25 8 percent of membership and should be represented at 9 each meeting of the IRB. This way we engage the 10 community of potential subjects. We give the 11 community a voice in the research that is done in this 12 community. And we democratize the process of IRB 13 review. We take it away from the scientific self- 14 interest of the Peer Review Committee and create what 15 was intended for IRBs, which was to be an IRB that 16 would appreciate community values and deliberate on 17 those. 18 So flexibility. I'm going to talk a bit 19 about flexibility. The biomedical lens or the 20 biomedical model and the ability to define and address 21 all human research. So we're going to talk about 22 refining definitions of key elements. We've talked 70 1 about the definition for research, human subject, 2 minimal risk, informed consent, expedited review and 3 I'm going to quickly run through some of these again, 4 because I think I bring a little different spin to 5 them. 6 102(d) research, means a systematic 7 investigation, including research development, testing 8 and evaluation and designed to develop or contribute 9 to generalized knowledge. I hate this definition. I 10 have to say that of all the things in the regulations, 11 this is the most poorly written part of it from my 12 point of view. Because it doesn't make any sense to 13 anyone who read it, who is knowledgeable of the 14 regulations, because what is a systematic 15 investigation? 16 It has never been defined. What is 17 generalizable knowledge? As Dan just said, it's a 18 student researcher whose research may not be intended 19 to be generalized upon initiation, but actually then 20 becomes generalizable knowledge. As the professor 21 says, gosh, you came up with good data. I'm going to 22 use it in my next grant proposal and my next journal 71 1 article. What is generalizable knowledge? Is a case 2 study generalizable knowledge? Why else would you 3 publish it? 4 Yet, it seems to be in the common 5 vernacular that case studies are not under IRB purview 6 nor would anyone want to make case studies under IRB 7 purview. So this is an area where guidance would be 8 very beneficial. 9 Who is a human subject? A living 10 individual about whom? I would pause at the clause 11 about whom is the core of this definition and I have 12 no idea what "about whom" means. The regulations 13 define intervention. They define interaction and 14 private, identifiable information. When is research 15 about a person? Opinions about events external to the 16 person? Is that about the person or are they about 17 the events that are external to the person? 18 Opinions about a work place. If you ask 19 me what it's like to work at UCLA, is that research 20 about me or is it research about UCLA? I won't answer 21 that question by the way. Are objective responses 22 defining who a human subject might be or subjective 72 1 responses? Where do we draw these lines? 2 When does a teacher become a human subject 3 in a school survey study? So you're going to go out 4 and you're going to interview a bunch of teachers and 5 give them a survey. When they answer questions about 6 how many students in the class, what's the average 7 GPA, how many people are tardy per day, how many 8 people miss class, is that about the teacher, is it 9 about the students, is it about the school? Who are 10 the human subjects? Are there any human subjects in 11 that survey? Remember, it's the "about whom" clause 12 that defines whether you have a subject, not the 13 intervention or interaction or the identifiable 14 private information. It's the about whom. 15 Let's say they start to ask how well do 16 your students perform after you do X, Y, Z? If you 17 lecture in the left hand part of the classroom as 18 opposed to the right hand part of the classroom, are 19 the students more responsive? Do they sit more on the 20 right hand side if you lecture to the right hand side 21 of the classroom? Is that about the teacher? Is that 22 about the students? 73 1 Private information. We talked a little 2 bit about this, Virginia Commonwealth was brought up. 3 I'm just going to bring up the questions, in case 4 you're not familiar with the case. I think it's an 5 incredible case. It actually focused IRBs around the 6 country on the about whom clause of the definition of 7 the human subject. Now, whether this was the right 8 case to do it or not, I don't know. And there are a 9 lot of curiosity seekers in the GeneX community who 10 are concerned about this. 11 But these were the questions outlined in 12 the OHRP letter that were asked of the subject about 13 the parents. Is the identifiable information about 14 the adult parent? Is the parent a human subject? 15 Should the IRB be deliberating about the about whom in 16 this case? Again, the Virginia Commonwealth issue is 17 an important one in terms of guidance. This is where 18 guidance is happening through a letter to an 19 institution. The guidance is happening to all 20 institutions rather than through guidance from the 21 agency itself. 22 Exempt from IRB review. What constitutes 74 1 normal educational practice? This drives everyone 2 nuts at my institution. The Department of Education 3 is great, but OHRP needs to collaborate with the 4 Department of Education to come up with definitions. 5 Is randomization among instructional techniques exempt 6 from IRB review? It would appear upon a first glance 7 of the regulation for exemption that it could be. Is 8 it specific to the school of the normal educational 9 practice? The school district? The USA? The globe, 10 we call Earth? 11 What constitutes public behavior in 12 exemption category (b)(2)? What defines publicly 13 available and to whom in category (b)(4)? Assessing 14 risk, physical, social, psychological, economic, there 15 are different types of potential risks and harms that 16 may require different protections. 17 Minimal risk. We determine the level of 18 review. This is a really important definition within 19 the regulations, because it's going to determine 20 everything that the IRB does. If it's minimal risk 21 and fits a certain category, it does through expedited 22 review. If it's more than minimal risk, it goes to 75 1 full committee review. It determines requirements for 2 informed consent and it determines documentation 3 requirements. 4 So probability and magnitude ordinarily 5 encountered in daily life are my two favorite phrases 6 here. We have heard about probability and magnitude, 7 but I'm going to talk about what does "ordinarily 8 encountered" mean? By whom? Whose daily life? I 9 traverse 14 miles every day back and forth across the 10 city of Los Angeles coming and going from work. Is 11 my daily life risk that I daily encounter different 12 than yours? I bet it is for a lot of you. Okay. I 13 go through certain sections of LA that a lot of people 14 probably don't want to go through, and I don't mean 15 Beverly Hills. 16 Informed consent, 117, the documentation 17 requirements. Informed consent shall be documented. 18 A written consent form must be approved by the IRB and 19 signed by the subject and a copy shall be given to the 20 person signing the form, shall be written consent 21 forms signed by the subject. How does a signed 22 document increase protection of the subject? Someone 76 1 needs to answer that question for me sometime during 2 my life working with IRBs. 3 It appears to me it's the protection of 4 the institution and the investigator. Yet, all the 5 other regulations are about the protection of the 6 subject. What is sacred about a signed consent 7 document when such a signature is not required by law? 8 And should there be a difference between social- 9 behavioral and biomedical research in the application 10 and guidance given for 117? Well, this is where we 11 really need some flexibility and guidance for the 12 requirements at home and abroad. 13 Who are our subjects? What are the basic 14 demographic assumptions at home? Who is majority and 15 who is minority in our communities? And here is the 16 case study, California, the place I'm most familiar 17 with, in the 1940s 89.9 percent Euro-American 18 population. David Hayes-Bautista showed in 1992 that 19 we need to question these assumptions, because in 20 California in the 1990, rapid and unoppressive 21 demographic changes challenged our assumptions, 22 because Euro-Americans are now the minority in the 77 1 State of California. 2 Similar trends exist in Texas, New York, 3 Arizona and New Mexico, Florida and Chicago area and 4 changing demographic trends challenge our assumptions 5 of minority versus majority, as well as homogenous 6 Euro-American value systems that may not be applicable 7 to communities of color. In fact, it may intimidate 8 people and force people out of research. So our case 9 study in California, 224 languages spoken in my state. 10 40 percent of the LA County residents are born in 11 another country. This is my home. 12 And the social/political context of 13 signing consent documents is really, really, really 14 different, because people come from communities and 15 cultures where signing documents aren't such a 16 profound noble thing as they are in our Euro-American 17 tradition. So 117 says we can waive signing form 18 consent when it is the only record linking the subject 19 and the research would be -- and the subject and the 20 research would be the consent document and the 21 principal risk would be potential harm resulting from 22 a breach of confidentiality. 78 1 There is flexibility necessary when other 2 methods are available to minimize risk and insure 3 confidentiality. Each subject will be asked whether 4 the subject wants documentation linking the subject 5 with the research and the subject's wishes will 6 govern. If the IRB finds that a signed document may 7 place the subject at risk as the unnoted, why would we 8 want to subject or give them the opportunity to place 9 themselves at risk? 10 If you're going into a crack house to do 11 interviews with crack addicts, you're going to ask 12 them if they want to sign a consent document that 13 identifies them as a crack addict who is being 14 interviewed about crack use? I mean, this just 15 doesn't make any sense. We need to ensure equity and 16 justice with human subjects. We need to recognize the 17 changing demographic and non-western, non-legalistic 18 approach to communication with people who are our 19 subjects. We need to acknowledge international nature 20 of research, both with drug research, device research 21 and social-behavioral research. 22 We need to acknowledge and address the 79 1 historical basis for the negative connotations many 2 people bring to signing documents. And as NBAC noted, 3 flexibility with respect to the documentation process, 4 especially when individuals can easily refuse or 5 discontinue participation would be appropriate. 6 Altering the required elements, you've seen what the 7 requirements are to alter or to waive the requirements 8 for informed consent. 9 So what if you want to do interviews with 10 nomadic people in the Himalayas? They don't have 11 telephones. They don't have electricity. They don't 12 have mail. But yet 116(a)(7) says you have to explain 13 to them who to contact for answers pertinent to 14 questions about the research and research subject's 15 rights and whom to contact in the event of a research 16 related injury of the subject. Does this information 17 have any worth to these people for participating in an 18 interview project about nomadic tribes in the 19 Himalayas? 20 This is the kind of stuff that drives 21 social-behavioral researchers insane. This is the 22 biomedical model that they complain about. This is 80 1 nonsense to people who live in the Himalayas who don't 2 have access to any of the tools that we may have here 3 in the Washington area. Research related injury. I'm 4 going to have to stop for a second and talk about 5 biomedicine. 6 Biomedical research of more than minimal 7 risk. 116(a)(6) says you are required to have an 8 explanation as to whether any compensation is 9 available to the subject should medical injury occur. 10 Now, this allows subjects to participate in more than 11 minimal risk biomedical research at their own risk and 12 relinquishes the sponsor of the institution and the 13 investigator of any responsibility to provide care for 14 injury occurring as a result of research procedures. 15 How are we protecting human subjects here? 16 This is one of the larger loopholes in the 17 regulations. So I could participate in a Phase One 18 trial where the purpose of the study is to find a 19 maximum tolerated dose of a drug, so they're going to 20 keep dosing people until someone gets really sick or 21 dies, and the institution, the sponsor and the 22 investigator can say we're not going to treat you. We 81 1 have no obligation to treat you. You participated in 2 this research at your own risk. 3 Expedited review, I'm just going to talk 4 about minor change. What constitutes a minor change 5 in already approved research? You're doing a Phase 6 One drug trial. 10 subjects have been approved to be 7 enrolled. The 10 subjects did really well. We all 8 have positive responses. You want to add three more 9 subjects to your Phase One trial. Is that a minor 10 change? 11 The short form. A lot has been said about 12 the short form so far. The allowed use of a short 13 form, typically used for non-English speaking subjects 14 in biomedical research. I can assure you that I have 15 not seen anyone use the short form in social- 16 behavioral research. It would be considered 17 incredibly ugly towards the subject to give them a 18 short form with an English translation. Are there 19 safety concerns with the short form? There is no 20 restriction to minimal risk research using the short 21 form. 22 Should we acknowledge the importance of 82 1 justice, but should use be limited to clinical trials 2 with therapeutic intent? Do we want to use a short 3 form for a Phase One trial? Does it promote two 4 classes of research subjects? And this was brought up 5 earlier and alluded to. Does the short form allow us 6 to have non-English speaking subjects who get a short 7 form with a full English language translation of full 8 English language consent document all the risks, 9 alternatives, etcetera and leave these non-English 10 speaking subjects with only a short form with the 11 rules outlined for them? 12 Alternative strategies. We need guidance. 13 Should the short form be limited to screening and 14 initiation of subjects followed by required complete 15 translated inform consent document? Is paper the only 16 acceptable standard? What about guidance for 17 regulations that provide for the use of new 18 methodologies and flexible strategies, such as audio 19 or videotaping the consent process and providing 20 subjects with a copy for reference as their document? 21 I can assure you that even in the most 22 poor sections of LA, everyone has a television set and 83 1 a video-player, but not everyone can read. 2 Suspension or termination of IRB approval. 3 "Grants IRB the authority to suspend or terminate 4 research." Shouldn't IRB have the authority to 5 suspend an investigator? We don't want them to 6 terminate an investigator, of course. But should they 7 be allowed to suspend an investigator? Currently, the 8 regulations don't provide for that. And shouldn't 9 institutional assurance be required to have policies 10 for such an action? 11 You can have a trial where you have 100 12 subjects on the trial. It's a biomedical trial and 13 the investigator has gone way out. But the subjects 14 are doing well. You want us to suspend that research? 15 Who has the authority to remove the investigator? By 16 regulation, not the IRB. 17 Central versus local IRB review. You're 18 going to talk about this this afternoon. 114 talks 19 about cooperative research. Was the intent of 114 to 20 share review among institutions within the same 21 geographic area if we go back to the National 22 Commission idea of what an IRB is or to create central 84 1 IRBs? The central IRB review undermined the local 2 community participation as contemplated by the 3 National Commission and the regulation. 4 Here is my laundry list and now social- 5 behavioral scientific-regulatory expertise in the 6 compliance section of OHRP. This is a necessity not 7 nationwide. Social-behavioral scientists need someone 8 to go to at OHRP who has social-behavioral expertise 9 in the compliance division, so that we can get a 10 guidance that is specific to that type of research 11 without at least the appearance of a biomedical model 12 being used to interpret the regulations. 13 Guidance. Level of detail and written 14 policies and procedures. You've heard about level of 15 detail required for IRB minutes. Level of review for 16 safety reports and adverse event reports. Adverse 17 events, of course, are not even contemplated in 18 Subpart A and standards for scientific review are 19 necessary for guidance. So the efficacy of IRB review 20 is predicated on improved federal guidance, from my 21 point of view. 22 The role of the institution and the 85 1 institutional official responsibility requires more 2 than compliance with the letter of the regulation, as 3 we have already heard. It also requires a willingness 4 to apply the ethical principles that are the spirit of 5 the regulations to create an institutional ethos that 6 governs the actions of all stakeholders and the 7 protection of human subjects. We need more guidance 8 on the composition of IRBs, including the inclusion of 9 community, flexibility of the regulations for specific 10 research paradigms, such as qualitative research. 11 Thank you very much. 12 (Applause) 13 CHAIR PRENTICE: Thank you, Steve. Would 14 the Panelists, please, assemble up at the table? 15 While you are doing that, I'm looking at the schedule. 16 Do we have anybody that's -- okay. I would like to 17 find out whether or not there are individuals here who 18 want to take advantage of the public comment period 19 and address SACHRP? No one has registered. And if no 20 one indicates an indication that they wish to address 21 SACHRP, I'm going to adjust the schedule accordingly. 22 In other words, I'm going to eliminate the public 86 1 comment section. 2 Okay. That will give us more time for 3 discussion. We do have somebody? 4 AUDIENCE MEMBER: Depending on what 5 happens during the question and answer. 6 CHAIR PRENTICE: Okay. Then I'll poll the 7 audience again closer to the break. Then we can 8 decide at that point in time. Okay. I always take 9 the Chair's prerogative of asking the first couple of 10 questions. Many of you had the same concerns, which 11 is, I guess, not surprising. Some of you emphasized 12 concerns to a greater extent than others. One of the 13 things that I found personally to be a real problem in 14 terms of creating what I call uncalled for regulatory 15 burden, and, Dan, you talked about this, and that is 16 when an IRB has reviewed a protocol and they have a 17 series of, let's say, clarifications or modifications 18 to the consent document or the protocol itself, in 19 other words usually it's a request for review form 20 asking more information about the recruitment process 21 or what have you. 22 It's currently the position, as I 87 1 understand it, the position of OHRP that unless the 2 clarification or modification is of a very, very minor 3 nature it has got to be taken back to the full IRB. 4 In consideration of this particular position, we end 5 up at my IRB tabling oh, probably about 40 percent of 6 our protocols, 40 percent. Quite frankly, I don't 7 believe that adds any value to human subjects 8 protection. It simply increases the workload of the 9 IRB members. It delays the initiation of research and 10 I am very comfortable with having the IRB chair or 11 even a subcommittee look at that and say, you know, 12 they responded. We don't need to take this back to 13 the full board. 14 And one of the problems, of course, is 15 that AAHRPP looks at OHRP standards and positions and 16 now, they are going out and looking at what IRBs are 17 doing. So we keep on tabling more and more protocols. 18 It's a catch-22 situation. The more detailed your 19 review, the more likely it is that you're going to 20 have changes, right? The less detailed your review, 21 the less likely it is you're going to have changes. 22 So I consider that to be a big problem, 88 1 but you're really the only one that emphasized that. 2 No one else seemed to have emphasized that and I found 3 that curious. So I guess that would be my first 4 question to the other three where it didn't seem to be 5 a problem at your institutions. Is it not a problem? 6 MS. SELWITZ: I would say that's an 7 excellent point. I think it is a problem. The reason 8 I didn't include it was that I was limited to 20 9 minutes and also, quite frankly, not just the time 10 limit, I was trying to focus on where I thought the 11 problems were perhaps in Subpart A itself. And I 12 don't think that's where the problem is. I think the 13 problem is how Subpart A is being interpreted, and I 14 do think minor should not be interpreted as narrowly. 15 And I think, again, the problem is that 16 the interpretation shifts from time to time given 17 other agencies, David DePay, as well on how it's 18 interpreted in terms of minor. But I agree with you 19 100 percent and I just didn't include it because of 20 time, and I don't think the problem is Subpart A. I 21 think it's the interpretation. 22 CHAIR PRENTICE: I have, obviously, lots 89 1 more questions, but I'm going to open it up for the -- 2 MR. PECKMAN: Could I respond to that one? 3 CHAIR PRENTICE: Yes, sure. 4 MR. PECKMAN: I would agree with Ada Sue 5 that it's not in Subpart A that the problem lies, but 6 in guidance and interpretation. However, it seems to 7 me that some of the guidance that has come out about 8 tabling or deferring proposals is appropriate in that 9 it forces the IRB to really consider what can be 10 approved by a chair, and it forces the IRB then to 11 give very specific information back to an investigator 12 rather than vague guidance saying well, tell us how 13 you're going to do this rather than saying this is 14 what we'll accept as an IRB. And by forcing an IRB 15 into that kind of statement with an investigator, it 16 then grounds the investigator in a place to make 17 decisions and discuss with the IRB how to negotiate 18 the terms of approval. 19 CHAIR PRENTICE: I appreciate that aspect 20 and that is an advantage, but it's also a 21 disadvantage. Okay. We're going to go with our usual 22 routine here. You raise your hands and I recognize 90 1 you and then you ask a question. Susan? 2 MS. KORNETSKY: I just wanted to follow-up 3 on that point, and thank you all for the 4 presentations. The issue that, Ernie, you just 5 brought up is an example where it isn't necessarily 6 the regulations, it's the guidance and this is a 7 catch-22. Yet, throughout everyone's presentations we 8 were saying more and more guidance is needed. 9 So in some respect and, you know, I have 10 been working probably as long as Ada Sue, 20 years, on 11 this stuff and years ago we didn't have to send all 12 that stuff back to the Committee. I don't think any 13 human subject is better protected and I don't know 14 whether I'm correct or not, but I don't even know if 15 there is any guidance on that, but that's something 16 that has come out in determination letters. 17 So in thinking about a lot of the things 18 that you said, the regulations actually give us lots 19 of flexibility. Yet, we're asking for more guidance 20 and when we sometimes get the guidance, then we're 21 sort of cornered in. So I am trying to sort of figure 22 out what the mix is and I don't know the answer and 91 1 you folks don't probably either, but I'm not sure just 2 asking for lots of guidance is the answer. Maybe it's 3 asking for a little more flexibility in being able to 4 interpret things at a local level. I don't know. 5 MR. PECKMAN: Now that I'm awake I can -- 6 I think what you heard this morning was that the 7 document is sound as Subpart A goes, and also that if 8 you ran through everything that we had to say, even 9 though we're each limited to 20 minutes or less, that 10 there is a lot of overlap and there is not a whole lot 11 of areas that need a lot of attention. But there are 12 some very basic and fundamental areas that need a lot 13 of concrete attention, and because there has never 14 been any guidance in these areas, I think is very 15 telling. 16 For the 25 years or so the document has 17 been in existence, there hasn't been any guidance 18 about the definition of a human subject or about the 19 definition of research, for example, and there are 20 very fundamental aspects of the subpart that need some 21 real attention. 22 MS. SELWITZ: You know, actually, I have 92 1 to say if you look closely, I only asked for guidance 2 in two areas and then I did it with some hesitation. 3 And so I, quite frankly, think that I don't ask for 4 guidance across the board, I mean, and it comes down 5 to what you believe and the faith of the system. And 6 I do think our system is being undermined with too 7 much detail and with expectations that we end up 8 spending so much time, guys, on the details that we 9 lose sight of the ethical issues and that core 10 question, sound, ethical decision making. 11 You know, I mean, it embarrasses me that 12 I have to sit in a meeting and count how many people 13 are there and make sure if you abstained, you have got 14 to leave the room and when you leave the room, I don't 15 mind leaving the room, but then you lose your quorum, 16 do we still have a quorum? That kind of thing 17 embarrasses me, because to me we're getting away from 18 the core issue, how do we promote sound, ethical 19 decision making? So I can't say that I agree with all 20 my esteemed colleagues on all the areas in which they 21 wanted guidance. Several of those areas I didn't want 22 guidance. 93 1 MR. FORSTER: I think you have to weigh 2 the need for guidance against the fact that there's 3 compliance and oversight by agencies. If we didn't 4 have that oversight and you didn't have to worry about 5 receiving citation letters, then I wouldn't want as 6 much guidance and I think that we could stick to our 7 core values and duties and protect human subjects, but 8 the fact is we have to do that as well as respond to 9 the agencies. Sometimes the two duties don't 10 dovetail. 11 CHAIR PRENTICE: You know, David, you 12 mentioned a 400 page minutes and I understand the 13 current philosophy of if it ain't written down, it 14 just didn't happen, okay, and I know how that arose 15 and institutions are under the gun to document, 16 document, document. Personally, I don't think that 17 adds any value to human subjects protection. 18 I think it's ridiculous that IRB minutes 19 are expanding, you know, exponentially. Back in the 20 old days you would have IRB minutes that might be 21 three or four pages. Institutions now have IRB 22 minutes that are 100 pages. So I wonder what you all 94 1 think about that. 2 Would guidance, specific guidance issued 3 by OHRP with respect to what has to be in minutes help 4 resolve some of this problem, you know, not recording 5 who goes to the bathroom at what particular time 6 during the IRB meeting? That's where we are now. 7 Don't you think that might help? 8 MR. FORSTER: You know, it depends on what 9 you think of the overall human subjects protection 10 system and if the agencies want a way to know what 11 we're doing, they can't get that unless they have us 12 document stuff. So if you think that the agencies' 13 oversight does help provide human subjects protection, 14 then the documentation is a way for them to come and 15 check on it. I think you have to weigh it in that 16 broader context. Does the IRB do a good job with or 17 without documentation? 18 MR. PECKMAN: I guess one of the 19 questions, Ernie, you have to ask yourself is if in a 20 clinical trial you are required to document whether 21 anyone is breathing or not, basically, when they came 22 in and what they did, blah, blah, blah, blah, why 95 1 wouldn't you want the IRB to document their decision 2 making process? And in a sense that's not a question 3 of whether it was documented and whether it happened 4 or not. We know things happen every day that aren't 5 documented and they happen, right? 6 I mean, everything I do isn't written in 7 the LA Times, thankfully, and I do a lot of stuff 8 every day. But the question is whether you can 9 demonstrate you did what you're supposed to do, and I 10 don't think that IRB minutes necessarily need to be 11 400 pages, but I don't think a three page IRB minutes 12 for the review of 20 protocols is going to be 13 sufficient either. And I think the old days in many 14 ways are thankfully gone, but there has to be a 15 balance. 16 MS. SELWITZ: I would agree there has to 17 be a balance, but I tend to balance it more along the 18 lines that Ernie -- I tend to agree with you, Ernie. 19 I think we have gone too far. The pendulum has swung 20 too far. And one of you, I don't remember which one 21 of you brilliant men said this, but one of you said 22 that maybe what we should be thinking about on minutes 96 1 is does it have to be the minutes as the way to 2 document? And that gets us back to interpretation and 3 flexibility of interpretation. I believe that I would 4 agree, Steve, with you that we need documentation, but 5 I'm not convinced that all that documentation belongs 6 in the minutes. 7 MR. PECKMAN: Sure. 8 MS. SELWITZ: And I think there are a 9 variety of ways to be able to demonstrate to an agency 10 how well we're doing without focusing only on the 11 minutes document. 12 CHAIR PRENTICE: Susan? 13 DR. WEINER: I think I want to support 14 what Ada Sue just said. If the problems really have 15 to do with administrative process and documentation of 16 requirements and not the spirit of what is in Subpart 17 A, then it seems to me that those are the places where 18 the focus should be made, so that what is 19 administrative and what is documented can be 20 validated, but that the energy -- there is a shift of 21 burden in thinking about the ethics of what needs to 22 be done and the protection of the subjects and the 97 1 enhancement of research rather than the paperwork. 2 And I don't know that that's a business for OHRP. I 3 think to some extent, those are in a way professional 4 standards. 5 MR. PECKMAN: I would just like to add one 6 thing, which is I have been a consultant to many IRBs 7 and the IRB minutes I read are mostly a recitation of 8 the protocol and they are not a description of the 9 controverted issues, what happened in the discussion, 10 in the IRB meeting. And if IRB minutes were focused 11 on what they are supposed to be focused on rather than 12 what they are not, I think you would see a lot 13 different IRB minutes. 14 CHAIR PRENTICE: Felix? 15 DR. GYI: We do end up talking a lot about 16 documentation and one of the questions that was raised 17 yesterday was be careful what we ask for in terms of 18 guidance, because we will get it and that may simply 19 add to our regulatory burden. And I think that is a 20 concern that we all have and that is certainly an area 21 that we're all wrestling with right now. 22 Part of what I have heard this morning is 98 1 also a reactive approach to not just defensive record 2 keeping, but defensive documentation for litigation 3 purposes. We heard David talk about being reactionary 4 to Maryland's law, but that's an interpretive issue, 5 because we had the occasion of talking with Jack 6 Schwartz of the Attorney General's Office just about 7 a week ago and we asked the question are IRBs outside 8 of Maryland held accountable and responsible for 9 providing the same kind of disclosure and giving up 10 the minutes, and his response was no. Only IRBs in 11 Maryland are held to that particular standard. 12 So you can see that there is a great deal 13 of interpretive reaction that we all engage in in some 14 way because of determination letters that either come 15 out from OHRP or warning letters that come out from 16 FDA and without real guidance in terms of its 17 application across the board, we end up using those as 18 bench marks of how we need to behave. 19 And I wonder if you had some thought on 20 how we might use that as an example of providing or 21 asking OHRP to give us guidance that's a little bit 22 more streamlined, but flexible in the nature that we 99 1 have been talking about. 2 MR. FORSTER: You know, sometimes I think 3 that OHRP could give us helpful guidance by saying 4 there is a variety of ways you can do this, and within 5 this spectrum we think this is okay instead of do it 6 this way. Now, that's hard to write and it's hard for 7 them to know what problems we're facing and what the 8 possible spectrum might be so, you know, they don't 9 have all that many personnel and they are not out in 10 the IRBs that often. So I think it would be hard for 11 them to issue that oftentimes. 12 MS. SELWITZ: You know, and again, one of 13 you talked about common sense in compliance and, you 14 know, I have to say I'm really sympathetic with OHRP. 15 I would much rather sit here and tell you how I think 16 you should do it than to have to do it, because you're 17 a public agency and your guidance is scrutinized in a 18 variety of ways, and I think that has dictated how the 19 guidance comes out where I think you have made real 20 efforts to be flexible but, in fact, you know, your 21 guidance is going to be looked at throughout the 22 United States and very closely, and so you have to 100 1 balance a lot of things. 2 And that is one of the reasons I'm a 3 little afraid to ask for too much guidance, because I 4 do think the pressures that are on any federal agency 5 that issues guidance will encourage them to put out 6 some fairly specific "you must, this is how" and you 7 all do put out lots of "here's a range." I mean, I 8 have seen plenty of guidance that does this, but I 9 think your question is a hard one, because how do you 10 craft guidance that doesn't get us back into this loop 11 where we're so tied up in the bureaucracy we're 12 missing out on what counts, which is real protection 13 of human subjects and sound ethical decision making. 14 MR. PECKMAN: Well, I will give you an 15 example. You have a social science researcher who 16 wants to go out and do surveys on stigmatizing 17 behaviors. So a longitudinal survey, you're going to 18 follow-up people three times a year for three years. 19 So there is identifiable data. The consent form 20 clearly is not the only link to the subject that could 21 cause a risk to a breach of confidentiality, but you 22 don't want to have people sign a consent document, so 101 1 you're stuck in 117(c). 2 It would be very helpful if OHRP would 3 issue a guidance for many IRBs that says look, if you 4 get a certificate of confidentiality, if you implement 5 various types of these standard confidentiality 6 procedures, it's possible to lower the risk to minimal 7 risk, right? So it's the nature of the guidance that 8 we need help with, because this will help the 9 investigators in the IRBs struggle with very important 10 issues that speak to the core of the protection of the 11 subjects. The research should be done. It should be 12 done in an ethical way that protects the subjects, but 13 yet the requirements are such with not enough 14 definition that allow the IRB to make these decisions 15 without being fearful. 16 CHAIR PRENTICE: Mark? 17 MR. BARNES: Yes. Thank you all for a 18 very interesting discussion and for all the time you 19 put into a review of the regs. I think that was 20 great. Just following up on what Ada Sue said and 21 then what you said. You know, if the problems that 22 are identified are that there are particular parts of 102 1 Subpart A that are odd in their application to a wide 2 variety of research projects and that's the problem, 3 then the solution to that may be not to have guidance 4 about it, but to change the requirement to make the 5 actual requirement itself in Subpart A a thematic 6 requirement rather than one with such specificity that 7 it yields all of these problems. That way you, 8 hopefully, vindicate. 9 Now, you know, not leaving aside the 10 practicality of revising Subpart A, what I am saying 11 is if you could do that and if you could achieve 12 uniformity among the federal agencies, which is 13 something that Ada Sue pointed out, then that would be 14 a better result than having more and more guidance to 15 remedy problems of over-specificity in the bedrock 16 regulation itself. 17 I am sympathetic to the idea that, you 18 know, the more guidance you have, the more complicated 19 everything becomes, because it is metaphysically 20 impossible to issue guidance that will satisfy all 21 possible permutations and combinations of 22 circumstances. So what you always have to go back to 103 1 are really first principles and figure out what it is 2 that you want to achieve thematically and then let the 3 local IRBs or central IRBs, which we'll get to later, 4 let them figure out their own path to that goal. So 5 that's one thing I wanted to say. 6 The other thing I wanted to say is that, 7 you know, in my experience counseling a lot of IRBs, 8 I will tell you that, you know, whether the IRBs are 9 compliant with every jot and tittle of Subpart A or 10 Subpart C and in all the guidance, nobody can actually 11 comply with everything. It's not possible, frankly. 12 It's just not, but the problem is in terms of 13 substantial compliance, in my experience, is not the 14 IRBs. The problem is the investigators. 15 If I see problems out there, it's not that 16 the IRBs are not doing their work. It's that 17 investigators are sloppy, are not getting informed 18 consent in situations where they promised they would 19 get informed consent in their presentations to the 20 IRB. Investigators are getting continuing review 21 applications and letting their studies lapse, but they 22 continue to implant the devices. I mean, that's the 104 1 stuff that I see all the time and that stuff is truly 2 bad stuff. It's not just paperwork stuff. 3 So there is also a question, I mean, of 4 emphasis and whether the emphasis really ought to be-- 5 if what you want to do is protect subjects, whether 6 the emphasis should be on the IRBs at all, at this 7 point, or whether the emphasis ought to be placed on 8 those who actually are supposed to implement what the 9 IRB directs. 10 MR. NELSON: Well, I'm hearing an inherent 11 tension here. You're helping crystalize it for me. 12 We're wanting to have our cake and eat it, too, I 13 guess and maybe that's the nature of the beast. 14 There's a lot of mixed metaphors there. Sorry. 15 On the one hand, we appreciate the 16 flexibility and, in some cases, the vagueness, the 17 framework that it gives and then we work within that 18 and let us do what needs to be done. On the other 19 hand, and I think you have a very OHRP friendly panel 20 up here, because we have all worked with you and 21 respect the work you're doing, and I will add to the 22 recognition that you're in a difficult spot as a 105 1 regulatory agency, because when it comes down to those 2 determination letters, now we get down to very 3 specific issues and you are effectively writing case 4 law whether we like it or not. 5 So on the one hand, we like this vague 6 framework over here, but on the other hand we all have 7 to look at what other institutions are being held 8 accountable for and saying ah, it's not quite as 9 flexible as we had hoped it was or thought it was or 10 that doesn't match where we have fallen in that 11 flexible application. 12 CHAIR PRENTICE: Susan? 13 MS. KORNETSKY: One of the points that 14 Mark had, and I'm sort of thinking about things as we 15 go along, is whether it's helpful to think about in 16 the future sort of guidance in a much more sort of 17 performance based on not prescriptive way. And I 18 think that there are certain things that you probably 19 can and can't do depending on the regulations. 20 You know, two examples, the issue, you 21 know, the minutes and what Ada Sue said. You know, 22 you can come to our files and through other types of 106 1 documentation follow the discussions or whatever. It 2 may not just be in the minutes. Even the issue of 3 sort of secondary subjects and deciding, you know, to 4 bring up that there are issues, they need to be 5 addressed. There are confidentialities, there are 6 issues, but then to allow some flexibility in there as 7 opposed. 8 I'm just very concerned about more and 9 more prescriptive types of guidance and, again, I 10 think we're going a lot by the determination letters, 11 the case law, Dan, that you said that, you know, we 12 read it and oh, my goodness, we're doing that wrong. 13 That isn't even guidance yet, yet it's, you know, some 14 determination that had been made in a particular case, 15 and sometimes we don't have all the details behind it. 16 I mean, sometimes, you know, there's more to it than 17 what's written there. Yet, we're all very reactive 18 like oh, my goodness, we're doing something wrong. 19 And I think some of the accreditation 20 agencies have tried to think more where they can about 21 performance based rather than very specifics to prove 22 that you're meeting a certain intent or principle, and 107 1 I'm just thinking whether there is some way of trying 2 to incorporate that into guidance as we go forward. 3 CHAIR PRENTICE: Mary? 4 DR. POLAN: Well, first I would like to 5 thank everybody for the presentation and it helps me 6 understand what IRBs are concerned about, and I 7 actually wanted to speak from the perspective of a PI 8 and particularly after Mark's comment, because in my 9 experience PIs really do want to do the right thing 10 and they really don't want to break the law and they 11 definitely don't want to get sued or put in the 12 newspapers, so they try. They just are very busy and 13 they don't understand what they are supposed to do. 14 And the other things that you said that 15 rang true were some kind of an association with all 16 the institutions and agencies that PIs deal with and 17 in particular, after Steve talked about living in 18 California, and although northern California isn't 19 quite as heterogenous as southern California, it's 20 close. You know, if you apply for an NIH grant, you 21 are required in the renewals to put down what the 22 ethnic distribution of your subjects are and you can 108 1 get your grant pulled if you aren't really doing that 2 appropriately. 3 But you can't get people to sign consent 4 forms in many cases. And so clearly, you're listening 5 to your investigators and that would be one major 6 plus, which I guess, as Susan has suggested, doesn't 7 have to come as guidance, but a broadening of the 8 ability to get consent that is documented 9 appropriately without doing it in just our 10 anglocentric way. 11 But the question I have in response to 12 Mark's statement that investigators sometimes make 13 mistakes, and I think that the intent is not to make 14 a mistake, the intent is to do it right, but just the 15 inability or not enough time or not the knowledge, 16 what have you all heard from your investigators? If 17 you were to phrase this as what would we do to change 18 the way we conduct research, not to change the 19 regulations or the guidance, but how would we do it to 20 conduct research that is ethical, protects patients 21 and helps the investigator? What things would you 22 suggest? 109 1 MR. PECKMAN: I suggested a whole lot of 2 them for social-behavioral researchers. I think they 3 are the researchers who are in the biggest bind right 4 now due to lack of guidance and because of certain 5 peculiarities within the regulations that actually 6 speak through a biomedical model, which I don't buy 7 into the whole subpart does, but there are certain 8 aspects of it that do. 9 I would also say that I agree with you 10 that the PIs, I think, that Mark is referring to is a 11 real outlier. Now, clearly, that's probably when Mark 12 gets called in is when that outlier exists, and when 13 I have those outliers they take up 80 to 90 percent of 14 my time, and so they impact me very much directly, but 15 that's not the biggest problem. 16 To me the biggest problem is the concern 17 for a philosophy of an institution that I have seen on 18 many coffee mugs as the kind of research for fun and 19 profit. That is the problem, is where money starts to 20 rule the game rather than the ethical responsibility 21 of those who are participating in the game. What 22 concerns our investigators the most is the ability to 110 1 get the research started, and I don't think that's a 2 part of the regulation. I think it's a part of the 3 institutional commitment to help the process work for 4 the investigator and the IRB. 5 DR. JONES: Well, I want to follow-up 6 about researchers, because I do think, you know, I 7 might agree more with Mark that it's not that, I don't 8 think, researchers always understand the principles 9 that they are really trying to fulfill and I don't 10 know if we actually did a survey. I think, Ada, you 11 said, you know, if that's the bottom line that we want 12 to do ethical research, you know, how many researchers 13 really understand some of the nuances? 14 And I think it really has come to home 15 when we have been going over Subpart C that, again, 16 we're throwing things back on the investigator that 17 they should be able to do and, frankly, it's not that 18 they don't want to do it. It's just where are they 19 going to carve out both the knowledge set that they 20 can do it adequately and really understand the ethics 21 and the philosophy behind it, as well as the time and 22 the manpower? 111 1 You know, and then again you go back to 2 who is going to fund all these other people that you 3 have to do to make sure that you can fill out all the 4 paperwork that's necessary, you know, which may or may 5 not actually protect the subjects. But I do think a 6 better knowledge base of some of the principles of 7 human subjects projection is really needed. 8 So on that end, someone pointed out the 9 need for education. You know, how would we bring 10 along -- I think, you know, most of the concerns that 11 we have heard today have been institutional or IRB 12 mechanism concerns rather than the research community 13 themselves. So how do we get to them? You know, how 14 do we help it work in Subpart A, so that they can 15 actually do what the intent is, you know, to protect 16 human subjects? 17 MS. SELWITZ: You know, and if I had an 18 answer I could probably retire, at this point, but I 19 think it's a core question and, actually, I do think 20 we need to be putting more emphasis on the 21 investigators not because I don't think they want to 22 do the right thing. Actually, it's for the very 112 1 reasons you articulated. They want to do the right 2 thing. Many of them just don't have the time. They 3 don't have the background. And what I think 4 complicates it is all this bureaucratic detail, 5 because investigators, you ask what do I hear from my 6 investigators, well, I hear why can't we do this 7 faster? You know, well, we can't do it faster, 8 because we need to make sure we're full reviewing. 9 I mean, you know, you get caught into this 10 loop, but, you know, you hate to say well, education 11 is the answer, but I actually think education is a big 12 part of the answer. But what do we educate them on? 13 That's what worries me. Have I got to spend three 14 days getting them to learn about these regulations and 15 the ins and the outs that they don't really care about 16 to begin with? They do care about protecting human 17 subjects. 18 So how do we get at those core issues? 19 And I do think those issues are in the criteria for 20 approval. I mean, I believe if you look at those 21 criteria, how do we get investigators to apply that 22 criteria in how they design their research and how 113 1 they assess their own research? 2 CHAIR PRENTICE: Tom, would you like to 3 yield to the lady from New York who wants to say 4 something? 5 DR. WEINER: As a follow-up to this, my 6 question really is to what extent do you think that 7 accreditation efforts really answer or resolve some of 8 the education and other issues which we talked about 9 this morning? And I raise that from a public 10 perspective, because in a way that may constitute a 11 shorthand for public willingness to participate and to 12 trust institutions and investigators who are 13 associated with them. 14 MR. FORSTER: I'll take that one on in the 15 context of the Panel. We went through accreditation 16 with AAHRPP and it was a lengthy process involving an 17 incredible amount of documentation. We had to fill in 18 some gaps. They identified things that we had not 19 covered as far as their accreditation standards. It 20 resulted in more paperwork and more procedures and 21 guidance. And right now, what I consider my job as a 22 compliance officer is I keep us compliant with FDA, 114 1 NIH, ICH, HIPAA and our accreditation standards. 2 And, quite frankly, our Board members 3 can't remember where the requirements come from. You 4 know, we need some regulatory staff in there saying 5 well, that's because of ICH or that's because of 6 accreditation. I think, you know, the idea behind 7 accreditation is proper and I think it made us a 8 better IRB. It didn't make us a simpler IRB. It 9 didn't reduce our amount of documentation or SOPs or 10 anything else. 11 As far as how did the accreditation help 12 with education? Not much. We were already doing a 13 pretty good job of education, I think. We didn't 14 change any of that due to accreditation. And the 15 accreditation did require us to go out and ask 16 investigators about their training. It didn't set a 17 standard for what training investigators have. And 18 it's tough to train on ethics. By the time everybody 19 is done with the compliance issues and the good 20 clinical practices, the ethics is kind of considered 21 a luxury and it should be the opposite unfortunately. 22 CHAIR PRENTICE: Okay. Tom? 115 1 MR. ADAMS: I would like to thank the 2 Panel. 3 CHAIR PRENTICE: Not until later. 4 MR. ADAMS: Sorry. 5 CHAIR PRENTICE: No, go ahead, Tom. 6 MR. ADAMS: I would like to thank the 7 Panel as well. It was terrific. I also appreciated 8 a little bit of a levity, which is always nice in 9 these proceedings. Each of you spent a great deal of 10 time talking about guidance. And I was hearing it 11 from two directions. One was there is a lot of 12 guidance that is out there, whether it is coming in 13 through determination letters or messages that OHRP 14 sends out. At the same time there was need for more 15 guidance in certain areas. 16 Someone mentioned the potential of a 17 compilation of the guidance which is out there. I 18 wondered if you could comment further as to what type 19 of emphasis you would put on that? Is that something 20 that would be absolutely of interest across the board 21 or does the present method of getting it out via 22 different methodologies seem to work for you? 116 1 MR. NELSON: Well, I'm afraid that was me, 2 so I'll take the first stab at it and then invite my 3 colleagues to jump in and agree or disagree. I guess 4 my observation was that yes, there is a lot of good 5 guidance out there. Having said that, there are a lot 6 of people out in the field who don't know its there or 7 don't have the time to wade through and somehow pull 8 that into their daily practice, so I think there's a 9 gap between. 10 I think, actually, OHRP is already over a 11 period of years addressed a lot of things and they 12 have come pretty close to compiling it, if you will, 13 at least on a single webpage. But I think you could 14 make a whole career out of wading through that webpage 15 to distill it down and then to translate that is, I 16 think, where the gap is into the daily practice of 17 IRBs. I can't afford to hire one or two people to 18 spend all day on the OHRP website going through and 19 figuring out where the gaps are in terms of 20 translating that into our SOPs. 21 I guess I hang that out there as a 22 question. I'm not sure how to improve on what it is, 117 1 because they have done a pretty good job with what's 2 there. The truth be told, it's a big job. It's hard 3 work, as our President says, and, you know, how to 4 make that translation is the challenge, I think. 5 MR. PECKMAN: I think the website is very 6 user friendly. I actually -- we haven't had any 7 problems and the IRBs I have worked with, actually, 8 have found the website to be very useful. I think 9 where we did a lot of our education regarding the IRBs 10 was right out of the IRB Guidebook. There are case 11 studies. There are examples that have proven to be 12 very helpful. There's only one problem. The IRB 13 Guidebook was published in 1993, 11 years ago, and I 14 have no idea how many years it took to actually write 15 it. 16 It is in clear need of an overhaul and 17 there is a lot of areas, I think, that we have heard 18 today where it would be much appreciated from IRBs to 19 have that kind of guidance within the guidebook 20 itself. I think it's a real fundamental article of 21 information for IRBs that is still useful, but needs 22 an update. 118 1 MS. SELWITZ: Can I say one thing? You 2 know, yes, I liked your idea, Dan Nelson. I thought 3 that was a good idea. My problem is I need it for 4 FDA. I need it for NIH. I need it for the CDC. I 5 need it for NSF. I need it for the Department of 6 Education. I mean, that's my problem is that it's not 7 just an OHRP issue in terms of total human research 8 protection. It's that I need it from everybody. I'm 9 not going to get it from everybody, and the truth is, 10 you know, I don't want a lot of guidance. But what's 11 there, I would like to have it consolidated so it's 12 easier to access and it's easier to find. 13 MR. BARNES: You just don't want to be 14 told what to do. 15 CHAIR PRENTICE: Okay. The last comment 16 will come from Bern. 17 DR. SCHWETZ: Just a couple of comments. 18 One on the IRB Guidebook. We realize that it is out 19 of date and to write a counterpart today means that it 20 takes longer to write a document of that kind than it 21 does for the input to change. No matter what we 22 produce, it would be out of date, because the 119 1 interpretations and things change so quickly today. 2 So what we're looking at instead is a web-based 3 information center that will have the information that 4 a new guidebook would have had in it if we had taken 5 that to completion. So hopefully that will be a 6 resource that is available to you that can be changed 7 on a daily basis, a weekly basis, whatever it would 8 be. 9 I want to comment about the system that we 10 have for compliance in this country for human subject 11 protection. Thank you for sharing your experience, 12 Panel Members, because it is valuable to get your 13 insight and I'm impressed by how consistent your 14 message is between you. I think one feature that is 15 unique about the human subject protection system that 16 we have is that it's built almost totally on trust. 17 If you look at other compliance systems we 18 have in this country, we have inspectors in every port 19 for looking at imports. We have meat inspectors in 20 every plant, every slaughter plant, every packing 21 plant. So the decisions are made at a local basis on- 22 site and they are backed up by a larger system, but we 120 1 can't operate that way with human subject protection. 2 We don't have an inspector to put in every research 3 site. 4 So while there are a lot of things that 5 were identified today that have to do with either 6 over-interpretation or under-information from the 7 standpoint of needing more guidance or changes to 8 regulations, I think SACHRP looks at this issue of 9 what to do with Subpart A from this point on. Having 10 heard all of this input, it should be taken into 11 account based on a system that is focused on trust, 12 not that we write sufficient guidances that there will 13 be an inspector in every research site to stand there 14 and watch to make sure that every investigator does 15 the right thing, according to the new guidance. 16 It makes it particularly challenging to 17 come up with a system that is based on trust where you 18 don't remove the trust by providing too much guidance 19 or you don't allow it to work well, because you don't 20 have enough. It's a very difficult balance to strike. 21 CHAIR PRENTICE: Okay. In the interest of 22 figuring out what we do next, I would like to once 121 1 again ask is there anyone here from the public that 2 would like to address SACHRP? I have one request. I 3 need to know if there is anybody else? I have two. 4 Okay. All right. Then what we will do is we will 5 take a -- we're a little bit off schedule now, so I'm 6 going to cut the break a bit short. We'll take a 12 7 minute break. 8 MS. SELWITZ: I'm going to time you, 9 Ernie. 10 CHAIR PRENTICE: And then we will 11 reconvene with the public comment section. 12 (Whereupon, at 10:48 a.m. a recess until 13 11:05 a.m.) 14 CHAIR PRENTICE: We're going to reconvene. 15 Would people, Ada Sue, okay. Thanks, I'll talk to you 16 in a bit. All right. We have a request from three 17 members of the "public" to address SACHRP, and I would 18 like to invite them to do that at this time, beginning 19 with Dr. Ivor Pritchard and as is customary, come up 20 to the microphone, identify yourself and your 21 affiliation and try to limit your comments, please, to 22 no more than five minutes. 122 1 DR. GYI: Ernie? 2 CHAIR PRENTICE: Yes? 3 DR. GYI: Should we have the Panelists sit 4 up front as well, so that they can be ready? 5 CHAIR PRENTICE: No. 6 DR. GYI: No? 7 CHAIR PRENTICE: These are comments to 8 SACHRP. Traditionally, we don't have the Panelists 9 respond to comments from the audience, unless we need 10 to, and then we can always ask one or more individuals 11 to come up. Please. 12 DR. PRITCHARD: Is it on now? No? I 13 don't need it. 14 UNIDENTIFIED SPEAKER: Now it's on. 15 DR. PRITCHARD: Okay. Ivor Pritchard, 16 Office of Human Research Protections. I would just 17 like to ask SACHRP to add one more technical item to 18 the list of issues that have to do with Subpart A. 19 This is an issue that has to do -- that I'll try and 20 identify by asking two questions. The questions are 21 about Section 101(b), which has to do with applying 22 the exemptions and making the exemption decisions. 123 1 Now, 101(b) says that "Research activities 2 in which the only involvement of human subjects will 3 be in one or more of the following categories are 4 exempt." The first question I have is where is the 5 complete list of categories of human subjects 6 involvement that are not the exemption categories, but 7 are categories of human involvement in research? If 8 the answer to that question is nowhere, and I take it 9 that that's the case, then how can anyone judge 10 whether the subjects "only involvement" will be in the 11 exemption categories? Thanks. 12 CHAIR PRENTICE: That's very interesting. 13 I never noticed that before. Thanks for bringing that 14 to our attention. Okay. Please. 15 MS. CABO: Is this on? I'm sorry that 16 neither Celia nor Susan are here in the room right 17 now, but I'm sure you can pass this on. My name is 18 Mary Ann Cabo. I'm with the Society for Research and 19 Child Development and actually my comment now bridges 20 what we have heard this morning about Subpart A with 21 the ongoing work of the Subcommittee on Children. 22 What I want to just request, I guess, for 124 1 SACHRP to pay attention to as it goes forward from 2 here is some fundamental issues with the 3 developmentally appropriate use of the assent 4 regulations. As a psychologist at that time working 5 in pediatric oncology and also a member of the IRB, I 6 have been heartened to see how well investigators and 7 IRBs have tried to honor children's capacities. But 8 I think it has gone a little awry over the years, 9 particularly with clinical trials and Phase One 10 studies where many institutions are requiring written 11 assent from very sick children. 12 Without enough developmental expertise on 13 the part of investigators to write those assent forms, 14 there is unnecessary delays in the research getting 15 started. And at the same time, without enough 16 developmental expertise on the part of IRBs, there 17 isn't a quick solution either. In some cases, kids 18 who have never been asked to sign their name before 19 are being asked to sign their name for beneficial 20 research when they are really quite sick. 21 So I just want to raise that whole 22 spectrum of us leaning a little bit too far toward the 125 1 idea of regulation and protecting the institution, 2 rather than the human subject issues that I think 3 drove that, those assent guidelines to begin with. 4 There is also a quite a bit of inconsistency across 5 IRBs in whether children have to sign assent forms or 6 they can be given -- they can give their assent orally 7 and so that's another issue that is related. That's 8 all. 9 CHAIR PRENTICE: Okay. Thank you for that 10 comment. That reminds me of the time that I went to 11 a very well-known children's hospital and interviewed 12 parents and asked them what they thought about written 13 assent. And, quite frankly, to my surprise, they all 14 indicated they did not feel it was appropriate to have 15 these sick kids sign assent forms. On the other hand, 16 my institution does have a written assent requirement, 17 so I went back to our Pediatric Oncology Department 18 and said, you know, what do you think about this? Why 19 don't you decide whether or not you want to request a 20 revision of our guidelines. 21 They discussed it and said we like the way 22 it works. However, I do appreciate the fact that it's 126 1 very difficult to develop the appropriate assent form 2 written at the appropriate cognitive level. So thanks 3 for your comments and I'm sure that the Children's 4 Research Subcommittee will consider that as they move 5 forward through to Subpart B. Okay. Where is John? 6 John, please? 7 DR. MATHER: Dr. John Mather from the 8 University of Michigan. I just want to make two 9 points. One is that the Panelists were excellent and 10 they have been around for many years involved in these 11 issues. I think that the comments about whether we 12 should or should not have guidance ought to be 13 tempered by the fact that in the field, at least with 14 respect to IRB administrators or those that run those 15 activities, I'm not sure how many IRBs around the 16 country right now, but maybe some 3,000 or 4,000 of 17 them, and maybe not all have been registered yet. 18 But I hear stories and people call me 19 about well, where do I find the new IRB administrator, 20 because we found this institution, but we don't want 21 OHRP to find out about some problems that we have 22 here, but we need somebody to get in and get going on 127 1 improving our process here. And finding somebody to 2 do that results in maybe finding somebody with an MPH 3 or, you know, a lawyer and then there is an intense 4 period of time, maybe six or nine months, before they 5 really get in depth understanding of what really are 6 the problems and what needs to be fixed. 7 Any good guidance that they latch onto is 8 very helpful. So I would encourage you to think 9 about, Bern, the need for this guidance and for this 10 website. Even if you just took all the stuff at OHRP 11 right now, forget about FDA, forget about the other 12 agencies, just your own stuff and get that codified 13 and on the website, I think that would be immensely 14 helpful to what I would call "the run of the mill IRB 15 administrator," not the excellent folks we had on this 16 particular Panel. 17 The second point is I want to resonate 18 with Steve Peckman's position about the institutional 19 ethos. I mean, it's not so long ago when we heard 20 comments that we need a culture of conscience not 21 compliance. That comes from, I think, Bern, your 22 predecessor, maybe the first time that was said. 128 1 Well, that's very difficult to pull off in an 2 institution and I would hate to see at least the 3 assurance, some sort of document that an institution 4 has to sign to commit itself to doing that, being sort 5 of in any way as denigrated. 6 We have had problems where people have 7 signed on their IRBs to our assurance, much to our 8 assurance, we have had to get that IRB's, which is 9 registered under the University of Michigan, assurance 10 actually moved away. So I think that very process is 11 a very, very important process in and of itself, 12 because it becomes a governance document. It means 13 that you can hold the institution official to some 14 degree of authority and accountability. 15 I would suggest that the institution 16 official's job is more than just insuring what we have 17 been talking about gets done, but extends to some 18 impact on who is going to be the day-by-day mentor or 19 guider to the investigator or researcher. An academic 20 center, I think, there has got to be better attention 21 given to the department chairman or the division 22 chief. It can't all just fall onto the IRB to make up 129 1 for making sure that things are tracked. 2 I'm not quite sure how it works. Well, 3 we've been thinking through how to do a better job of 4 mentoring the scientific process and how an individual 5 faculty person does their science. But it's got to be 6 much more of a shared responsibility and not just 7 evolve on the IRB if something goes wrong. I don't 8 know how it can be better articulated than what we 9 were hearing earlier about this is all falling down 10 onto, ultimately in the cascade, the researcher or 11 investigator, and they are the ones that if anybody is 12 failing in battle right now, they want to know well, 13 what is it I simply need to do? 14 When one says to them well, your 15 department chairman should be able to help you in that 16 regard, say, I've been to the department chairman and 17 he doesn't or she doesn't have a clue what it is I'm 18 really supposed to do. So, Steve, I would extend your 19 comment about the institution official having the 20 responsibility not just for the human subject 21 protection, but it's also got to deal with the full 22 academic enterprise where we're talking about an 130 1 academic medical center. 2 CHAIR PRENTICE: Thanks, John. Okay. The 3 reason why we have these Panels is to try to identify 4 issues that we think are important for SACHRP to 5 consider. It's very clear that we brought in four 6 experts and we thank you very much for your very 7 thoughtful and insightful presentations and 8 identifying the issues that you face every day. It's 9 very clear that some IRBs are over-interpreting the 10 regulations creating regulatory burden for both 11 themselves and the investigators that they serve. 12 Some IRBs are inconsistently interpreting 13 the regulations. They are all over the waterfront. 14 That creates problems in and of itself. Some IRBs are 15 under-interpreting the regulations. They are 16 utilizing a minimalistic approach. They are still 17 within the boundaries of the regulations, but they are 18 still utilizing a minimalistic approach and they are 19 not providing perhaps the best level of protection 20 they should. 21 Some IRBs incorrectly interpret the 22 regulations, which results in noncompliance on part of 131 1 the IRB and the investigator and it possibly increases 2 the risk to which human subjects are exposed. These 3 are all problems that we are aware of. We've been 4 dealing with these kinds of problems ever since we got 5 into this area, in my case back in 1980. So I think 6 it's time that we, you know, try to address some of 7 these problems. 8 Now, the way we normally approach charges 9 given to SACHRP by the Secretary is we have a Panel, 10 as we have had today, they identify the issues, then 11 we decide where we want to go. So I want to spend the 12 next 12 minutes or so before lunch with the discussion 13 among the SACHRP Members as to what you think is the 14 next step that we should take with regard to this 15 particular issue that has been brought before us. 16 Mark? Now, wait a minute. No, Mark, you 17 can't go first, because Celia never said a word for 18 the first time ever at any meeting, okay, and I 19 promised her that she would get the first say. 20 DR. FISHER: You requested. I didn't even 21 ask. So they just wanted to pep up, right, the 22 discussion. I thought the Panel was wonderful. I 132 1 must say that the conclusion I reached was that A as 2 written is pretty good. I think that was your basic 3 comments and that it has flaws, but that you weren't 4 recommending that it be rewritten. I did want to join 5 Mary and John and I think Steve in just giving a 6 little bit of the investigative perspective. 7 I agree with what everybody said. I think 8 that there is a frustration on the part of the 9 investigator that IRBs, especially social-behavioral 10 scientists, do not use the expedited review, that the 11 informed consent procedure is not participant 12 friendly, that most investigators view it as a legal 13 liability on the part of the university and that it, 14 in fact, hampers good research, limits the 15 randomization or breath and generalized ability of the 16 sample that can be tested. 17 So I think that that's a big problem. I 18 don't object as much in principle to the short form, 19 but I understand it is misused. But, in fact, it's 20 more participant friendly, even with non-English 21 speaking populations. In the real world, it's not 22 seen as an insult to them. But I can see as in 133 1 principle it could be perceived of as that way. 2 There is an overestimation of social- 3 behavioral risk and there is an emphasis by IRBs for 4 social-behavioral research to focus on the population 5 rather than the procedures themselves. So as soon as 6 it's a child or some other type of vulnerable 7 population, the notion is that population is 8 vulnerable, so no matter what you do, it is above 9 minimal risk. That bring me to the question that I 10 have. 11 Susan and I were talking about it, and the 12 extent to which the investigator's actions are part of 13 the equation in terms of determining risk, and I know 14 that there has been different OHRP guidance on that, 15 but I think this is a particularly important issue, 16 both with respect to the confidentiality issues that 17 if, in fact, you are putting in place procedures that 18 protect confidentiality for something that would be 19 great risk if it was relief, does it become a minimal 20 risk study? 21 And also I'm concerned because I realized 22 that as we were making recommendations about 134 1 reversibility for 406, remember that within 2 reversibility we were assuming that the second part of 3 reversibility is whether or not the investigator can 4 take actions that will prevent any long term type of 5 problems. So one of the things I think are important 6 is some type of clarification on the extent to which 7 an investigator's actions of minimizing risks, how is 8 that considered in terms of categorization of risk? 9 CHAIR PRENTICE: Okay. Thank you, Celia. 10 Mark, do you have some suggestions about how we 11 progress? 12 MR. BARNES: Well, I guess, what I want to 13 do is just ask you and Bern what it is that you want 14 to do. You know, without our guessing about it, I 15 want to know like what you think that you -- you know, 16 what's your idea? What do you want us to do? And 17 we'll probably do it. 18 CHAIR PRENTICE: Well, we know what we 19 want you to do, but we're trying to be democratic. 20 MR. BARNES: I don't want to. I just want 21 to be told. 22 CHAIR PRENTICE: And we're trying to let 135 1 you come up with what we want you to do by yourself, 2 so you feel better about it. 3 MR. BARNES: I would feel better if -- 4 PARTICIPANT: That's the worst part. 5 CHAIR PRENTICE: Yes. 6 MR. BARNES: I would feel better if I just 7 could do exactly what you want me to do, then I would 8 feel the best. 9 CHAIR PRENTICE: All right. All right. 10 I will tell you what I think and I'll certainly let 11 Bern speak on behalf of himself. I think it's 12 important to provide OHRP with an identification of 13 all of the issues that plague IRBs and investigators, 14 create regulatory burden, don't contribute to 15 protection to human subjects and there may be ways to 16 resolve some of those problems through a guidance 17 issued by OHRP and FDA and other, you know, entities 18 that enter into the equation. 19 OHRP is aware of some of these issues, 20 perhaps all of these issues, but they are not down in 21 the trenches. They are not down there operating at 22 the level of an IRB. They are not having to deal with 136 1 investigators constantly the way some of us are. So 2 I think it would be useful to have a subcommittee that 3 literally does what the Subpart B and Subpart C 4 Subcommittees are doing. They are analyzing the 5 regulations. And they are trying to figure out, okay, 6 what can we do to make this work better. 7 In the case of Subpart C, we have already 8 pretty much decided that, you know, we think there are 9 some short-term fixes that might work, but the long- 10 term fixes throw it out. I don't think that we think 11 that Subpart B needs to be thrown out. I think we 12 think it can be interpreted in such a way where we are 13 going to be able to facilitate pediatric research and 14 strike that right balance. 15 I think it is very clear from the Panel's 16 presentation and certainly I agree that Subpart A is 17 not broken. As a matter of fact, it's a masterfully 18 written regulation, but there are things that can be 19 fixed about it. You know, it doesn't require 20 revision, but there are certainly things that could be 21 fixed by the issuance of guidance. Maybe there are 22 areas that could be amended in Subpart A. For 137 1 example, the exempt categories. As David indicated, 2 that was a last ditch effort on the part of Charlie 3 under the gun to come up with some exemptions. That 4 seemed to work pretty well, but, you know, they are 5 still confusing. 6 You know, perhaps that could be expanded. 7 Perhaps the expeditable, you know, categories could be 8 expanded. We have done that once. Maybe we could do 9 it again. We certainly need guidance on waiver of 10 consent. I believe that there should be waiver of 11 consent, but, you know, how do you interpret that? 12 And you are quite right, I think, Dan, when you said 13 right. Okay. Everybody has the right to give 14 informed consent and IRBs are basically waiving that 15 right and how can they do that? Maybe it was Ada Sue. 16 It was one of the two that you said that. 17 You know, that's a very interesting 18 observation. We don't really address, you know, the 19 appropriate criteria for waiver of consent. We don't 20 understand what it means to do that. We need guidance 21 in that area and that doesn't require revision of the 22 regulations necessarily, unless we eliminated "right" 138 1 and just stuck with welfare, which perhaps could 2 happen. 3 So that's a lot of work. I think we have 4 to have a subcommittee, personally, to be able to 5 provide the necessary, you know, guidance to OHRP that 6 would result in a product that would resolve some of 7 these problems. It's not like some of the other 8 Panels where we said okay, let's get Mark to write a 9 paper, okay, and we did that for HIPAA. We did that 10 for HIPAA. I don't want to saddle you with the 11 rewriting of Subpart A, Mark. 12 MR. BARNES: I really appreciate that, 13 Ernie. 14 CHAIR PRENTICE: So that's what I think. 15 Yes, Bob? 16 DR. HAUSER: I agree with you. I would 17 like to -- I would hope that we could prioritize 18 things a bit. There is a lot that we could redo. 19 From a priority standpoint, I think, I would like to 20 work on those things that would improve human research 21 protection in Subpart A, particularly those areas 22 where there is some confusion and where the process 139 1 may actually expose people to unnecessary risk. The 2 second part which is important and there is no denying 3 that is to reduce the burden on the IRBs. 4 CHAIR PRENTICE: Bern, do you want to make 5 a comment? 6 DR. SCHWETZ: Yes. I think the 7 information that we heard this morning packaged 8 together with other things that you didn't even say 9 this morning that are floating around in the community 10 that are comments about Subpart A, I think what you 11 presented this morning is too rich and too important 12 to just let it sit as having been interesting 13 information that was presented to SACHRP. I think a 14 subcommittee is probably the best way to convert the 15 information that you've presented together with 16 perhaps other pieces that a subcommittee would fit 17 into that, into what I would envision just thinking 18 about it not a real long time, is some kind of a 19 spreadsheet that would organize the information in the 20 same context that perhaps Bob was suggesting. 21 There are a number of ways of categorizing 22 as you fill in the spreadsheet. One of them is 140 1 whether it relates to the burden or whether it relates 2 to really protecting human subjects. They are both 3 important. But it allows us to set a priority. 4 Whether or not it's an over-interpretation or whether 5 it's a lack of guidance and whether -- you know, there 6 are just a series of ways that you could, once the 7 information is in a spreadsheet, begin to analyze it 8 to help set priorities. 9 We collectively in the community, we all 10 own these problems. It isn't just you. It isn't just 11 OHRP. It isn't just OHRP and the federal community. 12 These are issues that we all have to deal with. And 13 I think this would be a good time to go through that 14 evaluation to identify priorities so that we don't 15 just work on one that perhaps could be fixed or worse 16 yet an intractable one and don't make any progress. 17 The community has higher expectations than that, and 18 I would certainly hate to waste the time of Panel 19 Members and Subcommittee Members on things where 20 everybody thought that the deck was stacked against 21 them, but they are working on an important issue. 22 So I think if we had a subcommittee work 141 1 on these issues to help set priorities and so that we 2 know clearly ahead of time whether or not this is 3 something that we can make some inroads in in this 4 system based on trust, and to make sure that we 5 understand whether they are simply a burden as opposed 6 to human subject protection or vice versa, I think all 7 of those are important things to look at as we 8 undertake working on some priorities to make a 9 difference. 10 I think the credibility of SACHRP, at this 11 point, is very high. I think if this kind of a task 12 is undertaken on Subpart A, that will also give a 13 credible reaction in the community to somebody working 14 on this as opposed to some other small group of people 15 who could perhaps do it as well, but I think this 16 would be a good way to do it. I think looking at the 17 question of whether or not this warrants a 18 subcommittee or some other approach to looking at it, 19 when you consider the -- you know, all the agencies 20 under the federal systems, buy into either Subpart A 21 or the FDA counterpart of it, makes it important. 22 It isn't just one agency that thinks this 142 1 is important. In the community that gets assurances, 2 you get an assurance per Subpart A and the rest you 3 decide whether or not you're going to go to the other 4 subparts. So those are the other two reasons why I 5 think this should be given a serious level of 6 attention, just because of the magnitude of the large 7 part of the enterprise force this is important. 8 CHAIR PRENTICE: Okay. It's lunch time. 9 So in the interest of lunch, I would be happy to 10 entertain a motion. Mark, you had your hand up first. 11 MR. BARNES: I told you I would do what 12 you wanted me to do. So I would like to move that we 13 have -- that we establish such a subcommittee, but I 14 would like to put on it a couple of caveats as part of 15 the motion. One is that the committee be very careful 16 in -- although, of course, as Bern said, identifying 17 the issues in a schematic or a systematic way would be 18 very, very important. That would be a contribution 19 itself and that would be the first priority, as I see 20 it, of the subcommittee. 21 But the caveat is that I think the 22 subcommittee, especially given all the discussion that 143 1 we had yesterday and today, the answer should not 2 always be immediately reflexively that we need more 3 specific guidance on everything. In fact, I think 4 that I would ask that the subcommittee in its work, 5 and this is part of my motion, to consider the extent 6 to which generalizing points instead of making points 7 more specific may be the way to go, one. 8 Two, whether guidance is actually 9 necessary from OHRP or whether the Committee itself 10 can look at the subcommittee itself, can look at 11 whatever the problems are and figure out ways itself 12 that these things either are not problems, have been 13 misinterpreted by IRBs, have been overly interpreted 14 by IRBs and investigators and lawyers and everybody 15 else who looks at these things. In other words, what 16 I don't want a committee to do in making the motion, 17 a subcommittee, is to come back to us with a list of 18 100 different guidance documents that are needed. We 19 don't need that. 20 We need identification and prioritization 21 of the real problems and then guidance documents, if 22 necessary, circumscribed, if detail is needed, or even 144 1 better generalized so that -- and one can, in fact, 2 have guidance documents that are more general rather 3 than more specific giving IRBs greater flexibility 4 rather than less flexibility. So that's my motion, if 5 that makes any sense. 6 CHAIR PRENTICE: Is there a second? 7 MS. KORNETSKY: I second. 8 CHAIR PRENTICE: There is a second. Okay. 9 Any further discussion? You want to discuss this 10 more? Okay. 11 DR. POLAN: Yes, I do. I have another 12 comment to add to that that I don't know that it has 13 to be in the motion. But I would very much like to in 14 some way assure that perhaps 50 percent of the 15 composition of that committee would be investigators. 16 That the perspective should be of people who run IRBs 17 and have SACHRP, but you're really dealing with 18 investigators. And if you can't make their lives 19 easy, then you've defeated the entire system. 20 CHAIR PRENTICE: Would you accept 25 to 21 50? 22 PARTICIPANT: 35? 145 1 DR. POLAN: I would accept 35 to 50. I 2 don't like the word substantial. 3 CHAIR PRENTICE: How about 30? 4 DR. POLAN: At least a third. Pardon me? 5 No, you see, I don't understand guidance. I just know 6 that you need to get the research done and you have to 7 make the funders work with the regulatory people and 8 somehow get the information. 9 CHAIR PRENTICE: Okay. Now, how about 30? 10 DR. POLAN: A third is very close, Ernie. 11 CHAIR PRENTICE: Okay. Thank you. 12 DR. POLAN: Thanks. 13 CHAIR PRENTICE: I want to get another 14 clarification. Are we talking about investigators who 15 are not IRB members or are you talking about people 16 who can be an IRB Member, but they are also 17 investigators? 18 DR. POLAN: I think they can be IRB 19 Members, but they have to be PIs. 20 CHAIR PRENTICE: They have to be PIs. 21 DR. POLAN: Active researchers and PIs on 22 current grounds. 146 1 CHAIR PRENTICE: Okay. 2 DR. POLAN: And they can be biomedical and 3 social researchers. I mean, I'm not trying to cut out 4 the social researchers. But I think there is a 5 difference between, at the risk of offending anybody, 6 there is the difference between running an IRB and 7 having active investigative research and applying for 8 funding, because the other part of the research is 9 trying to get it funded, which is not so easy. 10 Actually, if it makes anybody feel better, it's a lot 11 easier to get the IRB approvals than it is to get your 12 research funded. 13 CHAIR PRENTICE: Okay. How about we say 14 at least 30 percent? That gives us lots of 15 flexibility. That means we can go higher than 30 16 percent. 17 MR. BARNES: It's my motion. 18 CHAIR PRENTICE: At least 30 percent. At 19 least a third. All right. At least a third. All 20 right. At least a third. Now, that's 33 percent, so 21 it's at least a third, so we've gone up 3 percent. 22 Mark, you made the motion. You have to be prepared to 147 1 accept the friendly amendment. 2 MR. BARNES: I accept the one-third. 3 CHAIR PRENTICE: You accept one-third. 4 All right. Okay. Any other amendments? Yes, Susan? 5 DR. WEINER: Yes, of course, mine is the 6 mere image, which is -- I mean, you preempted my 7 comment to say that there really should be public 8 members on this group. I won't insist on a third, but 9 I would strongly recommend two, and that those two not 10 include me, and that we have, you know, them come out 11 of a different community, is my suggestion. But 12 that's my first point. 13 My second point is a more general point 14 about the burden of work that is coming before SACHRP 15 in its Panels today and in the coming months. I think 16 that, you know, I work a lot in the friends' community 17 and, you know, there is this joke that says that there 18 are 100 committees for 50 people. And so everybody is 19 serving multiple hats and I think we have to be very 20 careful about that, Ernie. So I would urge that this 21 committee have strong leadership, but also dip deeply 22 into the community. 148 1 CHAIR PRENTICE: Yes, okay. Thank you. 2 Mark, will you accept that? 3 DR. GYI: I just wanted clarification 4 involving community members. If we're looking at 5 Subpart A, Susan, I wonder how you see the role of the 6 community member impacting on what it is that the 7 subcommittee is supposed to be doing. I can see the 8 role of IRB Members and principle investigators, but 9 I'm a little concerned about what a community member 10 really brings to this particular group. 11 DR. WEINER: At least a perspective on 12 what they may be flexible about and what they may be 13 willing to yield or not yield, in terms of 14 requirements. I think it is important to have someone 15 whose job it is to be grounded in the values and the 16 end product. I think that, you know, it keeps 17 people's eyes on the process. 18 DR. GYI: I still have a difficult time 19 with that, so that would be one amendment or 20 modification that I would have a difficult time 21 digesting. 22 MR. BARNES: Well, I mean, it seems to me, 149 1 I think Susan is on to something that the structural 2 problem is that investigators and IRBs and IRB 3 administrators and members are the ones who tend to 4 understand the requirements, because they are actually 5 the ones who have to comply with the requirements. 6 It's not the subjects or the subject populations that 7 have to comply with the requirements. They can be 8 either beneficiaries or the victims of the application 9 of the requirements. So the structural difficulty is 10 finding community representatives who would understand 11 enough, having enough background about what is being 12 discussed and what is being cataloged and prioritized 13 to bring to bear that point of view. 14 So I'm not -- I actually agree with -- I'm 15 not disagreeing with you, Susan. I'm agreeing with 16 you. I'm just giving a little gloss to it and some of 17 my concerns. So I think that what I guess I would 18 accept is or ask Susan if she would accept would be 19 either that there be -- that the phrasing of it would 20 be that there be significant representation or that 21 there be one or two representatives or something like 22 that. At least one, Susan, one. 150 1 CHAIR PRENTICE: Okay. 2 DR. WEINER: There is an additional point 3 which is that when you have educated public 4 representatives, they bring resources that are not 5 just related to the discussion. They bring ideas from 6 other sectors that may solve the practical nature of 7 the problem. So you should have two. 8 CHAIR PRENTICE: Just a clarification, 9 Susan. How about a community IRB Member? They are 10 already on an IRB, so they would have some familiarity 11 with the regulations? 12 DR. WEINER: Why don't we leave that to 13 the discretion of the -- 14 CHAIR PRENTICE: Okay. So you're not -- 15 okay. 16 DR. WEINER: Okay. 17 CHAIR PRENTICE: So that's -- all right. 18 We've got to get to lunch, people. Yes? 19 DR. POLAN: Just one more comment. As you 20 constitute this committee, it would be nice to have 21 broad rep or ethnic and a socioeconomic background in 22 your community representatives. Because I was 151 1 astonished at the numbers of white males, white anglo- 2 saxons who sit on IRBs, that surprised me. I didn't 3 realize it was that high. And since we're all trying 4 to look at all segments of population in our studies, 5 it would be nice to have an actual understanding of 6 what communities need. 7 CHAIR PRENTICE: Okay. 8 DR. POLAN: Besides ours. 9 CHAIR PRENTICE: That's a good point. 10 MR. BARNES: Those data were only about 11 whites. They were not about anglo-saxons. I just 12 wanted to point that out and there is a big 13 difference. 14 CHAIR PRENTICE: Okay. All right. There 15 is a motion on the floor. I think we all -- do you 16 want to talk, too? 17 MR. ADAMS: I just wanted -- I do. I just 18 wanted to know if you wanted to cap the size of the 19 committee at 38? 20 CHAIR PRENTICE: Okay. All right. All 21 right. There is a motion on the floor. It has been 22 seconded. I assume that there is no further 152 1 discussion. So I would like to take a vote now. How 2 many people in favor of the motion? 3 (Show of hands.) 4 CHAIR PRENTICE: Any opposed? Any 5 abstentions? Motion carries. Now, listen, before we 6 adjourn for lunch, we're starting at 12:30. We have 7 a tight schedule for the rest of the afternoon. 8 You've got to eat lunch quick and get back here in 9 time to start. Okay? 10 ALL: Okay. 11 (Whereupon, the meeting was recessed at 12 11:42 a.m. to reconvene at 12:33 p.m. this same day.) 13 14 15 16 17 18 19 20 21 22 153 1 A-F-T-E-R-N-O-O-N S-E-S-S-I-O-N 2 12:33 p.m. 3 CHAIR PRENTICE: Welcome to the afternoon 4 session of SACHRP and we once again have a full agenda 5 beginning with our Panel discussion on the definition 6 of research versus what activities are not research. 7 And we have four Panelists and for the benefit of the 8 Panelists, the way we work this is that I will 9 introduce you very briefly. I will not go through 10 your bio in the interest of time. I will ask you to 11 come up to the computer and present. Please, stick 12 within the allocated time, so we can remain on time. 13 And then at the end of the four presentations, we 14 would ask all of the Panelists to assemble up at the 15 table there and respond to questions. 16 So the first presenter will be Dr. Michael 17 Carome. Dr. Carome is a Commander in the U.S. Public 18 Health Service. He is also the Associate Director for 19 Regulatory Affairs at OHRP and he is well-known, I 20 think, to everybody here. Michael, welcome. We're 21 looking forward to what you have to say. 22 DR. CAROME: Well, good afternoon and 154 1 thank you. My role just to quickly lay out some 2 background, first to make sure everyone is aware of 3 the regulatory definition of research, which sets the 4 framework for this, and then to give you a quick 5 overview of how OHRP views the relationship between 6 research, human subject research as defined by the 7 regs and other activities that are called something 8 else. 9 So I'm going to quickly walk through how 10 our regulations apply to a particular activity, give 11 you the definition of research and talk about then our 12 view of the relationship between human subject 13 research and other activities. And those other 14 activities might be clinical practice, quality 15 improvement activities, public health or public health 16 surveillance activities, as well as other activities 17 you could think of. 18 I think the core message I want to leave 19 you with is that there is overlap between research as 20 defined by the regulations and other activities which 21 go by another name, and that you want to avoid getting 22 into the thought process of always saying these are 155 1 mutually exclusive domains. It's important to 2 recognize that different domains can overlap. So 3 that's the core measures I want to leave you with. 4 And so the regulations we're talking 5 about, as everyone knows, are at 45 CFR Part 46, and 6 activities covered by these regulations are those at 7 research involving human subjects conducted or 8 supported by our Department, that is not otherwise 9 exempt, and they also apply to non-exempt human 10 subjects research conducted at an institution that 11 holds an institution that holds an assurance that 12 applies to those activities, whether they voluntarily 13 extended their assurance to all research, regardless 14 of sponsorship. 15 When assessing a particular activity, we 16 believe it is important to ask three questions and to 17 ask those questions in this order. First, does the 18 activity involve research? Second, does it involve 19 human subjects, if it is research? And if it's human 20 subject research, is the activity exempt? And so, 21 first, what's the definition of research, and this was 22 gone over this morning, but it is a systematic 156 1 investigation, including research development, testing 2 and evaluation, designed to develop or contribute to 3 generalizable knowledge. 4 And I'm going to disappoint some of you 5 and not go on further to define some of the terms that 6 are in this definition. So systematic investigation, 7 design, generalizable knowledge, those are terms that 8 are not further defined and for the sake of time, I'm 9 not going to further define them. 10 PARTICIPANT: You can write a guidance 11 document for us. 12 DR. CAROME: We'll do that later. It's 13 important to recognize that that's not the end of the 14 definition in the regulation. The definition goes on 15 to say that activities which meet this definition 16 constitute research for the purposes of the 17 regulations whether or not they are conducted or 18 supported under a program which is considered research 19 for other purposes. For example, some demonstration 20 service programs may include research activities. 21 What this says is there may be activities 22 that are not labeled research, they are something 157 1 else, they exist for some other purpose, but if they 2 involve some component that meets the definition of 3 research, then it's covered under the definition and 4 under the regulations. So what's the relationship 5 between that definition of research and other 6 activities? And this is my key point, it can overlap. 7 Well, this research can overlap with other 8 types of activities, whether that activity be X 9 activity, innovative clinical care, clinical practice, 10 public health surveillance activities, public health 11 activities, quality improvement activities, it's 12 important to recognize sometimes these two domains 13 occur at the same time. When that is the case, the 14 regulations apply depending upon the answer to those 15 other two questions. 16 Now, it's important to recognize how does 17 one go about determining whether an activity involves 18 research and I don't have time to go into this in 19 great detail. It is important to recognize that 20 whether something involves research is not the 21 dependent upon the risk benefits of that activity. 22 The risk benefit analysis follows the determination of 158 1 whether something involves human subject research. 2 It's not defined by any specific type of research 3 design or procedure. 4 You don't necessarily need randomization 5 or control groups to have an activity be researched. 6 The presence of such factors might lend credence to 7 something being research, although the presence of 8 randomization may not be a definitive characteristic 9 of something, of an activity as being research. It's 10 not dependent upon the testing or evaluation of an 11 experimental, innovative or new procedure, treatment 12 or intervention. Research can involve comparing 13 standard interventions, known interventions, but 14 trying to determine which one might be more effective 15 than another. 16 It's important to recognize that research 17 may not be the primary objective of a particular 18 activity. Publication is not a definitive feature of 19 research. There is research done that is never 20 published and there's lots of publications out there 21 that have nothing to do with research. You know, the 22 newspapers publish every day. What is described in 159 1 the newspaper doesn't mean the activities described 2 were research activities. 3 So after you evaluate whether something 4 involves research, you next have to define whether or 5 not it involves human subjects and the definition was 6 reviewed this morning. So here to further 7 characterize the Venn diagram, you have within the 8 research domain those that involve human subjects, and 9 again there can be overlap with the other activities 10 at times. Then the last question is whether the 11 activity is exempt. And so within the human subjects 12 research domain a certain amount of that is exempt, 13 but you have this overlap of non-exempt activities 14 that can overlap with other activities. 15 Last, I'll just make a few points that are 16 made in the Belmont Report. And the Belmont Report 17 talked about the boundary between research and 18 clinical practice or practice. It stated in part that 19 the distinction between research and practice and we 20 can substitute perhaps X activity is blurred partly 21 because both often occur together. 22 By contrast, the term research designates 160 1 an activity designed to test the hypothesis, permit 2 conclusions to be drawn and thereby develop or 3 contribute to generalizable knowledge. Researchers 4 usually, but perhaps not always, describe in a formal 5 protocol that sets forth an objective and a set of 6 procedures designed to reach that objective. 7 Research and practice may be carried on 8 together when the research is designed to evaluate the 9 safety and efficacy of a therapy. This need not cause 10 confusion regarding whether or not the activity 11 requires review, and that means review by an IRB. The 12 general rule is that if there is any element of 13 research in an activity, that activity shall undergo 14 IRB review for the protection of human subjects. 15 So in conclusion, what I've tried to do is 16 quickly present to you the definition of research, how 17 you should determine whether an activity is covered by 18 the regulations by asking three questions. Is it 19 research? Does it involve human subjects? Is it 20 exempt? And I have tried to communicate how human 21 subject research activities can overlap with other 22 activities and it's important to recognize that 161 1 overlap. Thank you. 2 (Applause) 3 CHAIR PRENTICE: Thank you, Michael. Our 4 next speaker is James Hodge, who is an Associate 5 Professor and Executive Director of the Center for Law 6 and the Public's Health at the Johns Hopkins Bloomberg 7 School of Public Health. Welcome to SACHRP, James. 8 MR. HODGE: Thank you very much, Dr. 9 Prentice. Let me get myself situated with the 10 microphone here, as I may wander around the room just 11 a little bit as I'm used to in my lecture classrooms 12 at Hopkins. Is that picking up well? Can everybody 13 hear me all right? Dr. Prentice, Dr. Schwetz, thank 14 you for this invitation to join you here today. Dr. 15 Carome as well, thank you. Thanks to OHRP for their 16 input on the project that I'm about ready to discuss 17 with you and just thanks to everybody for their 18 interest in this particular topic. 19 Let me, if I may, try to set the 20 parameters of what we accomplished and that speaks to 21 a group of folks, not just myself, but additional 22 scholars, including my colleague at Georgetown and 162 1 Johns Hopkins, Professor Larry Gostin, and a full 2 panel of folks that was assembled by the Council for 3 State and Territorial Epidemiologists. 4 Back this prior year, we just finished our 5 work in May of 2004 to essentially look at the key 6 question. The one question that is riddled public 7 health authorities for ever since the Common Rule, 8 actually, and now with the HIPAA Privacy Rule 9 routinely what are the distinctions between public 10 health practice activities and public health research? 11 The implications of that, I think, we heard about 12 earlier today. I will give you some additional 13 interest or ideas as to why this is such an essential 14 question and one of which we are looking forward to 15 input from this particular Committee. 16 One of the principal objectives of this 17 project was to produce a report that would begin to 18 assess the key distinctions. Not stop short of 19 recognizing the difficulties that this question causes 20 to public health practitioners at the federal, state, 21 local and tribal levels, but to take it a step further 22 and say there are key distinctions, here they are and 163 1 here is how we can utilize those to help draw clear 2 ideas between what is public health practice and what 3 may constitute human subjects research. 4 So one of the ways that we attempted to do 5 that was to assess the legal and ethical environments 6 to which underlie these particular practices. Public 7 health practice is, as you'll see in a few minutes, 8 driven by very different legal frameworks than human 9 subjects research. I'll explain that in just a 10 moment. We wanted to clarify the existing definitions 11 of public health practice and public health research. 12 What we are not attempting to do as other scholars 13 have suggested is to argue for sheer changes in the 14 existing definitions of what we see in the Common Rule 15 or the Privacy Rule. 16 We're accepting those definitions for the 17 moment and simply trying to draw distinctions between 18 them. We tried to provide meaningful cases on 19 practice and research. I believe the Members of the 20 Committee and certainly in the back of the room you 21 have a full copy of our report, 10 cases laid out 22 there based on very real facts showing the type of 164 1 problems that arise when we look at public health 2 practice versus public health research, and I think we 3 do that quite demonstratively trying to provide some 4 key lessons there, and then we attempt to make 5 distinctions. 6 There are some major assumptions here at 7 work as well. This report, our checklist, the work 8 that we have done is not for every type of public 9 health practice nor is it about every type of issue in 10 regards to human subjects research done by public 11 health authorities. In fact, while we are accepting 12 the existing legal environment and principles, we are 13 very much focused on only what we call or consider 14 core public health activities. And our core public 15 health activity definitions, I'll get to in a minute, 16 is very much about where and under what circumstances 17 we acquire, use or dispose identifiable health data 18 for these various activities. 19 One of the things that underlies this 20 report that I will hint about, that I want to make 21 sure you know, we also recognize the difficulties that 22 some entities are having in applying the Common Rule 165 1 standards in conjunction with the Privacy Rule 2 standards, because the Privacy Rule does take some of 3 what we see in the Common Rule to a larger level. Our 4 report attempts to develop a checklist or principles 5 that work for both major types of federal laws. 6 In that respect, I think you will see how 7 we weave that in here. So there are very significant 8 and key differences in the legal support for public 9 health practice and research. This classification of 10 these activities between research and practice without 11 doubt leads to multiple complications. Several of the 12 cases note how the same activity described in the very 13 similar fashion is quite simply characterized as 14 practice in one setting and research in the other. 15 The very same activity in the very same time period, 16 this leads to varying complications. 17 There are varying legal and ethical 18 standards for performing human subjects research that 19 is contrary to public health practice, and those 20 issues arise, of course, in the Common Rule and 21 Privacy Rule. Widespread variation in these existing 22 models and I can assure you there is extensively 166 1 widespread application of how entitles from 2 Institutional Review Boards to public health 3 practitioners to the best of that authority in some 4 cases actually make these distinctions, at least to a 5 host of problems for public health goals in 6 particular. 7 You know one of the things that Mike 8 mentioned and he's very right about this, one of the 9 problems, as the Belmont Report notes, is that this is 10 difficult to make some of these distinctions, because 11 public health practice, specifically practices in 12 which we acquire use and disclose identifiable health 13 data, can kind of look like or feel like human 14 subjects research. There is no doubt about it. 15 One of the things that we are attempting 16 to do in this is to recognize that there are these 17 very clear similarities. They are conducted in 18 manners consistent with the idea of protecting 19 individuals. Public health practitioners acquire 20 these data for the purposes of protecting the public, 21 but also to respect individual rights, just as human 22 subjects researchers may. They are justified 167 1 completely as laudable, communal activities that 2 further the public good in many cases. 3 That's specifically true, as you see the 4 definitions of public health practice and research as 5 we've defined them. But in reality, as we all know, 6 public health practice is not human subjects research. 7 Be definition, it's not meant to be human subjects 8 research. As we note in this report the different 9 methodology, the different objectives, the legal 10 support underlying these types of activities is very 11 different, even the ethical frameworks from principle 12 to bio-ethics to principles of public health ethics, 13 very different. And the design of many of these types 14 of activities is quite, quite unique. 15 What we attempted to do with the report is 16 to set what we think is a state of the art 17 conceptually strong definition of public health 18 practice that is distinct from public health research, 19 and here it is. We consider it the collection and 20 analysis of identifiable health data by public health 21 authority, and there I'm using the term very 22 consistent with what we might see in the HIPAA Privacy 168 1 Rules. For the purposes of protecting the health of 2 a particular community where the benefits and risks 3 are primarily designed to accrue to the participating 4 community. 5 Contrast that if you would with the 6 definition that we build forth public health research. 7 It's the collection analysis of identifiable health 8 data. There is your common link. By public health 9 authority, yes. Public health authorities do engage 10 in research, there's no doubt about that. But the 11 primary purpose is to generate knowledge consistent 12 with Mike's definition as found in the Common Rule 13 that will primarily benefit those beyond the 14 participating community, who bear the risks of 15 participation. 16 Those two definitions are so critical to 17 our analyses, because we're really trying to show the 18 key conceptual differences and hopefully you see that 19 built into these various ideas I'll provide here in a 20 second. The existing approaches, I need not summarize 21 in great detail here. The report does a nice job. 22 Governmental, private sector, academic, we looked at 169 1 a lot of these approaches and, quite frankly, there's 2 just a few general observations. I can tell you about 3 them. 4 They are vastly different. What IRBs and 5 institutes, like mine at Johns Hopkins, used to 6 distinguish public health practice from public health 7 research is far different than what the CDC uses at 8 present. It's far different than what many scholars 9 have suggested to use. One of the things that we 10 definitely found in regards to our scholarly research 11 on this particular issue is that these approaches 12 really are quite divergent. And that alone has led to 13 lots of problems for lots of folks in public health 14 and in human subjects research areas. 15 The legal frameworks I mentioned earlier. 16 Public health practice very much grounded in the 17 constitutionally-approved authority of Government to 18 protect the public safety and general welfare. Let me 19 share with you the common message that I hear from 20 public health practitioners consistently, and this is 21 not just through this report, but in the essence of 22 what I hear from them in working with them all the 170 1 time. 2 What we do to practice public health in 3 our particular state, tribal Government or local 4 Government is constitutionally-approved. It's our 5 duty to make these things happen. We don't have a 6 choice. We can't be stopped in our tracks by some 7 Institutional Review Board that wants to assess 8 whether these activities go on as research or 9 practice. 10 I'm not suggesting I share this view. 11 What I'm suggesting is that's a very common 12 perception, that what we have to do in surveillance 13 and Epi investigations and otherwise must go on 14 without impediment. It's a constitutionally mandated 15 issue in many of the states. 16 Contrast that with public health research 17 as we know it, any type of research, human subjects 18 related, grounded very much in the principles of the 19 Federal Common Rule, which might very nicely summarize 20 for us the focus on protecting individuals in the 21 pursuit of knowledge. Don't doubt for a second though 22 that public health practitioners are very much 171 1 focused, as well, on protecting individuals. They are 2 not out in the conducting of these activities through 3 which they acquire the sometimes highly sensitive 4 identifiable data to harm individuals. They are 5 collecting it to protect the public's good and their 6 intent is to protect the individual as to the extent 7 possible in every case. 8 The Privacy Rule, of course, as we know 9 provides very different rules for the disclosure of 10 protected health information to public health 11 authorities for public health purposes as contrasted 12 with to public health authorities for research 13 purposes. One of the things that underlies that in 14 the Privacy Rule and the Common Rule that is 15 recognized routinely by public health authorities, 16 there is a strong incentive to characterize the 17 activity that you are acquiring that data for as 18 public health activity. 19 Why? Because you get virtually a carte 20 blanche in the Privacy Rule. Public health 21 authorities get the data. That is basically the 22 answer. If it's research that underlies the activity, 172 1 oh, be sure there's a host of things that you must 2 show before you can get the data without written 3 authorization pursuant to the similar standards that 4 we see in the Common Rule. 5 Now, again, as I mentioned, we could have 6 stopped short there and said those are a lot of 7 difficult issues and we have laid all those issues 8 out, so folks have a good clear sense of what we mean 9 in regards to how these issues play out in the 10 research versus practice arena for public health, but 11 we didn't want to stop short. We tried through this 12 report to assess and provide a functional framework or 13 a series of guidelines to help resolve these types of 14 issues. Perhaps a national standard, which has not 15 been approved by any particular federal agency, I want 16 to assure you, but is one that we think helps to 17 resolve the easy and the hard cases. 18 We're going to present Stage 1, Essential 19 Features of Public Health Practice and Research, Stage 20 2, Enhanced Guidelines and a checklist. The two stage 21 analysis is really geared towards accomplishing and 22 providing guidance to resolve what we call the easy 173 1 cases. There are so many examples that we saw in our 2 research that we all, I think, know about. They are 3 just easy cases. This is not human subjects research 4 and it can't be human subjects research because of 5 this element, but the Institutional Review Board 6 doesn't pick up on that or the public health 7 practitioner or the person funding the activity 8 doesn't pick up on that and, as a result, there is 9 misclassifications. 10 But then after those, 90 percent of the 11 cases, there is the 10 percent of those cases that are 12 really hard. They are not simple. They do involve, 13 as Mike was saying, bundled activities, a single 14 activity that might have research and practice within 15 it. That's what we try to address with our Stage 2 as 16 well. 17 So let me give you the approach here very 18 quickly and then I will be happy to defer to any 19 questions and comments. What we try to do is assess 20 what the fundamental foundations of public health 21 practice are versus public health research. What are 22 the foundations? What are the things that are 174 1 essential to public health practice that if you're 2 engaged in these things for these reasons, you can 3 feel at least confident under an easy case scenario 4 that you have a public health practice activity. 5 Well, first of all, public health practice 6 involves in many cases very specific legal 7 authorization at federal, state or local levels. The 8 state may say to its public health entity you will 9 collect these data. You will do it for the purposes 10 of surveillance for HIV or other types of conditions. 11 This is a surveillance public health activity. Some 12 states like New York in some cases actually go so far 13 to spell it out in the statute that the state 14 considers that activity to be public health and not 15 research. 16 It includes a corresponding Governmental 17 duty to perform the activity to protect the public's 18 health. The Government can't step away from this. 19 The legislatures require these activities to occur. 20 It involves direct performance or oversight by a 21 Governmental authority in public health or its 22 authorized partners, who recognize in the Privacy 175 1 Rule, private sector entities can be public health 2 authorities and accountability to the public for its 3 performance. 4 A couple of additional foundations. It 5 can legitimately involve persons who did not 6 specifically volunteer to participate. Your data by 7 public health authorities can be collected without 8 your informed consent, without any written 9 authorization for a public health purpose and that's 10 legitimate. It's constitutionally permitted. It's 11 been approved by the Supreme Court as we all know and, 12 to say the least, that's a foundation of practice 13 supported by principles of public health ethics, as 14 well, but I won't go into great detail, but have a 15 very different feel to them, very different dimension 16 than bylaw faces. 17 Foundations of human subjects research. 18 These are core things, things you have to find before 19 you have identifiable data being used for human 20 subjects research. It involves living individuals. 21 It involves, in part, identifiable private health 22 data. It involves research subjects who voluntarily 176 1 participate or participate with the consent of their 2 guardian absent a waiver, of course, supported by 3 principles of bioethics that focus on individual 4 interests while balancing the communal value of 5 research. 6 Now, when we went beyond those easy cases 7 and, believe me, a lot of these issues are resolved by 8 those simple principles we suggested well, what are 9 the criteria that we would use to help distinguish the 10 harder cases? Well, there are several that we found 11 and we looked at them very systematically and said 12 well, these really aren't helpful. Mike has mentioned 13 some and I think other commentors will say, as well, 14 the performance, publication, urgency, funding and 15 data collection methods we found to be inconclusive in 16 every case as to whether it helps distinguish practice 17 versus research in the public health setting. 18 Performance. Who is performing the 19 activity? Who cares? It might be a private sector 20 entity in a particular state that's under contract 21 with a public health authority at the local level to 22 do a very public healthy thing. It might not look 177 1 like public health on first glance, because they are 2 not a Governmental authority, but the Privacy Rule 3 very clearly suggests that entity can be engaged in 4 public health surveillance. 5 Publication. It does not matter whether 6 there is an intent to publish or not. It's just 7 inconsequential. Public health authorities routinely 8 publish their results in JAMA and MMWR and other 9 places, so do researchers. 10 Urgency. We saw in a few examples, some 11 in the Governmental level, that a sense of urgency 12 might substantiate a distinction or an activity being 13 practices versus research because of the sheer 14 necessity of moving quickly. We don't think so. 15 Urgency may require expedited review by an IRB. It 16 may require moving quickly in certain instances, but 17 it's not a distinguishing factor. 18 Funding. In the end the Common Rule is so 19 universally applied and because one of our primary 20 assumptions is that we do have a Governmental public 21 health authority involved in the activity, funding, as 22 far as the source of the funding, is not a linchpin to 178 1 distinguishing practice versus research. 2 And the data collection methods. The 3 tools that we use to collect the data might look very 4 similar in public health practice versus research, not 5 a good distinguishing criteria. 6 So let me be very quick, if I may, with 7 what we consider to be the enhanced guidelines, 8 general legal authority, specific intent, 9 responsibility, participant benefits, what we call 10 experimentation and subject selection. And again, I 11 will be quite brief here. 12 General legal authority. You know, you're 13 going to find in many cases specific legal authority, 14 the legislature or the Department of Health spelling 15 out the exact legal authority to conducting a public 16 health practice activity and sometimes you're not 17 going to find that. You're going to just find general 18 legal authority. The State Department of Health is 19 authorized to do what it needs to do to prevent 20 communicable diseases. 21 In some cases, that might lend towards a 22 distinctive factor of a particular activity being 179 1 viewed as practice, because there is a general legal 2 authority that supports that particular finding, but 3 that's not conclusive, of course. A public health 4 authority may use research methods to accomplish 5 things under that legal authority as well. 6 Specific intent. CDC very clearly thinks 7 intent is a key element to making these distinctions 8 and we tend to agree with that. Unfortunately, the 9 intent driven criteria that CDC uses is too 10 generalized based on our assessment. We're suggesting 11 here to use something much more key to the specific 12 elements of intent for research and practice. So the 13 intent of public health research is to test a 14 hypothesis and seek to generalize the findings or 15 acquired knowledge beyond the activity's participants. 16 Contrast that with the intent of public 17 health practice, assuring the conditions in which 18 people can be healthy, that Institute of Medicine 19 definition we know well, through public health efforts 20 that are primarily aimed at preventing known or 21 suspected injuries, diseases or other conditions or 22 promoting the health of a particular community. 180 1 What we're suggesting here with the intent 2 criteria is very important. We're suggesting it's a 3 critical part of the assessment, but it's not the only 4 criteria. It's not even the primary criteria. And if 5 you're going to use intent as one to distinguish these 6 activities, you have got to be able to specifically 7 show that that's what underlies the activity that 8 you're engaged in, because if it's closer to research, 9 you probably have research on your hands. 10 Responsibility. What we're here assessing 11 is that in many cases, of course, you see 12 responsibility for the health, safety and well-being 13 of research participants falls on somebody. There is 14 an identified individual. Typically, we call them the 15 principal investigator and that person really is 16 responsible for the welfare. No matter what other 17 researchers, underlying that principle investigator, 18 do it's the PI at the end that could be very much held 19 accountable. 20 Responsibility for individuals' welfare in 21 a practice activity might not fall on any particular 22 individual. The Government may hold that responsibly. 181 1 It may be a Governmental function or duty to protect 2 that person, but you might not be able to pinpoint and 3 say that person there, they are the one who is 4 responsible, because they quite simply may not have 5 that sort of legal or ethical responsibly in a 6 practice scenario. 7 Participant benefits. Whenever additional 8 risks are imposed on participants to make the results 9 generalizable beyond the participants themselves, the 10 activity can be classified or it should be classified 11 very much as research. Public health practices are 12 premised on providing some benefit to persons though 13 or the population of which they are members. These 14 benefits can be very limited due to failures in design 15 or implementation. The practice activity may not go 16 as planned. It's an unfortunate circumstance, but the 17 idea here is we're practicing to provide a benefit to 18 the people who are impacted by the practice 19 themselves. 20 Experimentation. Research, as Dr. Carome 21 was mentioning, typically involves introducing 22 something very non-standard to the research subjects 182 1 or their identifiable health data, something that's 2 new. And if you're not applying something new or 3 doing something different with these data, question 4 the value of the research involved. Practice is not 5 driven by that. There is no experimentation 6 underlying practice, although it certainly can be 7 innovative. 8 Practice is very much geared towards 9 applying standard, known and accepted and proven 10 interventions to address these suspected particular 11 public health problems. I'm not trying to say that 12 public health practice doesn't deviate from these 13 things. What I am trying to say is that it's not the 14 focus to apply something very experimental to the 15 particular data at hand. 16 Subject selection. Research subjects can 17 be selected randomly, so that the results are 18 generalized to a larger group or to reduce bias. This 19 is kind of an underlying underpinning of research. 20 Public health practice doesn't involve such random 21 selection. These people are self-selected individuals 22 with diseases, conditions that participate to seek 183 1 benefits in the programs themselves or on behalf of 2 their community to benefit them as well. 3 So the checklist that you see, in brief, 4 provides a series of steps that you can utilize to, 5 hopefully, begin to make these distinctions. Check 6 our key assumptions. If these key assumptions don't 7 apply to the activity you're doing, the checklist is 8 not about you and that is an important thing to 9 mention. 10 Assess the foundations of public health 11 practice is Step 2, part of the easy case scenario. 12 If you survive that particular step, if you have not 13 been told to stop because you definitely have a 14 practice activity, go to Step 3, assess the 15 foundations of human subjects research. Again, if you 16 survive this step, you have a hard case, what we call 17 a hard case. It's not so easy. Step 4, consider the 18 enhanced guidance, but try to introduce that in a very 19 pithy sort of way in the checklist to summarize what 20 the much, much longer report does in that regard. 21 So in brief summary, I think 22 distinguishing public health practice from research is 184 1 not always easy, but nor is it always hard despite the 2 fact that many institutions and other places do get 3 this calculation wrong on many occasions. Varying 4 legal standards, methodologies and frameworks 5 certainly complicate the distinctions, and the public 6 health community all typically agree that 7 clarification is needed. 8 We're not suggesting with this report or 9 anything I may mention here today that this single 10 guidance resolves all of the cases under existing 11 definitions, but I think this sort of guidance does 12 provide clarity where clarity counts, resolving the 95 13 percent of the cases that still result in several 14 complications. 15 You have a full copy of our report, I 16 think, in your binder, but additional copies are 17 available online there. And with that, I thank you. 18 (Applause) 19 CHAIR PRENTICE: Thank you very much, 20 James. Our next presenter is Dr. Mary Ann Baily and 21 Mary Ann is an Associate for Ethics and Health Policy 22 at the Hastings Center. Welcome, Mary Ann, to SACHRP. 185 1 We're looking forward to your presentation. 2 MS. BAILY: Let me see if I can get this 3 on. Okay. Well, like Jim Hodge's presentation, this 4 presentation is also based on a project. Is this 5 working? Okay. You have a handout, which describes 6 the project in a little more detail and gives you a 7 list of the task force members. 8 Unlike Jim's project, however, our project 9 is still in progress and we have had three meetings so 10 far and we are working on a draft report to be 11 presented at the fourth meeting, which will be held in 12 November. But since our project isn't done yet and, 13 in fact, we have another year of it, my remarks here 14 reflect the deliberations of the project members, but 15 I want to emphasize that I am not speaking for the 16 project members, at this point. 17 All right. Now what do I do, press here? 18 Okay. Now, the impetus for the project, the reason 19 that AHRQ gave us the money was the fact that people 20 were having problems in applying the research 21 regulatory framework to QI activities, and our mission 22 was to step back and do some fundamental conceptual 186 1 analysis of the nature of the problems. What are we 2 trying to do here and what would be the best way to go 3 about it? 4 And that's why these two questions capture 5 what we were trying to do in the project. What makes 6 the QI activity ethical or unethical and what social 7 arrangements should be in place to ensure that QI 8 activities are conducted in an ethical manner? Okay. 9 The project is not about all the activities that could 10 possibly improve the quality of health care. The 11 boundaries are fuzzy, as Mike Carome indicated, but 12 basically it's about activities that attempt to bring 13 about positive changes in the clinical care setting in 14 a systematic way, a data guided way. In other words, 15 it's that kind of quality improvement that can be 16 viewed as somewhat research like. 17 It's a broad range of activities of 18 varying degrees of complexity and so on. It includes, 19 for example, retrospective data reviews, limited 20 innovations in work patterns in small teams and large- 21 scale data analyses of prospectively determined 22 interventions across multiple sites. 187 1 Now, the process that we engaged in in 2 this project was, first of all, we compared and 3 contrasted the three kinds of activities listed on 4 this slide, clinical practice, QI and research. We 5 then looked at the generally accepted ethical 6 requirements for research and asked to what extent 7 they did and didn't apply to QI. 8 We drew conclusions from that analysis 9 about the kind of ethical review and oversight of QI 10 that would make sense on a conceptual level, and the 11 final step in the project will be to explore whether 12 the approach that we come up with can be implemented 13 on a practical level and what it would look like. 14 Now, given our mission, which was to take 15 a step back and look at the conceptual framework, we 16 felt that we had to allow ourselves to contemplate 17 changes in the regulatory framework. At the same 18 time, our mission was to be of practical use, to draw 19 conclusions and make recommendations that would help 20 people manage the problems. Since changing 21 regulations is not simple, we would like to find ways 22 of managing QI that are as compatible as possible with 188 1 existing structures. 2 Okay. Now, let's turn to these 3 overlapping categories and first, let's look at 4 clinical practice versus QI. First, I should note 5 that we are using the term clinical practice here in 6 a somewhat broader sense than the Belmont Report used 7 it, which the Belmont definition emphasized the 8 physician/patient relationship in the individual 9 patient and it focuses on interventions in individual 10 patients. 11 Now, for care delivered in organizational 12 settings like hospitals and managed care plans and, in 13 fact, even for an individual physician in solo 14 practice, a significant part of clinical practice is 15 about having systems in place for managing the care of 16 many patients. So perhaps I should say that we're 17 really talking more about clinical care delivery, not 18 just clinical practice in that narrow sense. 19 A key insight in our project has been the 20 importance in a clinical care setting of managing 21 change throughout health care. There is constant 22 adaptation and innovation going on in health care, 189 1 constant adaptation to changing background conditions, 2 especially changing technology. Now, managing this 3 constant change is a core responsibility of the people 4 and the organizations that make up the health care 5 delivery system. Whatever is going on in a particular 6 clinical setting is the product of thousands of small 7 and large decisions about handling change that could 8 have been made somewhat differently in this setting 9 and have been made differently in other settings. 10 Traditionally, in the health care system 11 such adaptations have been introduced on a "just do 12 it" basis, so to speak, without careful attention to 13 all of their effects. The QI movement, as it is 14 sometimes referred to, is about making this process of 15 continual adjustment more self reflective and 16 systematic and, thus, ensuring that it does, in fact, 17 produce positive change. In other words, when you add 18 a systematic experiential learning dimension to 19 managing the ongoing change that is intrinsic to 20 clinical practice, you have QI. 21 Now, turning to QI versus research. As we 22 have said repeatedly today, research is a systematic 190 1 investigation with the goal of generalizable knowledge 2 and the knowledge gained is implemented or not by 3 others. So in order to get the social benefits, 4 dissemination is essential and publication is, 5 therefore, important. 6 QI on the other hand has the goal of 7 immediate local improvement. Nevertheless, it is a 8 systematic investigation guided by data and it does 9 use methodologies that are sometimes quite similar to 10 those used in research. That means that combining the 11 results of QI activities in different settings can 12 yield insight about how to bring about positive 13 change, which can be a kind of generalizable knowledge 14 or at least that the kind of experience that people 15 outside the local setting can learn from. 16 Therefore, dissemination of QI results can 17 be socially useful and QI practitioners ought to have 18 the possibility that the results will be worth 19 disseminating in mind when they begin the activity, 20 and the intent to publish does not distinguish QI from 21 research. I don't have to belabor this point since 22 it's been made already for public health. 191 1 Okay. Now, turning to some of the 2 conclusions we came to. A critical difference for 3 ethical review and perhaps the most critical 4 difference is this one. QI activities are strongly 5 oriented toward the goal of immediate local change. 6 They take place in the clinical setting. They are led 7 by the people who work in that setting and they 8 typically incorporate rapid feedback of results to 9 bring about positive change in that setting. That 10 change is often incremental and unlike in research in 11 which there is a fixed protocol that's implemented for 12 a time period that can be years long, QI methods not 13 only allow, but expect modifications in the initial 14 protocol as experience accumulates over time. 15 The term continuous quality improvement, 16 which is used almost interchangeably with the term QI, 17 highlights the fact that QI is not so much the 18 implementation of discreet projects as it is an 19 ongoing process of continual self-conscious change. 20 That means that QI is an integral part of the ongoing 21 management of the system for delivering clinical care, 22 not an independent knowledge seeking enterprise. In 192 1 other words, it's part of normal health care 2 operations. 3 Clinical research with human participants, 4 the kind of research that is the most similar to QI, 5 is often located in the clinical care setting, but it 6 typically has a different relationship to the health 7 care delivery system. Usually, it's conceived, funded 8 and managed as discreet projects led by a principal 9 investigator who takes primary responsibility for the 10 design and conduct of the project, and the resources 11 for clinical research come from outside the 12 organization in which it is carried out or if from 13 inside, from a separate research budget, not from the 14 clinical care resources. 15 When the research participant would be 16 receiving health care in any case, the provision of 17 therapeutic care is intermingled with the search for 18 generalizable knowledge about a physical process or a 19 treatment effect. In this dual role, physician 20 researchers have a potential conflict between concern 21 for patients and concern for the success of the 22 research, which is why there is such an emphasis on 193 1 protecting human subjects within research. 2 A second critical difference for ethical 3 review is that a fundamental principle of research 4 ethics is that research is optional. It's a separate 5 enterprise. That's why there is such an emphasis on 6 consent within the research regulations. 7 The ethical paradigm for QI is different. 8 Health care professionals and health care managers and 9 health care organizations owe it to the patients as 10 part of their mission to serve the patient's interest 11 to be constantly trying to improve their practice. QI 12 is a particularly effective way to improve practice 13 and, therefore, it is not completely optional. 14 Patients also have an ethical obligation 15 to participate in quality improvement. Health care 16 delivery is a collective, cooperative enterprise. A 17 person seeking care from the health care system cannot 18 be free to refuse all cooperation in QI activities, 19 because QI is a critical ingredient in the creation of 20 the benefits that the patient actually is seeking. In 21 a world of constant change, adaptation is an integral 22 part of producing quality health care, and adaptation 194 1 is more likely to be beneficial when it's guided by 2 data as in the methodology of QI. Yet to be 3 effective, QI activities require at least a minimum 4 level of cooperation from participants in the care 5 delivery system, at least in allowing the collection 6 and use of information about their health conditions, 7 the treatments they receive and the results. 8 Okay. In practical terms, what this means 9 is that specific informed consent is not required 10 every time a human participant is included in a QI 11 activity. Since QI is part of normal health care 12 operations, consent to it is part of the consent to 13 receive clinical care. Consent to receive health care 14 should include consent to a minimal level of 15 cooperation with the ongoing QI activities of an 16 organization providing care and there should be public 17 education, so that people are already aware of this 18 when they seek care. 19 When they present themselves for care, 20 they should be reminded of this background expectation 21 and informed in general terms about the QI that goes 22 on and how they can obtain more information about 195 1 projects. When the QI activity imposes a greater than 2 minimal risk or burden, however, their specific 3 consent is required and when the results of QI 4 activities are published, they should be guaranteed 5 confidentiality or they should have to give their 6 consent. 7 So these are examples of the major project 8 conclusion, which is that the requirements for ethical 9 QI differ from the requirements for ethical clinical 10 research, and these differences justify a somewhat 11 different approach to oversight of QI. Now, we 12 arrived at this conclusion by considering carefully 13 the requirements that are accepted internationally for 14 the ethical conduct of clinical research and then 15 carefully analyzing them from the point of view of QI 16 using the analysis I just summarized very briefly 17 about the relationship between QI, clinical practice 18 and research. 19 The article that I just put up there lays 20 out seven requirements for ethical conduct of clinical 21 research, and the authors of the article say that five 22 of these are substantive, the first four and the last, 196 1 whereas two are important, but they are more in the 2 nature of procedural requirements. In other words, 3 namely, those two are independent review and informed 4 consent. What they mean by that is that they are 5 designed to help ensure that the other requirements 6 are met and even for research, they are a means to an 7 end and other means may sometimes be appropriate. 8 Now, as it turns out, the major 9 differences that we saw in the ethical requirements 10 for QI from those for research were exactly in these 11 two areas, and we're not saying -- we did not conclude 12 that these ethical requirements of independent review 13 and informed consent do not apply to QI. What we 14 concluded was that the requirements ought to be 15 somewhat different. For example, specific informed 16 consent, as I said, we thought is not always required, 17 but sometimes it is, depending on the nature of the 18 activity. 19 Similarly, independent review is required 20 in most cases in the sense that someone other than the 21 person who came up with the QI activity should review 22 it for ethical acceptability, and some projects ought 197 1 to require quite extensive review. But in many cases, 2 it would make sense to have the review take a rather 3 different form than the current system of IRB review. 4 And the problems of IRB review of QI are 5 not totally dissimilar from the problems of IRB review 6 with some kinds of research. For example, the fact 7 that it has high transactions costs and, of course, 8 many QI projects are done by people whose day job is 9 to deliver clinical care and the resource cost of 10 going through the IRB process might actually kill the 11 project, because it would be large relative to the 12 actual resources devoted to the project. 13 Second, the IRB process does not fit the 14 structure of QI activities very well. QI activities 15 are a part of normal health care operations. They 16 require incremental change as they go over, as they 17 are carried out over time, whereas IRBs are more used 18 to reviewing a discreet proposal that is submitted, 19 reviewed in its entirety, has a fixed protocol and so 20 on and then is implemented and comes back to the IRB 21 only if changes are required. So it doesn't fit the 22 structure of QI activities very well. Third, the 198 1 ethical standards for QI are different than for 2 traditional research, as I have already begun to argue 3 and in the report we eventually write, there will be 4 a lot more discussion of that. 5 But finally, the most serious problem, I 6 think, with requiring IRB review is that it creates a 7 disincentive for systematic monitoring and evaluation 8 of change. And the reason for that is that in most 9 cases, you can avoid IRB review of a QI project by 10 simply making the underlying change that the QI 11 project is monitoring and investigating without the 12 systematic investigation of its consequences. 13 QI is mostly about changes that are within 14 the discretion of health professionals and managers to 15 make, for example, adopting guidelines and figuring 16 out to introduce national guidelines into a managed 17 care plan, doing a minor change in managerial 18 arrangements that might possibly have a negative 19 effect on patients, but probably won't. You normally 20 can do that kind of managerial change without any kind 21 of review or anything until you introduce the 22 systematic investigation and collection of data, 199 1 whereupon it then becomes something that looks rather 2 like research. 3 Well, unfortunately, what you do then is 4 you create an incentive to simply not do the 5 systematic investigation of the consequences, and this 6 is particularly a bad thing when we are afraid that a 7 particular change that is being introduced in the 8 health care system might actually harm patients, 9 because those are exactly the changes that you might 10 not want to send through an IRB and you can avoid 11 doing that by simply removing the investigation part. 12 Okay. So stepping back then. Needed to 13 ensure ethical conduct of QI are three broad things. 14 One is guidance on ethical requirements, and here I'm 15 talking more broadly than just OHRP. We need 16 guidance, we need people out there in the quality 17 improvement world and the ethics world to think in 18 greater depth and detail exactly what the guidance 19 should be. 20 Well, I will take these one by one. By 21 guidance I mean how to interpret and apply ethical 22 requirements. For example, with respect to 200 1 methodology, one of the requirements for ethical 2 research is that the methodology be properly 3 structured to achieve the goals of the research. The 4 same goes for QI, but there is somewhat less clarity 5 around what kinds of methodologies are appropriate and 6 what kinds aren't. 7 Informed consent. The issue of when 8 informed consent is and is not required and finally, 9 ethical review, what kind of ethical review is 10 appropriate. And it's very important to emphasize 11 that the people responsible for reviewing QI to ensure 12 that it is ethical, no matter how decent they are, 13 they need some standards to apply in order for ethical 14 review of research or of QI to be socially beneficial. 15 We also need an environment, an economic 16 and regulatory environment that supports ethical 17 conduct of QI in the sense that the responsibilities 18 of all parties are clearly established and the 19 incentives are in the right direction. 20 And then finally, we need accountability, 21 and by that the important thing to remember when you 22 think about accountability is that managing the 201 1 quality of care is a core management function, and 2 that core management function includes managing the 3 quality of QI activities. 4 So my conclusion is that because this is 5 a core management activity, it is not ideal to send it 6 out, to outsource it, so to speak. The conclusion we 7 seem to be coming to is that rather than exporting QI 8 to the IRB based system for protection of human 9 participants in research, we should be wanting to 10 import the system for the protection of human 11 participants in QI into the accountability system for 12 clinical care. 13 Okay. Now, what I want to emphasize here 14 is this is not about letting QI off the hook. There 15 has to be an effective accountability system for 16 ethical conduct of QI, and this implies that managers 17 and health professionals have to be held to 18 appropriate standards. The ethical review and 19 supervision is the responsibility of the health 20 professionals and managers, but there also has to be 21 accountability from external sources for the conduct 22 of QI just as there is for clinical care. In fact, 202 1 what we need to do is we need to improve the clinical 2 care accountability system and, in fact, requiring the 3 people in systems delivering clinical care to do QI 4 and to do it ethically is an integral part of 5 improving the clinical care accountability system. 6 What we really need to have, finally, is 7 integration of accountability for the ethical conduct 8 of clinical care delivery, QI and research. There are 9 many settings in which all three of these activities, 10 clinical practice, QI and research, with a capital R 11 as I would put it, go on side by side. If we had a 12 properly integrated system, which left a lot of the 13 responsibility with the people in that clinical 14 setting to supervise and on an ongoing basis, what was 15 going on, we would then spend less time and energy on 16 struggling with the definition and have the ability to 17 pay more attention to the overall pattern of these 18 three closely related activities and their impact on 19 the well-being of patients within a clinical care 20 setting and the organization within which it is 21 located. 22 So I leave you then with a question and 203 1 this is the question that our group will be taking up 2 at the November meeting and for sometime thereafter, 3 I suspect. Can this be done? And if so, what would 4 be the best way to do it? And we also are here and I 5 want to thank SACHRP for allowing us to be here, me to 6 be here, on behalf of the project. We want to 7 cooperate as much as possible with OHRP in trying to 8 figure out a way to incorporate QI into the human 9 subjects protection system in such a way that people 10 are less confused than they are right now, and the 11 activities can go forward as they need to. 12 (Applause) 13 CHAIR PRENTICE: Thank you, Mary Ann. 14 That was very interesting. Our next presenter is 15 Peggy O'Kane and Peggy is President of the National 16 Committee for Quality Assurance known by the acronym 17 NCQA, which is an independent, nonprofit organization 18 whose mission is to improve the quality of health care 19 everywhere. Welcome to SACHRP, Peggy. We're looking 20 forward to your remarks. 21 MS. O'KANE: I really appreciate the 22 opportunity to be here today. I have learned a lot 204 1 from listening to Mr. Hodge, because I think there are 2 some similarities in considerations between QI and 3 public health. You may hear a lot of echoes between 4 Mary Ann's presentation and mine, because I sit on the 5 panel that the Hastings Center has put together to 6 deliberate about this issue, and we tend to be in 7 agreement on much of this. 8 What I'm going to do today is, first of 9 all, I'm going to introduce you to NCQA a little bit, 10 and then I'm going to set a bit of a context for why 11 quality improvement is so urgent in health care, and 12 then I'm going to make some comments that really are, 13 I think, very much aligned with what Mary Ann was 14 saying and then make some recommendations. 15 My thinking, at this point, is still 16 evolving. I do appreciate the complexity of the issue 17 and do see that there aren't bright lines here, and so 18 I think we have to have a bit of humility about the 19 process that we're going through, so I appreciate the 20 chance to have a dialogue with you about some of these 21 issues. 22 Just to say a few words about NCQA, we're 205 1 a 501(c)(3) nonprofit. We measure and report on 2 health care quality. We bring different stakeholders 3 around the table to achieve consensus that we found 4 that with complex issues, that's really the best way 5 to move forward. Our mission is to improve the 6 quality of health care and we do that in a very 7 specific way through measurement, transparency and 8 accountability and we are all about that agenda 9 throughout health care. 10 I should mention that we also have a 11 project, which we do jointly with the Joint 12 Commission, called the Partnership for Human Research 13 Protection, which accredits research programs on their 14 treatment of human subjects and we have created a 15 separate entity. It has a separate Board of 16 Directors. I am on the Board and it is simply 17 dedicated to improving the protection of human 18 research subjects. Its primary activities are 19 accrediting HRP programs and IRBs, and it also 20 provides publications and education services. 21 I am not here today to speak on behalf of 22 PHRP. I realize that, you know, the relationship is 206 1 complex, but I am really speaking today as the 2 President and CEO of NCQA, from that perspective. And 3 I would recommend this to you. There are handouts. 4 I'm not going to go into this, because I want to sort 5 of stick closely to the subject at hand. 6 I like this diagram of, basically, the 7 purpose of health care and I think it represents a 8 very different paradigm than was alluded to earlier in 9 the Belmont Report. The paradigm of health care, at 10 that point in history, I think was very much a one-on- 11 one transaction between a health care provider or a 12 physician and a patient, and I think what we have 13 learned over the years is that that paradigm often 14 results in gaps in quality. And so let me just walk 15 you through this. 16 If you think of a population being the 17 population of the United States, the Medicare 18 population or a panel of patients being taken care of 19 by a particular physician, at any particular time the 20 majority of the population is healthy or at low risk. 21 For various reasons, they may move into an at risk 22 category, into a higher risk category. They may 207 1 develop early symptoms and they may have active or, 2 indeed, very complex and often fatal disease at this 3 end of the spectrum. 4 In our point of view, the purpose of 5 health care is to exert pressure from the right side 6 of the picture to the left side of the picture, and so 7 thinking about how to deliver quality does involve 8 thinking about populations and not even thinking about 9 them in terms of tradeoffs with individual benefit. 10 So if you think about the diabetics that are in a 11 particular health plan, you need to know who has 12 diabetes in order to know if people are receiving the 13 standard of care. So there is a more complex, I 14 think, paradigm of quality that we have today that 15 really needs to be part of how we think about it, and 16 I just wanted to make sure we got that point across. 17 Now, we know and the IOM, I think, has put 18 out some very compelling reports that we have today in 19 American health care serious quality gaps. Beth 20 McGlynn's article in the New England Journal last year 21 documented that Americans receive recommended care 22 about half the time. There are preventable errors 208 1 that result in 44,000 to 98,000 deaths and cost 2 between $17 and $29 billion each year. That's from 3 the IOM Report on Medical Errors. And providing 4 optimal care could save lives and increase 5 productivity and lower costs. 6 We have just released, every year we 7 release, HEDIS results on health plans that publicly 8 report and we have been working on a concept called 9 the quality gap, which is the gap between the plans, 10 the top 10 percent of health plans that are reporting 11 to us and what we have in terms of the average 12 baseline in the United States, and that gap we see as 13 a surmountable gap because, obviously, 10 percent of 14 plans have surmounted it, and it cuts across all the 15 things that we measure. 16 And let me just point out that what NCQA 17 is measuring usually is not areas of great debate in 18 health care. They are areas of strong scientific 19 consensus where we know that the interventions make a 20 big difference. You know, because of all the 21 difficulties of quality measurement, we have been very 22 strategic and parsimonious about what we're measuring. 209 1 I'm not going to go through these numbers, 2 but we have made estimates using published information 3 on the avoidable deaths. If we were to close that 4 quality gap, remember the gap I just mentioned, our 5 estimates, depending on how conservative you are, go 6 from 42,000 to 79,400 deaths per year in the United 7 States that could be avoided. 8 The costs, $1.8 billion. I think this is 9 very conservative. This is only related to medical 10 costs, not other externalities. And in terms of 11 productivity and absenteeism, 66.5 million days, which 12 translates into how many FTEs, Sarah, do you remember? 13 But it's 300,000 FTEs and it's the total employed 14 population of Ford and General Motors. So the toll of 15 poor quality in terms of suffering and waste is huge, 16 and so I think I put this out there as kind of a way 17 of thinking about, you know, what is the cost of not 18 improving quality, and to really make the case that 19 there is a strong ethical imperative to improve the 20 quality of health care. 21 Failures in health care come from a 22 variety of sources. Failure to apply clearly 210 1 beneficial medical treatment, so I'm giving you some 2 examples that we measure, and not giving beta blockers 3 after a heart attack. Failure to apply a beneficial 4 treatment effectively. So maybe you start someone on 5 antidepressants, but you fail to follow-up and make 6 sure that they are taking their antidepressants. 7 There is a whole set of measures around that related 8 to a practice guideline. Applying a treatment that is 9 not clearly beneficial. 10 So we have a measurement, inappropriate 11 use of antibiotics, but there are other documented 12 examples, people getting surgery that don't need it 13 and so forth. Applying unproven treatments. You 14 know, we had the whole episode of bone marrow 15 transplants for women with advanced breast cancer. I 16 won't go into it, but even screening mammograms for 17 women aged 70 and over, there is not good evidence 18 based on that today, so doing that constitutes waste 19 and possibly harm to patients. 20 Failing to address patient needs. We very 21 strongly believe that patients need to be fully 22 informed of risks and benefits of whatever 211 1 intervention is being offered to them and that failing 2 to take their preferences into account is a failing of 3 health care. 4 The other point is systems can improve 5 quality, so we need systems that design important 6 processes to promote adherence to standards of care. 7 You have heard about computerized order entry systems 8 perhaps, registry use for patients with chronic 9 disease. So how do you make sure that all your 10 patients with active heart disease are getting the 11 appropriate secondary preventive care? 12 Monitoring to detect variation from 13 standards and, you know, we have certain areas of 14 health care where we have very clearly defined 15 standards. We know as McGlynn's article and many 16 other publications in the quality world show that we 17 fail to do that at least half of the time. Let me 18 just report on one other thing. There is also the 19 example of where you don't have a strict practice 20 guideline and you have great variation in what's going 21 on in the process of care. 22 I want to give you this example, because 212 1 I think it's a compelling one. Brent James at 2 Intermountain Health Care looked at the administration 3 of antibiotics related to surgery, and he found that 4 there was a huge variation in the time of delivery of 5 the antibiotics, what the antibiotics were, and so 6 forth. And by systematically examining these 7 different processes and the outcomes of care, he was 8 able to ascertain that the two hour window prior to 9 the surgery was the point most when these antibiotics 10 should ultimately be administered. That has become a 11 standard of care. 12 Now, it was the data mining of information 13 that Intermountain Health Care was routinely 14 collecting that allowed this to happen, so I put that 15 there. Let's sort of bookmark that for more 16 discussion later. And I guess we could call that -- 17 I think that there is an ethical responsibility for 18 surveillance. I want to make that point. 19 So if we know that health care, that 20 quality doesn't just happen, then I think there is an 21 affirmative responsibility for surveillance of patient 22 outcomes, including their symptoms, their functioning 213 1 and satisfaction. And that is the kind of information 2 that can lead to the advancement of medical knowledge. 3 And reporting to health care professionals and the 4 public is another very important way. I mean, 5 physicians receiving information on people in their 6 panel who need a particular intervention is one of the 7 ways that health plans, for example, have improved 8 quality. 9 What is quality management? I liked Mary 10 Ann's presentation on this. What it is it's very 11 systematic approaches to monitoring and improving on 12 the performance of the health care system and health 13 care professionals. It's most successful when it's 14 linked to operations and it's very important to get 15 the engagement of the people that need to be 16 implementing the improved quality systems that are 17 developed. 18 It needs to be supported by leadership and 19 it needs to involve health care staff at all levels. 20 As we know, health care is a team sport. That 21 paradigm of the old dyadic relationship, I think, is 22 really not the one that leads to optimal quality. We 214 1 see that repeatedly in the best practices out there. 2 Examples of QI and they are going to go 3 back to the chart I showed you were the failures. 4 Implementing proven standards or well-accepted 5 standards, exploring variations, you know, James being 6 an example of that, and they are primarily directed 7 towards improving health care operations as a means of 8 improving patient outcomes. 9 And for those of you, you know, that live 10 and breath in the research world, for me, I think that 11 the most compelling thing is we have spent a lot of 12 money, a lot of effort, a lot of patients have 13 participated in trials to develop medical knowledge. 14 And the deployment of medical knowledge absolutely 15 depends on having excellent systems to deliver. So, 16 you know, I think if I could speak to you in those 17 terms. 18 Good QI shares some important 19 characteristics with research. So I would say it's 20 designed to create generalizable knowledge. You know, 21 I think that there has been all this kind of worry 22 that if I'm talking about generalizable knowledge, 215 1 it's research and it needs to go through the IRB. I 2 think Mary Ann was very good in pointing out that the 3 unintended consequence of that may be well, if I just 4 do this kind of amateurish little thing or this thing, 5 that I'm just going to go off and do by myself without 6 talking to anybody about it, that that somehow doesn't 7 qualify and doesn't need to go through IRB review. 8 The consequence being that QI activities 9 can be less effective than they would be if they were 10 more systematic. So just as in research, there is an 11 identification of an eligible group, an observation or 12 an intervention with individual patients, staff or 13 systems, systematic data collection and adequate 14 statistical analysis. The ethical underpinnings, I 15 think, are very similar. Beneficence, clearly, is a 16 goal of health care and of research, not doing harm. 17 Minimizing risks and equity, I think, are all 18 important common attributes. 19 But QI differs from research in important 20 ways. It occurs in the context of health care 21 provision. Now, Mary Ann made this point about local 22 improvement of quality, but I would want to just put 216 1 out there, again for our discussion, what if you had 2 kind of a consortium of individual organizations 3 improving quality? There, you might actually be able 4 to move forward in a more rapid way. That I think the 5 underlying ethical problems with quality would argue, 6 would be terrific. So I don't want to tie this too 7 much to the local situation, so I put that there for 8 discussion. 9 Patients have consented to treatment, you 10 know, pretty much across the board. The health care 11 provider has access to patient private information as 12 a part of providing care. They need that information. 13 And the health care provider has responsibility to 14 provide appropriate treatment. And I think no one 15 argues about that. 16 So we also agree that QI needs review and 17 oversight, but we need to encourage best practices in 18 QI, both in effective design and measurement. We 19 don't want people fooling themselves that 20 interventions worked when they didn't, because they 21 had, you know, analytical problems. Leadership and 22 involvement of staff, so we need to consider existing 217 1 standards of care and health care operations. And we 2 need to protect patient rights without unnecessarily 3 constraining the use of data. It's a hollow victory 4 if your privacy is protected, but you lose your life 5 or if somehow your health is compromised. I don't 6 want to be too melodramatic, but I feel pretty 7 passionate about this. 8 So I don't believe that research IRB is 9 optimally suited to evaluating QI. I think it assumes 10 that the investigator is external to the care process. 11 There are problems with responsiveness and operational 12 time pressures. IRB membership may not be familiar 13 with operational issues. Remember, I do believe that 14 the people that are -- the deployers of the new 15 systems really ought to be involved with the QI 16 process. 17 I think I'm going to skip this, because 18 I'm worried about time, but I just want to make the 19 strong point that because something is reducing costs, 20 I don't think should make our antenna go up and say 21 that this is something that is necessarily to be 22 subjected to a higher level of scrutiny. We know that 218 1 we have a lot of inefficiency in the current system. 2 There is a lot of waste. We have also 45 million 3 uninsured Americans. We certainly are spending plenty 4 of money compared to the rest of the world, much more 5 per capita than other countries. 6 So I would say that this is something that 7 needs to be on the agenda, but I'm afraid it's not 8 essential to the conversation. I'll be happy to talk 9 about it in the discussion. So I agree with Mary Ann 10 that there needs to be, I like the way she framed it, 11 oversight of QI activity. You do not want lone 12 rangers off doing quality improvement projects in 13 whatever setting you are in. But I think what we need 14 is the appropriate infrastructure or development of 15 the appropriate capability within the quality 16 infrastructure to make sure that patients' interests 17 are protected. 18 We need to be aware of the failures in 19 health care. We need to promote systems to address 20 these failures. We need to develop structures that 21 ensure that QI efforts are using the state of the art 22 in their analytics and we need to eliminate review 219 1 processes that hinder QI implementation. So that's my 2 pitch. Thank you very much. 3 (Applause) 4 CHAIR PRENTICE: Thank you, Peggy, for an 5 interesting presentation. Would the Panelists all 6 assemble up at the table? Now, while they are 7 assembling, I want to change the schedule around 8 slightly. It's about 1:45 right now. We're scheduled 9 for a break at 2:30. I want to reserve the -- last 10 break at 2:30. Yes, I want to reserve the last 10 11 minutes of this time period for an opportunity for the 12 public to ask any questions that they wish. So I'm 13 putting that into the schedule, so SACHRP Members, 14 please, bear that in mind when you ask your questions. 15 And then, once again, the only benefit I 16 have of being Chair of this August group is I learn a 17 lot, but I also can ask the first question. I think 18 it's a very, very important issue. I can't tell you 19 how many times I get calls from people at my 20 institution asking me am I doing research, and I was 21 hoping that Mike Carome would tell us what was meant 22 by all of those terms and the definition of research, 220 1 but he declined to do that. So that didn't help me 2 very much. 3 So I still don't really know the answers 4 to a lot of questions, although I've certainly got an 5 awful lot of information from the Panelists. So I 6 want to pose one scenario for you, which would be 7 typical of something I might get asked. I get a call 8 from let's say it's the director of the NICU and they 9 say I want to do a quality outcome assessment, a 10 quality initiative improvement, and I'm going to do it 11 systematically. I'm going to actually perform a 12 systematic study. 13 I got a protocol to do that. And I 14 actually want to randomly assign patients to two 15 standard methods, interventions, procedures, whatever 16 you want to call them, and I'm going to collect data. 17 I'm going to analyze that data and I'm going to draw 18 some conclusions and I'm going to implement some 19 changes as necessary in the delivery of health care. 20 I'm going to make this better at this institution. Am 21 I doing research? 22 The next question I ask is okay, other 221 1 than trying to improve the delivery of health care in 2 the NICU, do you intend on publishing this data? Are 3 you going to present this data at a national meeting? 4 And if the individual says yes, that's about the only 5 thing that I can hang my hat on, at least up to now. 6 I couldn't hang my hat on anything else. I say well, 7 you know, you're going to have to go through IRB 8 review. If he or she says no, then I say well, you're 9 not doing research or if they say maybe, then I often 10 say well, okay, tell me when you make up your mind. 11 And then it may be a situation where 12 you're doing prospective research or maybe a situation 13 where it's retrospective and you're simply analyzing 14 existing data could be exempt if you strip off the 15 identifiers. Okay. So that's my problem. That's 16 been my response to investigators for a number of 17 years. So I would like to get the Panel's reaction to 18 that, because listening to you, I'm still not quite 19 sure whether I'm doing it right or not. Sure, go 20 ahead. 21 MS. O'KANE: Well, one thing I would like 22 to say is that, you know the thought of having 222 1 patients exposed to two separate processes and 2 randomly assigned may sound kind of horrifying. But 3 I think what we have to think about is what happens 4 when we don't intervene, which is that we have a 5 proliferation of processes that may be as varied as 6 the number of providers or combinations of 7 permutations of them working with each other. You 8 know, so there is a kind of chaotic quality to care 9 processes that have not been defined. 10 Now, I have a strong point of view that 11 publication should not be the threshold issue. I know 12 it has been used as such, but if we know that we are 13 in a situation now where there is an urgency, we know 14 enough about quality that there is an urgency to 15 improve it and that knowledge needs to be shared as 16 much as possible, as long as the patient's privacy is 17 protected, it seems to me -- I don't understand why 18 publication in itself should constitute -- 19 CHAIR PRENTICE: So you would say no. 20 MS. O'KANE: Right. 21 CHAIR PRENTICE: Right. Dr. Carome? 22 DR. CAROME: Based upon what you 223 1 described, you used the words systematic study, 2 perhaps systematic investigation could have been 3 substituted. It sounds like what the investigative 4 person was doing was a single site study comparing 5 Intervention A to Intervention B in a systematic 6 randomized manner. It's collecting data to draw 7 conclusions and that sure sounds like research. I 8 don't think it is indistinguishable from research, 9 because it sure sounds like research, because it is 10 research. 11 CHAIR PRENTICE: Okay. 12 MS. O'KANE: Okay. 13 CHAIR PRENTICE: Now, we've got two 14 different opinions here. All right. Mary Baily? 15 MS. BAILY: Okay. Well, I think Mike 16 would also agree that it is quality improvement. It 17 is definitely QI, because you are, in fact, looking at 18 two existing treatments that are within the standard 19 of care and that you could -- that are already in 20 place. I take it these are -- the way you described 21 it, they are already being -- they are doing them 22 already that way. 224 1 CHAIR PRENTICE: At least they are 2 standard. 3 MS. BAILY: Right. 4 CHAIR PRENTICE: They are standard. 5 MS. BAILY: They are standard. Our 6 project has not taken up the issue of randomization 7 yet and we do intend to, but the fact is if these both 8 -- if you had two different doctors who might -- and 9 each one picks one of these standards and uses it, 10 that is what is going on in the system and if you then 11 decide to find out which one works better in order to 12 improve care so all the babies do as well as possible, 13 that would certainly be a QI project. I think Mike 14 would probably say that it is quality improvement 15 research and it falls into one of his overlapping 16 categories and should go to an IRB. 17 Now, I might be prepared to say that it 18 was in that overlapping category. What I'm not so 19 clear about is whether we want -- first of all, I 20 think the publication thing, we should not focus on, 21 because whatever you call it, it surely would be 22 interesting to other NICU docs and you would want them 225 1 to think about that ahead of time and be willing to 2 share the information. Whether it was through a 3 publication or some other way, provided the patient's 4 personal information was protected. 5 You might also want to have some form of 6 review and some form of supervision within the 7 setting, because, of course, you would be doing things 8 within the NICU that would use some resources and 9 would need to be, possibly, explained to patients, 10 possibly not. But where I would probably hesitate is 11 is the IRB really the best place for this? And I 12 would like to see a somewhat different structure 13 within your organization where you could seamlessly 14 review both QI and research and stuff that is sort of 15 both and not have to spend a lot of time on the phone, 16 which you presumably do like everybody else. 17 The reason this project exists in the 18 first place is because people are spending a lot of 19 time on the phone in conversations like that. And if 20 you had a seamless review process, you might be able 21 to avoid some of that time on the phone, while at the 22 same time making sure that these activities go forward 226 1 properly. 2 CHAIR PRENTICE: Okay. So you agree with 3 Peggy. You are basically saying that there should be 4 some sort of review mechanism in place, but it should 5 not be the IRB, simply because we are publishing it. 6 James, do you want to throw in your two cents worth to 7 just kind of make this even? 8 MR. HODGE: Not particularly. No, let me 9 throw in two cents and I do want to agree with Mike 10 that this sure seems like, sounds like, looks like, 11 feels like research and there is probably a good need 12 for an IRB to be involved. However, I completely 13 agree with my colleagues to the left, publication is 14 inconsequential in the view of public health 15 practitioners, because of the sheer reality that what 16 is disclosed and done through public health practice 17 is routinely published using non-identifiable data, of 18 course, but is done just as a regular routine practice 19 the same as research. So in my mind there is no 20 linchpin to whether you have research or quality 21 improvement. 22 CHAIR PRENTICE: Okay. 227 1 DR. CAROME: Mr. Chairman, if I could just 2 add one. The type of activity described, the IRB 3 system is well qualified to handle. That's the type 4 of thing they see all the time and the literature, at 5 least the research literature, is filled with studies 6 that have compared two or more standard accepted 7 therapies to try to find which actually is better and 8 for which subjects or which types of patients is it 9 better. 10 The fact that neither is experimental or 11 innovative doesn't mean it's not research. NIH 12 sponsored the multi-center ALLHAT trial in which they 13 compared standard blood pressure medicines for people 14 with hypertension to decide which ones are more 15 effective, cost effective and in which sequence should 16 you use them. Again, all accepted therapies used, you 17 know, they way they were used in the research was the 18 way they were used in clinical care, but it was still 19 research and the IRB system handled that one quite 20 well. 21 CHAIR PRENTICE: Okay. I'll open it up 22 for SACHRP Members now. You had your hand up first, 228 1 Mark. 2 MR. BARNES: I always have my hand up 3 first, Ernie. I mean, what Mike says sounds good 4 until one realizes that there is kind of no end to 5 that process. Then the question becomes a slippery 6 slope. If those are your criteria, then what quality 7 assurance or quality improvement activity is not human 8 subjects research? Then, you know, and obviously, one 9 of the things that we want to avoid is the unintended 10 secondary and tertiary consequence of having every 11 single thing classified, every single activity 12 classified as human subjects research. 13 You know, just a final point, and that is 14 that one could take another example in public health 15 practice that would be similar to the example that was 16 used in the context of a NICU. I mean, one could 17 take, you know, the -- and I'm not sure by the way 18 that I -- we can talk later, but I'm not sure that I 19 buy all of the distinctions that you make, because I 20 think that there are many times -- just to give you an 21 example, there are many times that in public health 22 practice we do things like, that are basically 229 1 experiments, passing a regulation in the City of New 2 York to lock up and civilly detain people who are 3 recalcitrant in taking their anti-tuberculosis 4 medications never been done before, there's not a body 5 of knowledge that says we can do that. 6 We open a unit. We sequester them and we 7 report the results of that. I mean, is that research 8 or is it public health practice? I mean, you know, I 9 think everybody has thought that it is public health 10 practice, but, you know, my point is this, and that is 11 that these lines are very permeable. And if one just 12 accepts, you know, Mike's answer and doesn't probe 13 further with other hypotheticals, then one, I think, 14 inevitably ends up in a situation which everything 15 becomes research. And that itself then has unintended 16 consequences. 17 CHAIR PRENTICE: Okay. Celia? 18 DR. FISHER: Thank you for your very 19 illuminating talks. I agree with Mike, that sounded 20 to me like research, but I will give my justifications 21 for why. First of all, I just want to briefly say 22 that I agree that simply placing publications as the 230 1 criteria for whether something is research does not 2 make sense. My guess is, and we have people here that 3 can tell us if this was so, but that the frame is of 4 these guys, and to really looking at the motivations 5 of external researchers and thinking that by having 6 publication as some type of criteria it would rule out 7 the external kinds of research that you are 8 eliminating anyway from quality improvement. 9 I also agree there has to be some kind of 10 oversight, but what I want to do is I'm trying to 11 think about it and I think what has not been 12 presented, at least in a direct manner, is in some 13 sense an ethical framework for how we would decide 14 what is QI and what is not. And so I think, Mary Ann, 15 you attempted to begin to do that. You talked about 16 how an underlying justification for a lack of informed 17 consent is that both the system and the patient have 18 a mutual obligation to and an investment in system 19 operations. 20 Now, that's very interesting, except then 21 you say it is only true for minimal risk research. 22 And so for me that becomes problematic. Let me list 231 1 a couple of points and then you can answer. So I 2 don't see that therefore as some kind of a uniform 3 framework. In addition, it also assumes, and although 4 I certainly agree with Margaret that a cost 5 effectiveness study is not in and of itself unethical 6 or not quality improvement, but the model that I think 7 you were presenting, Mary Ann, has an assumption that 8 the system has the patient's best interest in heart, 9 at heart throughout any of the types of investigations 10 that it would do. I think there are aspects of cost 11 benefit analysis that may not be in line with how 12 patients might perceive best practices. 13 The other thing is that, for me as I have 14 struggled with this over the years, I have tried to 15 figure out how -- I tried to frame whether it is 16 research or not and whether or not the service 17 provided or not provided, because of a QI review, 18 would affect the patient differently than if a QI 19 review was going on in a way that might be 20 significant. 21 Now, in the types of examples that 22 Margaret was giving, failures in the health care 232 1 system as judged by best practices, to me that seems 2 like QI in the sense that it is really the physician 3 and the system that is the target, is the focus of the 4 research. The patient data is not changing from what 5 it was. It's almost a record review, unless I 6 misunderstood what you were describing. 7 On the other hand, in randomization or 8 perhaps some other types of methods, I would ask two 9 questions that would -- or two scenarios that would 10 mean to me that it is probably close to research. 11 One, will the patient be exposed to a procedure they 12 would not otherwise receive if QI was not going on or 13 will they not be receiving a treatment they would 14 otherwise receive, because a QI is going on? 15 The second question I would ask would be 16 in terms of informed consent is a clinical choice 17 removed from the practitioner or the patient that 18 would otherwise be there if QI was not being 19 implemented? So I'm just throwing out the way I have 20 begun to frame distinction. 21 MS. BAILY: I think you are raising a lot 22 of provocative questions that we would like to take a 233 1 whack at. 2 DR. FISHER: Yes. 3 MS. BAILY: I would like to reply to that. 4 First of all, we do have a much more elaborate ethical 5 analysis of these issues and I had to reduce it all to 6 20 minutes. And those questions that you raised are-- 7 I'm not quite sure, I still only have a few minutes, 8 so I can't really go into great detail. But I think, 9 for example, the issue of whether or not informed 10 consent is required does hinge very much on what you 11 said, which is to what extent is the person being 12 subjected to some kind of an intervention, which is 13 not minimal? 14 For example, just personal health 15 information, perhaps there is some kind of a procedure 16 that is not part of their normal health care and as a 17 result of being part of a QI project. You do want to 18 ask for their -- you want volunteers being willing to 19 do that. At the same time, unlike research where they 20 have an absolute -- they have no ethical obligation to 21 say yes, I think they would still have, even when you 22 do have to ask for their consent, I think they do have 234 1 an ethical obligation to consider participating that 2 they would not have in a research setting. 3 So there isn't necessarily a contradiction 4 there. And then there are many other things you 5 raised, which I don't even remember, but would love to 6 discuss them at greater length. 7 MS. O'KANE: If I might? First of all, in 8 randomization, I think the question is if you're 9 looking at a current state of affairs where there is 10 great variation in what's going on, some of it 11 appropriate, some of it inappropriate, and then you 12 are creating a group where what you are proposing to 13 do is ensure appropriate delivery of care, then what 14 will happen is, presumably, the intervention group, 15 you will study whether the intervention brings them to 16 a higher level of care than the control group, right? 17 So the control group is not any worse off 18 than they would have been. You know, I am not an 19 ethicist, which I think has become abundantly clear, 20 I just try to think straight about this stuff. They 21 will not be worse off. The intervention group you 22 will either demonstrate that the intervention brought 235 1 more appropriate care and then, presumably, would go 2 back and, you know, implement this broadly to the 3 whole group or you would say it didn't work and we'll 4 have to try something else. You know, is that 5 research or is it structured quality improvement? 6 DR. FISHER: Let me just say that, you 7 know, some of these arguments, I think, are very 8 appropriate, but when you start talking about, you're 9 really describing a treatment as usual control design, 10 so I'm not sure where it is research and where it is 11 not. So some of these examples may get very 12 difficult. In addition, just to throw in, there are 13 some other comments that are very much tied to 14 research as well. The need for generalizability. The 15 fact that informed consent of voluntariness may impugn 16 the ability to draw conclusions. Those in and of 17 itself do not argue that something is or is not QI. 18 MS. O'KANE: Right. No, I mean -- 19 DR. FISHER: I mean, Mike might have -- 20 MS. O'KANE: I think what we're stuck 21 with, I think, is that we have paradigms for thinking 22 about what is research that really overlap a great 236 1 deal with systematic delivery of care. And that, you 2 know, trying to draw the distinction too bright, too 3 sharply around research interferes with the systematic 4 delivery of care. I think that's the conundrum. 5 DR. CAROME: If I could make a comment? 6 You know, you already posed sort of one example and I 7 don't want my response to be interpreted to my saying 8 that what you described represents all QI activities 9 and therefore I think all QI is research. I think, as 10 Mark said, the discussion can't be framed by one 11 example. I think these activities span a wide range 12 of activities and we could probably lay out 100 13 distinct examples and some of them we would probably 14 say as an office are not research and some may be. 15 So let me give you an example of an 16 activity that I think is a QI activity, and I could be 17 wrong perhaps, that I think is not research. So it's 18 known that after someone has an acute heart attack, an 19 acute MI, that mortality is improved if you send those 20 people home on beta blockers, which is a type of blood 21 pressure medication. And so good delivery of health 22 care would be that everyone who had acute MI goes home 237 1 on a beta blocker unless they have a clear 2 contraindication of a reason not to be on a beta 3 blocker. 4 And so a hospital might say, you know, we 5 want to deliver the best care possible, and so we want 6 to make sure that we send people home on beta 7 blockers, unless it is contraindicated after they have 8 had a heart attack. So the first thing they want to 9 do is: (A) Assess how often do we do that? So are we 10 doing that correctly? And so they perhaps review 11 their charts of the last thousand MI patients over the 12 last year and find out how many went home on a beta 13 blocker and how many didn't. And suppose they find it 14 was 50 percent, so now they have their baseline. 15 And their sole goal of doing this and the 16 subsequent activity I'm going to describe is to 17 improve their delivery of care to their patients, not 18 to contribute to generalizable knowledge. They want 19 to make sure everyone goes home on something that is 20 proven and known. Go home on a beta blocker. And so 21 now having done that finding only 50 percent of the 22 people who had a heart attack go home on a beta 238 1 blocker, they say we got -- that's not good. We've 2 got to put in place an intervention, a quality 3 improvement intervention, perhaps is the term some 4 would use, to make sure we get that as close to 100 5 percent as possible, because that's good care. 6 And so perhaps they set up a system in 7 which a time of discharge by the physician and maybe 8 that's done through a computerized discharge order, 9 there is a computer flag that says your patient had 10 MI, make sure you put them on a beta blocker or not. 11 And also, when the nurse does their sign out of the 12 patient, the nurse is triggered to make sure that they 13 confirm that if someone is on a beta blocker going 14 home and if not, that there's a reason why. There is 15 some kind of indication documented. 16 And so that process is implemented for the 17 physician and the doctor and after a year they 18 reassess well, how many people went home on a beta 19 blocker? Now, it's 99 percent. So they implemented 20 something to approve care and their goal was to 21 improve care not to contribute to generalizable 22 knowledge. And so what I just described, I think, is 239 1 an example of QI that's not research. 2 MS. BAILY: And then if they write it up? 3 No, but, I mean, they could write it up and they could 4 say -- 5 DR. CAROME: I think they could write it 6 up just like researchers write up a retrospective 7 series of, you know, their analysis of their last 100 8 biopsies, you know, related to some disorder and in 9 that retrospective chart review, the activities that 10 occurred and are documented in the clinical record 11 that are being reviewed weren't research activities. 12 Likewise, I think these QI activities, someone went 13 back and looked at them, that doesn't make those 14 activities research. 15 The research might be if they are 16 analyzing now identifiable prior information to test 17 hypothesis to contribute to generalizable knowledge, 18 then that record review would become research. But it 19 might be exempt and you actually might be able to do 20 it in a way it doesn't involve human subjects. 21 MS. BAILY: But let me push you a little 22 bit, because suppose what they wanted to write up was 240 1 how they managed to make this happen, and so there was 2 an intervention. They did something. They want to 3 write up how they managed to make it happen. And, 4 obviously, they would make it unidentifiable when they 5 published it, but it would have involved using 6 records. Would that make it research? 7 No, you're still comfortable calling that 8 -- I mean, I am, too, but I'm just pushing you. No, 9 I want to find out, because one of the things you 10 might be interested in is that something very like 11 that was considered. It had to do with the project to 12 adopt national guidelines to local standards. There 13 was an article published about an effort that wasn't 14 unlike that. It was trying to figure out to take a 15 national guideline, adopt it to local conditions and 16 then figure out how to implement it. 17 Did this have to go to an IRB? And 18 overwhelmingly the survey, I wish I could -- I'm not 19 good at rattling off all the numbers, but most of the 20 IRB and internal editors they tested this on thought 21 it was research and should go to an IRB. 22 DR. CAROME: Well, let me -- 241 1 MS. BAILY: So that project that you 2 described is very like that. But you say no, that's 3 pretty clearly QI. 4 DR. CAROME: Yes. 5 MS. BAILY: I thought so, too. 6 DR. CAROME: Well, let me take my example 7 and turn it into research. 8 MS. BAILY: Okay. 9 DR. CAROME: Suppose a hospital system 10 wants to find out what is the best way to make sure 11 people go home on beta blockers. And so now I've got 12 Intervention A. We're going to target the physician 13 to make sure the physician writes the prescription and 14 order or we're going to target the nurse and make sure 15 the nurse is -- you know, where is the best place to 16 flag this process and catch it. And so now we're 17 going to at perhaps two of our hospitals which are 18 similar, two others we're going to make the physician 19 being the person we're going to flag versus the nurse, 20 and now we're going to collect who had a better 21 outcome for being place on beta blocker. I think I 22 have turned that into research. 242 1 MS. O'KANE: But see now that's the 2 trouble. 3 MS. BAILY: That's the essence of QI. 4 MS. O'KANE: Because something which if 5 you only did one of those, you can do either one of 6 those without keeping track. 7 DR. CAROME: But if your goal is to assess 8 -- let me ask the question. Which of these works 9 better so that I can learn from that and disseminate 10 it to the community at large so that everyone can pick 11 the best, most cost effective way of doing it, I would 12 turn that QI activity also into a research activity. 13 MS. BAILY: But you have taken normal 14 management of health care organizations -- 15 DR. CAROME: But again, I have -- 16 MS. BAILY: -- and turned it into 17 research, which is very tricky. 18 DR. CAROME: As I said before, research 19 can compare standard normal things for the purpose of 20 learning and contributing to generalizable knowledge. 21 MS. O'KANE: Okay. Let me ask you a 22 question. If you have a monitoring and surveillance 243 1 system and you discover that patients of a particular 2 category are getting care that varies all over the 3 place, is it ethically appropriate to say but I'm not 4 going to do anything about that? 5 DR. CAROME: Of course not, but I don't 6 think -- that's not the question of whether a 7 particular activity is research or not. Of course, 8 you ethically should be doing something about that if 9 you discover poor medical care and that might be 10 through QI intervention. It might be, you know, a 11 variety of other techniques. 12 CHAIR PRENTICE: Okay. We've got a couple 13 more questions. Bob, you're up next. 14 DR. HAUSER: Ernie, I think the case that 15 you described is clearly research. We took a group of 16 patients to the intensive care unit. We divide them 17 into A and B. We intervened either A or B, the 18 hypothesis was that one or the other may improve 19 quality, that's clearly a research project and 20 involves experimentation. I think Mike just cited two 21 really good examples how you differentiate quality 22 improvement from research. 244 1 And I can't think of a better example. I 2 am on the Quality Committee of our health system and 3 we have 2,000 beds. We have a very large primary care 4 clinic, as well as some specialty areas. If we had to 5 subject our QI process to an IRB review, it would 6 cripple our efforts. 99 percent of what we deal with 7 are root cause analyses or sentinel event analyses and 8 these have to be dealt with right away. We're talking 9 about wrong-side surgery. We're talking about wrong 10 patient surgery. We're talking about adverse drug 11 reactions, hypoglycemia, hemorrhage because of heparin 12 overdose, etcetera, etcetera, etcetera. 13 And you have to get at these things right 14 away. You're not testing hypothesis. You're finding 15 where the problem is. That's what I call quality 16 improvement. If, on the other hand, you need, as part 17 of your quality improvement process, to conduct an 18 experiment, then it goes to the IRB. Just as Mike's 19 example illustrated. 20 CHAIR PRENTICE: Susan? 21 MS. KORNETSKY: One of the points that I 22 want to bring up that, Mary Ann, you mentioned is I 245 1 think a lot of the problem also, and I'm not blaming 2 journals, but a lot of what I have noticed is concern. 3 People who are doing quality improvement and then want 4 to write this up are being asked to sign all different 5 types of things away, you know, from the journals. 6 And, you know, we're trying. We've taken an approach 7 that we, at our own institution, don't -- I've sent 8 copies of our policies, our procedures. 9 So far it has not been challenged, but I 10 think this is a big problem and driving investigators 11 crazy. And also, having IRBs say well, you know, 12 journals question whether IRB is required, you know, 13 submit it to us. So I'm not sure what to do about 14 that, but wanted to bring that up. 15 MS. BAILY: I would like to comment that 16 that is one of the practical things we are trying to 17 accomplish in this project is to straighten out how 18 journals ought to look at QI and we do have some 19 people on the project who are well-placed to help us 20 with that. 21 Could I also mention that one of the 22 things that bothers me is that if you are setting up 246 1 this dramatic difference between the kind of review 2 that something requires, if you do it as QI and if you 3 do it as research, I think, I would be very tempted to 4 say well, I'm not even going to try to do that. I'm 5 not going to try to find out which works better for 6 the beta blockers. I'm just going to implement one of 7 them, because that way I don't have to go through an 8 IRB. Because you said that clearly was QI. 9 And also, what's to stop me from 10 implementing it and then deciding, you know, six 11 months later I didn't think it worked as well as I 12 hoped it would. Now, I'm going to implement the other 13 one and then I can see how that would happen. Well, 14 now, that wouldn't be research, at that point, because 15 it would be QI. And then at the end of the year, I 16 would have the results as if I had implemented them 17 both side by side or would I then have to say oops, I 18 was sly here and I have now committed a serious 19 ethical error, because what I did was equivalent to 20 research retroactively? 21 And I think that that kind of thinking, I 22 think we are wasting the time and the resources of the 247 1 people who do this very useful work by having them 2 have to do that kind of thinking. It would be much 3 better to have an integrated system that did not 4 privilege or penalize. In other words, if IRB review 5 and normal ethical oversight of QI were more closely 6 integrated so that you could get the right kind of 7 review in accord with the nature of the project, and 8 that is something that kind of bothers me about your 9 example. 10 CHAIR PRENTICE: Okay. We have time for 11 two more questions. Three and a half minutes. Keep 12 your questions brief and, please, keep your responses 13 equally brief. Mary Polan? 14 DR. POLAN: Well, I think what you said 15 just now is exactly right. If you had a coordinated 16 process, because what Bob says is totally true. In 17 the real world when somebody, a couple of people, you 18 know, get infections, you do a QI process and you have 19 to have that in place for JCAHCO approval. So you 20 have already got that set up. And then the same 21 people who sit on the QI Committees are also in 22 academic centers. Faculty members who sit on IRBs. 248 1 So they are conversant with both processes, so you do 2 a QI process. 3 You do a review and then somebody says 4 hey, that's really important. We should publish that. 5 Then you go to your IRB and you get a waiver as an 6 exempt process, so that you can publish it, because 7 you just have to have that IRB approval to get it 8 published. And it would be ever so much more 9 efficient and nicer to have some kind of, from the 10 very beginning, process that said okay, if we want to 11 publish it, we've already got the IRB approval. And 12 if it turns out not to be interesting, forget it. 13 It's just a QI process. 14 CHAIR PRENTICE: Bob? 15 DR. HAUSER: I think there is a clear 16 distinction between QI and research. QI oversight 17 belongs at the chief executive level. It belongs with 18 the Board of Trustees or the Board of Directors of the 19 health care organization. The key question is is it 20 research and if it's research, it goes to the IRB. If 21 it's QI, it comes under an entirely different 22 oversight or organizational element, and I get very 249 1 nervous when I hear that we're going to somehow 2 integrate or import or cross whatever, because these 3 are two very distinct activities. 4 CHAIR PRENTICE: Okay. Thank you. I 5 would like to thank the Panel for their presentations. 6 They were fascinating. It simply shows that there is 7 a lot on confusion over the definition of research and 8 we have a lot of work to do to try to provide 9 appropriate guidance. So if you would, please, join 10 the rest of the audience. 11 (Applause) 12 CHAIR PRENTICE: Now, I would like to ask 13 whether or not any of the members of the public who 14 are in attendance would like to address SACHRP. We 15 have one individual who has expressed an interest. 16 Are there any others? I already know about you're the 17 one individual, Bob. Is there anybody else that 18 would? Okay. 19 Bob, please, go to the microphone. There 20 is one behind you over there or over here. Would you 21 state your name for the record and your affiliation, 22 and we look forward to your comments. 250 1 MR. LEVINE: I'm Bob Levine. My voice is 2 not working well today, because I have been overusing 3 it. It's appropriate punishment. One of my -- oh, 4 I'm at Yale University School of Medicine. One of the 5 assignments I had many years ago was to be a special 6 consultant to the National Commission for the 7 Protection of Human Subjects, etcetera, and the first 8 assignment they gave me was to write the definition of 9 research and to distinguish it from practice. 10 And it might interest you to know that two 11 of the criteria that we explicitly rejected, one was 12 the idea that you could define this based on the 13 intent of the individual who is doing it. I put 14 intent in my first draft, and was argued out of that 15 by one Joe Brady, a radical behaviorist. And those of 16 you who know a radical behaviorist will know they 17 don't believe in intent, because you can't see it. 18 And so that's why the word that ended up in the 19 definition is designed not intent. Another thing we 20 rejected right off the bat was whether or not 21 something was destined to be published, and you have 22 heard some of the reasons for that here at this 251 1 meeting. 2 But I want to talk about the continuing 3 efforts over the years to redefine research. I think 4 what the National Commission's assignment was was to 5 distinguish research from medical practice, and I 6 think it did a reasonably good job at that, but it's 7 very clear that this is not helping us deal with some 8 other sorts of problems that we have confronted since 9 the 1970s. 10 I have been Chair of the External Review 11 Committee for a CDC which reviewed this. I am part of 12 Mary Ann Baily's QI project, etcetera, etcetera, 13 etcetera, and every arena I find myself in I try to 14 plead with the participants, don't write a new 15 definition, because there's other people like you 16 doing different things all over the country and they 17 are all trying to write a definition, and each one of 18 them is trying to write the definition so that it 19 excludes some cherished activity that their group 20 carries out, but if they get their way it's really 21 going to screw you and if you get your way, that's 22 what it's going to do to them. 252 1 And so what I have been recommending 2 instead is that they go the exemption route. The 3 example I use is the exemption from coverage by 4 federal regulations for research evaluating Public 5 Benefit Programs. Many of you know the history of 6 this and Ernie will give me enough time to tell it to 7 you, but in this exemption the category of activity 8 that they are exempting presents far greater risks 9 than anything I have heard about this afternoon. And 10 yet, for an important national priority, they created 11 this exemption and it doesn't have to go through 12 anything. 13 The final comment I want to make is when 14 you exempt an activity from coverage by the 15 regulations, that does not mean you absolve them of 16 all ethical responsibility for doing their work. And 17 I think you heard very clearly from Mary Ann in 18 particular how she wants to have an accountability 19 structure built in there, something that would do the 20 kind of job that the IRB does for research, but it 21 would do it with special sensitivity to the particular 22 problems of quality improvement. Thank you very much. 253 1 CHAIR PRENTICE: Thank you, Bob. We 2 appreciate your comments. You have been part of the 3 history of the development of the regulations and for 4 those of you who are as old as I am, you're aware I'm 5 sure of Bob's definitive book. It was the first one. 6 The Ethics and Regulation of Clinical Research is the 7 book that I went to when I first got involved in IRBs, 8 as many of us did, to try to find out, you know, what 9 these regulations meant, so we all owe a debt of 10 gratitude for Bob's comments and his presence here 11 today, because he is going to be speaking to us in a 12 little while about different IRB models. 13 Okay. Are there any other individuals who 14 wish to address SACHRP? All right. We're going to 15 take a break and I would really not want to go any 16 more than 15 minutes for a break. So would you, 17 please, reconvene at 2:40. Thank you. 18 (Whereupon, at 2:24 p.m. a recess until 19 2:42 p.m.) 20 CHAIR PRENTICE: If people would take 21 their seats, please. Okay. Now, we have a panel on 22 models of IRB review. As everyone knows, we started 254 1 off back in the early '80s with one basic concept of 2 IRB review. That was local IRB review. And the 3 entire climate of clinical research has changed over 4 the last 23 years and now, we have multiple versions 5 of IRBs. 6 So this should be a very interesting 7 Panel. The organizers of the particular Panel are 8 Susan Weiner, who is a member of SACHRP, and Amy 9 Patterson. So before I call up the first speaker, I 10 would like to ask Susan to make a couple of comments 11 concerning this issue. Susan? 12 DR. WEINER: Thank you, Ernie. I would 13 like to thank the Panel in advance for coming today 14 and for sharing their expertise with us. I just 15 wanted to make one comment about this topic from the 16 patient and public perspective. I have had the direct 17 experience of participating in the design and approval 18 of a protocol that took between two weeks and 18 19 months to approve for a multi-site study for children 20 with cancer. I don't think anything brings home more 21 dramatically that contrast. 22 Efficiency is a vital thing from a 255 1 patient's perspective. Time has a different meaning 2 for us than it does for research and, indeed, I have 3 been both and understand it from both sides of the 4 table, but it's our wish today not to rewrite the regs 5 perhaps, as the other panels discussed at various 6 points in the morning, but really to come up with 7 greater ways of making this process of review, 8 appropriate ethical review, as science becomes more 9 sophisticated just to make it more efficient. Thank 10 you very much. 11 CHAIR PRENTICE: Thank you, Susan. All 12 right. Everybody knows the format. We have 13 individuals come up here and present and then 14 everybody else, at the end of the Panel, go to the 15 table and then we will respond to questions at that 16 particular time. 17 So I would like to invite Dr. Michael 18 Carome to come up and he was a previous speaker, so 19 I'm not going to introduce him again. So, Michael, 20 we're looking forward to your comments. 21 DR. CAROME: Okay. I think for OHRP this 22 is a less controversial area than the definition of 256 1 research. What I'm going to quickly do in a few 2 minutes is just make you aware of the relevant 3 regulatory provisions and OHRP guidance document that 4 is relevant to the topic of having joint IRB review of 5 cooperative projects. And so I'm going to talk about 6 the relevant regulatory provision. I'm going to 7 mention how the FWAs facilitate doing this and 8 reference a guidance document on IRB knowledge of 9 local research context, and you all know the 10 regulations we're talking about. 11 The key provision relevant to this topic 12 is found at 45 CFR 46.111, cooperative research, and 13 it has three sentences, which say the following. 14 First of all "Cooperative research projects are those 15 projects covered by the HHS regulations, which involve 16 more than one institution." So the regulation first 17 defines what a cooperative research project is, and it 18 could involve two institutions, it could involve 19 1,000. 20 Secondly, 114 goes on to say that "In the 21 conduct of cooperative research projects, each 22 institution is responsible for safeguarding the rights 257 1 and welfare of human subjects and for complying with 2 the HHS regulations." So all the sites involved in a 3 research project, despite which IRB reviews it, have 4 certain obligations to protect the rights and welfare 5 of the subjects participating at their site. 6 And then here is the key part of this 7 provision. "With the approval of the department or 8 agency head, an institution participating in a 9 cooperative project may enter into a joint review 10 arrangement, rely upon the review of another qualified 11 IRB or make similar arrangements for avoiding 12 duplication of effort." 13 So I think the key point of that provision 14 is that the regulations are not an impediment to 15 having cooperative review of research by one IRB on 16 behalf of multiple sites. Whether that's a central 17 IRB or some regional IRB, the regulations were written 18 to permit this. And under the most common assurance 19 process we have today, the Federal-wide Assurance, the 20 ability to do this is fairly streamlined. 21 Under the FWA, an institution can 22 designate any registered IRB of another institution 258 1 and in that mechanism, we approve that and allow that 2 other IRB to review on your behalf and multiple 3 institutions can designate under the FWAs the same 4 IRB. When this is done, there should be an IRB 5 Authorization Agreement between the institution that 6 holds the FWA and the institution whose IRB is being 7 relied upon. But with those mechanisms it's very 8 simple to set up the arrangements for joint review. 9 The last thing I would like to note is 10 when an IRB that's reviewing on behalf of an 11 institution is geographically distant, it's important 12 that there be mechanisms in place for that IRB to be 13 knowledgeable about the local research context of the 14 institutions they are reviewing for, and we have a 15 guidance document that outlines mechanisms for doing 16 that. 17 Now, when a single IRB perhaps, a central 18 IRB as it's sometimes referred to, is reviewing upon 19 multiple, a large number of sites such as maybe 50 or 20 100 like the NCI Central IRB, our guidance document 21 talks about a mechanism in which there can be a shared 22 review process and that's outlined here, Section B.(2) 259 1 of this guidance document. The institution relying 2 upon another institution's IRB has the responsibility 3 to ensure that the particular characteristics of its 4 local context are covered either through knowledge of 5 its local context by the reviewing IRB, and the rest 6 of the document talks about how that can be done or 7 through subsequent review by an appropriate designated 8 institutional official, such as the chairperson and/or 9 other member of its local IRB. 10 So again, I tried to give you an overview 11 of the relevant regulatory provision and guidance from 12 our office on this topic and I'll stop there. 13 CHAIR PRENTICE: Thank you, Michael. 14 (Applause) 15 CHAIR PRENTICE: Our next presenter is Dr. 16 David LePay. Everybody knows Dr. LePay. He is the 17 Senior Advisor for Clinical Science and Director of 18 Good Clinical Practice Programs within the FDA's 19 Office of the Commissioner. Welcome, David. 20 DR. LEPAY: I'm having problems with this 21 set up, so I'm going to let her proceed here. Very 22 good. I'm going to start out pretty much reiterating 260 1 the point that Mike had made, and that is certainly 2 from FDA's perspective we have always allowed for 3 cooperative review agreements and made provisions for 4 cooperative research in the regulations. Oh, I'm 5 sorry. Thank you. 6 I titled this deliberately "Perspectives 7 on Central and Local IRBs" just to make a couple of 8 basic points. First, from the standpoint of 9 terminology, the terms central and local IRBs are not 10 regulatory terms. They don't appear anywhere in FDA 11 regulations. We define Institutional Review Board. 12 And from that, that really leads to our second 13 perspective based on expectations. 14 Because we have a single definition of 15 IRBs, we accept review of research by any of the 16 constructs that one can anticipate that are compliant 17 with this particular definition. Be it central, 18 commercial, academic, local IRBs and so forth, the 19 expectations from a regulatory standpoint are 20 identical for any kind of IRB, if you will, because, 21 again, the common definition, that is all IRBs must 22 meet the requirements at FDA set forth in 21 CFR Part 261 1 56. 2 A point that I did want to raise, because 3 again with respect to FDA regulated research, we are 4 in the majority generally outside of the academic 5 sphere. More research is conducted in support of FDA 6 applications in nonacademic than in academic settings 7 and, indeed, just to reinforce, it is in FDA 8 regulations the clinical investigator's responsibility 9 to select an IRB that is compliant with FDA 10 regulations. Sponsors have the ability to select the 11 clinical investigators. The sponsors do not select 12 the IRBs. That is under the regulations a 13 responsibility of the clinical investigator. 14 The expectation, however, is that if the 15 clinical investigator's site is affiliated with an 16 institution that has its own IRB or institutional 17 policies regarding that IRB review, such as an 18 assurance, the clinical investigator is expected to 19 first approach that institutional IRB. 20 A fourth perspective deals with 21 cooperative agreements, and that is something we need 22 to remember. Any IRB can serve as a central IRB. We 262 1 don't have to have a function that you are identified 2 as a central IRB for marketing purposes or whatever. 3 Cooperative research agreements, basically, were 4 developed with the point of trying to streamline the 5 process where possible for reliance on joint review in 6 these kind of multi-center protocols. 7 A couple of other very important points 8 here. First of all, again, no local IRB is ever 9 required under regulations to enter into a cooperative 10 research agreement for centralized review. That is at 11 the discretion of the institution. And secondly, and 12 I think very importantly, cooperative agreements under 13 FDA regulations do not need to be all or none. That 14 is agreements may apportion part of the review 15 responsibilities between a central IRB and a local 16 IRB, which allows for several constructs that we may, 17 indeed, want to talk about and think about as how 18 processes can best be conducted. 19 With respect to cooperative review 20 agreements, there are just some very basic 21 expectations. Obviously, subject protections must be 22 maintained. Any local IRBs having jurisdiction over 263 1 the investigator needs to be in agreement to the 2 arrangement. There should be documentation of these 3 cooperative review arrangements, and the reviewing IRB 4 must be appropriately constituted and adhere to all 5 U.S. regulations, as well as state and local laws. 6 Talking about documentation, we need to 7 recognize in these systems of cooperative review there 8 are at least five parties we have to consider as 9 having an interest in the research and, therefore, 10 having an interest in the way that this review will be 11 conducted, the investigator or investigators at the 12 site, the central IRB, if you will, the local IRB, the 13 local IRB's affiliated institution and the study 14 sponsor. 15 And of course, much like OHRP, we 16 recommend that written agreements be executed to 17 discuss these cooperative arrangements, in particular 18 between the local institution and the central IRB, 19 recognizing that is the way most such arrangements are 20 made. But clearly, this is not going to be an 21 effective system if copies of the agreement and an 22 understanding of the agreement is not made available 264 1 to the local IRB, the investigators at the site, as 2 well as the study sponsor. And surprisingly, that 3 does not happen in some cases. 4 As far as local IRB review or local 5 review, this is something we have had considerable 6 discussion about in FDA. We highlight, as our 7 regulations note, the need for IRB members to have 8 adequate knowledge of community attitudes, 9 institutional policies and state and local laws that 10 apply at the research site, but FDA has never 11 established local IRB review as an explicit regulatory 12 requirement. 13 As for compliance, well, this is something 14 I can't say a whole lot about in terms of a 15 distinction here, we do conduct a fair number of IRB 16 inspections a year, approximately 300, but because we 17 work off of a single definition, a single set of 18 compliance programs, a single set of regulations, we 19 don't subclassify IRBs and don't develop inspection 20 metrics based on any specific subset or type of IRB. 21 I can say for the IRB community as a 22 whole, as a result of our inspections there are 265 1 relatively few instances where we have to take an 2 official action. In general, we're talking about 2 to 3 3 percent of the inspections we do per year and many 4 of even that percentage are based on for cause versus 5 routine surveillance inspections. 6 This is not to say that we don't see value 7 in inspection from the standpoint of education, 8 because we, obviously, do recognize deficiencies in 9 our inspection system, be they at central IRBs, be 10 they at local IRBs that we still think deserve 11 attention and can certainly improve the process. So 12 my prospective Number 7 on compliance really goes back 13 to the issue. There are a very large number of good, 14 local, academic, central or commercial IRBs and there 15 are a few bad local, academic, central and commercial 16 IRBs. 17 Now, that's, of course, the issue of 18 performance from a compliance standpoint and, as I 19 said, we can't even make that breakout based on the 20 way -- that we don't have that kind of categorization. 21 Obviously though, compliance is only one of the 22 indicators we need to talk about when we look at the 266 1 relative advantages of systems. We do need other 2 indicators and we need data on these from the 3 standpoint of participant protection. 4 How well do we perform in that area? How 5 well do these systems perform in participant 6 comprehension, in responsiveness to the participants 7 and to things, either questions or things that may 8 happen in the trial, to the quality in terms of 9 staffing processes and products, in terms of 10 efficiency, in terms of transparency? 11 And hopefully, some of the other speakers 12 in this Panel may be able to shed a little light on 13 some of these performance measures, because certainly 14 I don't have them and I'm always interested in 15 learning what is out there and what is available. 16 So in terms of closing perspectives, I 17 think at least on the abstract, on an umbrella 18 approach, can central IRBs and local IRBs provide 19 equal protection of subjects' rights? Certainly, what 20 we have seen over the years at FDA would lead us to 21 believe or to say that the answer to that is yes. Do 22 central IRBs and local IRBs always provide equal 267 1 protection of subjects' rights? I don't think we 2 really have the performance data there to make a 3 complete answer, at this point. 4 And two, as we start to look at the 5 comparative systems, we may need to start pulling out 6 individual functions and go back and say how was this 7 better conducted in a central system versus a local 8 system and maybe learn, in fact, where differences 9 occur, how we can make the other system work better in 10 that particular category. 11 And so I'm finally going to close with one 12 last remark. Will we have more to say on the use of 13 a centralized IRB process and multi-center clinical 14 trials? The answer is yes, and FDA does expect to 15 provide additional guidance and I will say in the very 16 near future. 17 I will answer two questions that came up 18 earlier, one that I think was addressed through Ada 19 Sue and the Panel before, and that dealt with the 20 questions of flexibility. Obviously, the goal of our 21 guidance is to provide or to address flexible systems 22 and how to employ these flexible systems. 268 1 And secondly, of course, I think there is 2 still information that we need to acquire in this 3 process, but as we move forward with our guidance I 4 will emphasize we don't work in a vacuum anymore. 5 Anything that we put out or develop in guidance we 6 share quite freely with our colleagues at OHRP and NIH 7 before it goes out for public comment. 8 Even if we're not talking about joint or 9 simultaneous guidance, you can be sure that all of 10 these issues, all of these comparisons are done in the 11 process of any one of our agencies developing 12 guidance. So I'm sure I have used more than my five 13 allotted minutes, but thank you very much. 14 (Applause) 15 CHAIR PRENTICE: Thank you, David. We 16 appreciate your comments. Susan Kornetsky is our next 17 panelist. Susan, as you probably know, is the 18 Director of Clinical Research Compliance at Children's 19 Hospital in Boston and, obviously, a member of SACHRP. 20 Susan, we're looking forward to your comments. 21 MS. KORNETSKY: Thank you. I guess I was 22 asked to do this, because I have been involved, like 269 1 Ada Sue, with IRBs for the last 20 years, have seen a 2 lot of change, and really think about this from the 3 IRB's perspective and what it means. So what I want 4 to do is just briefly go over some of the issues that 5 face IRBs and some of the models, and I think the 6 speakers who follow me are going to go into these 7 areas in a lot more depth. 8 So one of the issues, I think Mike spoke 9 and David spoke about, the first one is can one IRB 10 accept the approval of another? Why is this an issue 11 now? Why is there resistance? What are the barriers 12 and then examples of a couple of models. Mike already 13 had this slide up here. I just want to add something, 14 sort of a side note to this. I originally did this 15 talk last February for bone marrow transplant 16 specialists in Orlando, and it was mostly 17 investigators who were in the audience, and I saw 18 people sort of ripping out of the slides this 19 particular slide that they were going to quick, run 20 back to their institution and say this can be done, 21 this can be done. And I think we all have to realize 22 that the regulations allow that and we need to begin 270 1 to use it. 2 So why now? Well, obviously, the way we 3 have conducted clinical research has changed. When 4 the regulations were written and when IRBs began, most 5 institutions, there was one project done in one 6 institution. There was a single institutional funding 7 and now, this has changed to multi-institution 8 research for many and good reasons. There has also 9 been shifts of funding from federal to corporate 10 sponsors. There is increased relationships between 11 academics and industry. There is a general increase 12 in clinical research. There is a rapid pace of 13 technology. We have more research, more things that 14 we want to try. 15 The other thing that no one has really 16 mentioned is mergers. This really hit home when lots 17 of institutions started merging with each other. 18 Everyone had their own IRB and all of a sudden there 19 are mergers and conglomerates and parent institutions. 20 And for awhile everyone was, you know, wanting to keep 21 their own IRB. And then also the idea that research 22 has moved to the community, so we have a lot more 271 1 research going out in the community who may or may not 2 have access to IRBs. 3 So why is there resistance? Why isn't 4 everyone -- you know, jointly, we have talked a lot 5 about regulatory burden. Why wouldn't IRBs say fine, 6 let someone else just do this? Well, IRBs and 7 regulations were born out of a culture for local 8 review and protectionism, and I'm going to talk sort 9 of a little bit about thinking about community a 10 little bit different than sort of a local context, 11 because I think that's something that isn't explored 12 enough. 13 IRBs are territorial. There is no 14 question about it. They have in their assurances, you 15 have to think about who comes in under your net. Who 16 do you define? Who are you responsible for? Who are 17 you going to review for? Liability and legal 18 implications also are always a big topic. As a side 19 note, at Children's Hospital our oncology patients are 20 physically joint patients with the Dana Farber. They 21 are in our hospital for inpatient. They are at the 22 Dana Farber for outpatient. It took years, not 272 1 because IRBs didn't want to sort of say let's stop 2 with two different IRBs reviewing this, it took years 3 to get over some of the concerns about legal and 4 liability. So I think that's a real issue in sort of 5 relinquishing your review to someone else. 6 IRBs, you know, we know our research 7 participant community best. You know, no one else 8 knows that. We know best. We need to review. That's 9 an attitude that's out there. And the regulatory 10 environment, frankly, and although it's changing, has 11 not been conducive to creativity. IRBs have spent a 12 lot of time being over-reactive, over-restrictive, 13 worrying about what's going on there and it hasn't 14 been easy to say let's think about something a little 15 more creatively. 16 And Dr. Schwetz talked about trust, 17 distrust of IRBs among other IRBs, and I think this is 18 something that we also need to think about. We need 19 to start thinking about trusting IRBs to feel 20 comfortable that someone else can review this and we 21 feel comfortable. So these, I think, are some of the 22 key areas of why there is resistance. I'm not sure 273 1 about the solutions or how to change it, but I think 2 from a very sort of everyday practical level, these 3 are the things that often go through the minds of 4 IRBs. 5 So centralizing the IRB review process. 6 I talked a little bit about sort of this idea of 7 bringing community. We often stop and say, you know, 8 we live in Boston, we know our community, but I think 9 we need to begin to think about community as far as 10 individuals with a disease or a disorder or a 11 condition. There may be some differences in receiving 12 chemotherapy or other things, Children's versus some, 13 you know Children's Hospital and another institution, 14 but there probably are a lot more similarities that we 15 can think about about thinking about community. So it 16 may not have to be that an institution has to sit 17 locally there. 18 And so some of the examples of different 19 models that have been used, and I'm not intimately 20 involved with any of these, I don't have a lot of 21 details, but sort of have heard about different things 22 and approaches that people are beginning to take. 274 1 Obviously, I think, the one that a lot of people know 2 about is the NCI model and this is actually now moving 3 to the COG, to the pediatric group, where they are 4 trying to get this started, so that there is some type 5 of central IRB for pediatric trials. 6 Then another model is sort of a group of 7 institutions getting together and one institution or 8 organization leading it or possibly rotating it. 9 There's a group called MACRO, and I will talk a little 10 bit about that, that has formed that is a group of 11 institutions on their own for whatever reason. I 12 don't know the reason that these particular 13 institutions chose to do this. Use of an independent 14 IRB, and you will hear more about that, and then also 15 models of creating IRBs from members from each 16 institution. And I'm sure there are a lot more 17 models. These were just some. 18 And so some of the advantages, we often 19 talk a lot about conflict of interest. Institutions 20 have their own IRBs and although we're autonomous and 21 we make our own decisions, many times within a local 22 context we are there acting on behalf of our 275 1 institution. So there is in some way some inherent 2 conflict of interest. A lot of the arguments I'm 3 hearing is if you utilize a central thing, they may 4 remove that somewhat. 5 Other arguments that I have heard is 6 attracting sponsors, and this is why I refer to issues 7 of industry and going to corporate funding, that it's 8 often attractive and becoming more attractive for 9 sponsors to be able to go to one IRB, as opposed to 10 200, and have that approved. So there may be 11 institutions that are going to be, you know, "losing 12 business" if you want to look at it that way, because 13 every institution, you know, has to review something. 14 And then also one of the important things, 15 and Susan alluded to this in her opening comments, is 16 the timeliness of review and just the elimination of 17 duplicative reviews. So one organization, BRANY, 18 which is the Biomedical Research Alliance of New York, 19 and I again have no idea how successful or 20 unsuccessful, but these are at least attempts and I 21 have to give people credit for attempts, established 22 by the New York Academy of Medicine. It's research 276 1 administration and central IRB services to sponsors 2 and institutions. There is one central IRB and IRB 3 members are from surrounding academic institutions, 4 and there may be people here who can answer questions 5 more about this model or are involved in it. 6 MACRO is another group. Now, why these 7 institutions chose to get together, Baylor, University 8 of Alabama, University of Penn, Vanderbilt, University 9 of Washington and they have set up a system of IRB 10 reciprocity. The five sites volunteered to work 11 together and for this particular model, each trial, 12 one IRB will serve as the primary reviewing 13 institution. There is no specially constituted IRB. 14 The information about the review is circulated to all 15 the institutions. 16 The process for determining which sites 17 assume IRB review may depend on the sponsor or which 18 centers are going to cooperate. And again, I think 19 one of the hopes here, again, was to attract sponsors, 20 so that these are well-established academic 21 institutions. You can go to one IRB and the trial is 22 conducted in multiple sites, at least with an 277 1 additional IRB. 2 The NCI model here, this is receive a 3 protocol from a cooperate group. IRB members are from 4 multiple academic and community organizations. There 5 isn't one individual from each institution, but from 6 the cancer community in general. They review and 7 approve the protocol and here, after that review, it 8 then moves to a local level where locally, the IRB, 9 there is a local PI. There is an individual 10 designated to review the materials to determine if 11 there are local concerns, whether to accept it, or 12 they may decide that they want to put it through some 13 type of review. 14 Consent changes are allowed at local 15 institutions. There is a method for notifying the 16 central IRB when the protocol is accepted. The 17 central IRB continues to conduct the subsequent 18 reviews and the local IRB and PI receives notification 19 of review. So this is sort of a two tiered system 20 that has developed and a similar system is being 21 thought about and developed for the pediatric oncology 22 group, COG, as well. The slides are out of order. I 278 1 apologize. 2 Central IRBs. You're going to hear more 3 from David. This is an independent, usually for 4 profit IRB. They review for individuals or 5 institutions with no IRB. There are a lot of places 6 that don't have an IRB. They are also utilized for 7 organizations who have designated a central IRB or an 8 IRB record either for some or all protocols, so that 9 you still can have your own IRB and designate a 10 certain subset to that, and David will be able to 11 answer that. 12 I just want to point out that the idea of 13 trying to think about levels and types and new ways of 14 review was referenced in the National Commission's 15 Advisory Report. In 2003 the Institute of Medicine 16 also talked about this and also, actually, the most 17 recent report in the pediatric group also and it's 18 particularly important in pediatrics, because there 19 are very small numbers of individuals, thank goodness, 20 in children, so when you do trials you need lots of 21 children. So you're automatically going to lots of 22 different institutions even outside of the oncology 279 1 group, cystic fibrosis and there are other disorders 2 as well. 3 So to sum up some of my conclusions is we 4 need to think outside the box. There is a growing 5 need to consider alternatives. We need to let go to 6 some of this baggage that we have accumulated about 7 thinking that, you know, we are the only institution 8 or the only IRB that can review it. I think we have 9 to strive to centralize IRB reviews. I think, in 10 thinking about it from an IRB, sort of this accept or 11 reject approach may help at least initially gain some 12 trust and confidence. 13 I think one of the important things, 14 whatever approach moves forward, is this idea of 15 informed consent. IRBs are very possessive about 16 their informed consents, and I think in order to 17 succeed there needs to be some mechanism that perhaps 18 there can be a central IRB reviewing, but then again 19 perhaps an IRB can then take that informed consent and 20 individualize it. I'm not sure of sort of all the 21 mechanics that we would use, but I think if anything 22 has -- in trying to be creative and move forward, that 280 1 this may help get over, you know, the bump with some 2 IRBs if somehow this can be incorporated. 3 And if the central review approach is 4 rejected, which I hope it isn't, at least try to 5 improve the communications between IRBs about review 6 outcome, and I think there are efforts underway, but 7 I sort of wanted to take that off, because I don't 8 think we should reject it. I think we should move 9 forward, but it's another alternative. 10 So I also will apologize. I need to leave 11 early, so I won't hear all the questions and answers, 12 but I will talk to my colleagues afterwards to make 13 sure that I'm caught up on things. Thank you. 14 (Applause) 15 CHAIR PRENTICE: Thank you, Susan. Our 16 next presenter is Dr. Robert Levine. He is a 17 professor of medicine at Yale and also Co-Chair of the 18 Yale University Interdisciplinary Project in Bioethics 19 and, of course, Dr. Levine is well-known to all of us 20 and really doesn't need an introduction. Welcome to 21 SACHRP, Bob. We're looking forward to your comments. 22 MR. LEVINE: I want to thank you very much 281 1 for inviting me to be here, and I especially want to 2 thank your Chairperson for his generous introduction 3 of me during the Q&A session of the last Panel. Ernie 4 has already presented my first slide, but I'm going to 5 review a little bit of history to show how the 6 original intent was to have this IRB process as a very 7 local process. 8 The IRB was named an IRB in 1974. This 9 was a compromise between the FDA's Institutional 10 Review Committee and the Department of Health, 11 Education and Welfare's Organizational Review Board. 12 I have tried to color code this to see which names 13 survived. The review was common to both, and 14 institutional came from FDA and board from DHEW. That 15 was quite an achievement in political compromise. 16 Now, in 1978, the National Commission had 17 this to say about review. The ethical conduct of 18 research requires a balancing of society's interest in 19 protecting the rights of the subject and developing 20 knowledge that can benefit society as a whole. I 21 think it's very important to keep in mind that at the 22 beginning of all this, the IRB was not seen at all as 282 1 an agency that simply protects research subjects from 2 investigators. You will see all through the National 3 Commission's report that it was expected to balance 4 the needs of science, of the institution with the 5 rights and welfare of the subject. 6 The Commission also went on to state its 7 concern with conflict of interest, but in a language 8 that may seem very alien to us today. Review is 9 required because investigators are always in potential 10 conflicts of interest by virtue of their concern with 11 the pursuit of knowledge, as well as their concern 12 with the welfare of the human subjects. 13 They went on to say that local committees 14 located in the institutions where research is 15 conducted compared to the possible alternatives of 16 regional or national review, local committees have the 17 advantage of greater familiarity with the actual 18 conditions surrounding the conduct of the research. 19 It went on as follows. Local committees can work 20 closely with investigators to assure that the rights 21 and welfare of the subjects are protected and that the 22 application of policies is fair to the investigators, 283 1 contribute to the education of the research community 2 and the public. IRBs should serve as local resource 3 centers concerning ethical standards and communicate 4 with federal officials and other local committees. 5 The IRB then is perceived as being in a 6 collegial relationship with the investigators who are 7 the faculty and students in the same institution, and 8 much was said about the necessity of having this, a 9 collegial relationship. This seems very odd in the 10 light of recent discussions about how they really 11 should be serving as guardians of the rights and 12 welfare of the subjects. The IRB is perceived as an 13 agency that would balance the interest of 14 investigators with those of the subjects. 15 I think DHEW anticipated inconsistencies 16 across IRBs with the requirements you have already 17 heard about for competence to judge a proposal in the 18 terms of community attitudes. It was really believed 19 that something that might be found approvable in one 20 part of the United States might be found unacceptable 21 in another part of the United States, and there has 22 been quite a bit of literature in which there are 284 1 efforts to document inconsistencies. And my response 2 often has been well, what did you expect? 3 Now, the non-Institutional Review Board, 4 I think, was the first name given to this odd 5 creature. This was a name that was coined by Sam 6 Berman when he was Chair of the FAC organization IRB 7 in Philadelphia, one of the very early CRBs, NRBs, 8 whatever. The original purpose of this was to respond 9 to FDA's new or new in 1981 requirement for IRB review 10 of all research. Until 1981, FDA only required review 11 of research that was carried out in institutions. But 12 so much of Phase Three studies in those days was one 13 in doctor's offices. 14 So much of it is also done in doctor's 15 offices now where we get information about how does 16 this drug devise biological whatever? How does it 17 work? What is its safety? What is it's use and 18 benefit profile under conditions that simulate the 19 actual practice of medicine? 20 Now, I would want to make a preliminary 21 comment on these non-Institutional Review Boards or 22 Central Review Boards. If all we consider are 285 1 measurable activities that indicate literal compliance 2 with regulations, there is nothing the local IRB can 3 do that cannot be done equally well, perhaps a bit 4 less conveniently, by the Central Review Board. It 5 can do all of those things, that's why, I think, they 6 fare very well on audits, on inspections by FDA 7 inspectors and so on. 8 What about measurability? Jerry Mashaw is 9 a professor in the Law School at Yale and the stuff 10 I'm going to present to you now is taken from a paper 11 he wrote for the President's Commission in 1983. The 12 President's Commission for the study of ethical 13 problems in medicine and in biomedical and behavioral 14 research. He had several models of decision making 15 groups. The one that he said was best for the IRB was 16 a blend of two of these models. The wise judge and 17 the paternalistic professional. 18 However, Mashaw said if we are to have 19 wise judges and paternalistic professionals, we can 20 neither specifically direct nor objectively evaluate 21 their behavior. And this presents a great problem, 22 particularly when there are public demands for certain 286 1 types of accountability. So much of what has happened 2 since the early 1980s has been to apply constant 3 pressure to force the IRB to operate according to what 4 Mashaw calls the "bureaucratic rationality model." 5 In that model, everything they do is 6 measurable, everything they do can be recorded. As 7 some officials from OPRR and subsequently OHRP have 8 said, "If it's not documented, it didn't happen." And 9 so much of what the original purpose of the IRB was 10 was to have people get reason together and come out 11 with mutually agreeable resolution of problems. In 12 fact, the IRB at Yale was started in 1960. When I 13 became chair in 1969, I introduced the innovation, 14 perhaps an exercise in quality improvement, of 15 numbering the protocols and developing a filing 16 system. That seems so other worldly. 17 Now, the attitudes in the '70s and '80s 18 surfaced on the IRB was regarded as desirable, an 19 opportunity for a professor or a student to make 20 important contributions. IRB membership included some 21 of the most senior and respected members of the 22 faculty. Students actually competed for appointments 287 1 to the IRB. We always had one medical student from 2 each class, so we don't meddle in the selection 3 system. 4 When the students come to Yale, they are 5 asked to identify, to develop a list of the 6 committees, the faculty committees they would like to 7 serve on. One year, 64 out of 104 entering freshman 8 medical students identified the IRB as their first 9 choice and the Student Council had an essay 10 competition. 11 Concerns about credibility. When I first 12 wrote about this in the first edition of my Ethics and 13 Regulation of Clinical Research Book, I said something 14 that I believe to this day, that, in my opinion, the 15 single most important factor that contributes to the 16 successful functioning of an IRB is its credibility 17 within the institution that it serves and within the 18 community that the institution serves. In that first 19 edition of the book, I listed some things that present 20 threats to the credibility of the IRB, and all I'm 21 going to do is read you the list of subtitles in that 22 passage of the last chapter. 288 1 I said "The IRB should be seen as an agent 2 of its own institution," not as one of my colleagues 3 put it a "deputy sheriff" of the Federal Government. 4 "The IRB should avoid double standards." By that, I 5 meant it should not say something is required if it is 6 paid for by NIH and not required if it's not paid for 7 by NIH. "The IRB must seem to support defective 8 regulations." Even in 1981, there were some things in 9 the regulations that most people in the IRB world 10 thought considered defective. 11 "IRBs invest too much time and energy 12 doing unimportant things." And that's greatly 13 escalated since. "The IRB is not a police force." 14 Some federal policies preempt the authority of the 15 IRB. "The IRB must have adequate membership." And 16 finally that "Perfection is not the proper standard of 17 performance for the IRB." 18 I was hoping -- oh, good. I link the 19 IRB's success to these factors. Firstly, that it must 20 be an agent of its own institution. It must be able 21 to stay within the university community. We are here 22 to assist our colleagues on the faculty in upholding 289 1 the values of the institution. These are the values 2 of our institution and we take them so seriously that 3 we want to provide the assistance of IRB review to you 4 to comply with all of these things. 5 I also pointed out that an extremely 6 important component of the IRB system was motivated 7 IRB Members. Well, what motivates IRB Members? I 8 think, first, there is a wish to be of service to the 9 discipline and to the institution. A wish to be 10 respected within the institution. Some people from 11 outside said well, you can motivate people to serve on 12 IRBs by adding to their salaries. But this really 13 doesn't understand what we think is the real coin of 14 the academic realm. If these intelligent people were 15 primarily motivated by money, they probably wouldn't 16 be working for universities. 17 They say you can give them release time. 18 Well, that might be good in the English Department, 19 but that doesn't work in the medical school, for 20 example, where every minute of every day you can't 21 really tell what you are doing, teaching or research 22 or clinical practice. Whenever you are doing one, 290 1 you've got a flock of students or residents or 2 something around learning from what you are doing. So 3 release time simply makes no sense. 4 What factors do I think have eroded 5 motivation? Well, in the last 10 years there has been 6 an increasing amount of public portrayal of the IRB as 7 an institution and the IRB Members, IRB individual 8 members as either incompetent or worse. We have seen 9 really severe criticism from Government Commissions 10 from the Office of OIG, and we have seen great 11 journalistic excesses, particularly around the 12 closings of the research institutions. 13 Members of a faculty have had a tendency 14 within these institutions that have been closed to 15 hold the IRB Members accountable for these closings. 16 They say well, if you had been doing your job 17 competently, we would still be doing our research. 18 And so all of this saps the motivation of the faculty 19 member to be an IRB Member. And then the requirement 20 that they must enforce questionable policies, I cited 21 that as a problem in 1981, but there was nothing then 22 that has provoked as much anger as the decision about 291 1 secondary research subjects after the evaluation of 2 the Virginia Commonwealth University Program. 3 NRG is staffed by low-yield chores, review 4 of adverse event reports of convened leadings, 5 periodic reapproval and so on. Let's put that aside 6 for a moment. The single most important thing that's 7 detracting from the motivation of faculty to serve as 8 IRB Members is the extreme financial crisis that there 9 is within the academic medical centers. Faculty 10 members are required to generate revenue. Even at 11 Yale, we are given a quota. We are told if you don't 12 make your quota, you will not get your annual increase 13 in salary. 14 In fact, if you miss it by enough, we will 15 cut your salary. This is a powerful motivator to 16 retreat from the spirit of community service and to 17 engage in things like clinical practice or sponsored 18 research that will bring in enough money to make your 19 quota. 20 My conclusions, I think there is real 21 cause for concern about the well-being of the local 22 IRB. I think it is becoming increasingly difficult to 292 1 recruit and retain high quality members. I think it 2 is also increasingly difficult to maintain the 3 credibility of the local IRB within the institution. 4 Universities and others, as Susan so clearly pointed 5 out, are already turning to the CRBs, NRBs for relief. 6 And I believe this is likely to increase. 7 This is not likely to be associated with 8 any measurable decrease in compliance with 9 regulations. However, something of immeasurable value 10 will be lost. Thank you very much. 11 (Applause) 12 CHAIR PRENTICE: Thank you, Bob, for a 13 very sizeable presentation. Our next presenter is 14 David Forster, who presented earlier. He is Assistant 15 Vice President, Office of Compliance at the Western 16 Institutional Review Board, and I think everybody 17 knows that that is literally the world's largest 18 independent IRB located in Olympia, Washington. 19 Welcome once again, David. We appreciate you doing 20 double duty. 21 MR. FORSTER: Thank you. 22 CHAIR PRENTICE: But we're not going to 293 1 give you double pay. 2 MR. FORSTER: Okay. And I'm going to talk 3 a little slower this time, because I didn't write too 4 many slides for this presentation. So I'll talk at a 5 more reasonable pace. I was asked to talk about the 6 role of an independent IRB and how institutions and 7 industry view us, and so that was a pretty easy 8 assignment for me and I was happy to get that as 9 opposed to some of the other ones that were banded 10 about. 11 A brief outline of what I'm going to do is 12 just talk about independent IRBs and multi-site 13 trials. I'll talk about how sponsors in institutions 14 view independent IRBs and my experience. I'll talk 15 about responsibility and liability and then some of 16 the classic questions, community attitudes and local 17 conditions, conflict of interest, source of funding 18 and resources and then IRB shopping and I'll also talk 19 some about complexity. 20 So in today's world of research, there are 21 many, many players. I think we had this point in 22 earlier presentations today. It is no longer a simple 294 1 institution with an IRB investigator creating their 2 own research and getting an NIH grant. We have these 3 parties here involved in research, not to mention 4 others, such as DSMBs, pharmacies, etcetera. I mean, 5 it's just a wide range of players in research. 6 And kind of a classic graph of this, you 7 have in the upper left hand corner there sponsors, 8 that could be a commercial sponsor or a Government 9 sponsor and then sometimes you have a contract 10 research organization and then I just gave, you know, 11 an example here for sites. On the left I put an 12 institution that has the investigator and the IRB 13 internally and then an institution using independent 14 IRB, and then you have an investigator who is just 15 enrolling subjects and going straight to an IRB and 16 then you have an SMO. 17 Now, multiply this by 100 or 300. Every 18 site has a different set of players involved or often 19 times does. It gets incredibly complex and then, of 20 course, you add in things, as we'll talk about, like 21 state law, etcetera. There is a lot of complexity to 22 multi-site research. 295 1 So sponsors were the first group, first 2 sponsors, commercial sponsors that used independent 3 IRBs commonly, and as mentioned by Dr. Levine, this 4 started out with the FDA requirement in 1981 that you 5 find an IRB no matter where you find it. And so, at 6 that point, independent IRBs were mostly providing an 7 interstitial role. Providing IRB services, you know, 8 that were outside of those institutions that had their 9 own IRBs. 10 I think from the sponsor's view, we 11 provide consistency of review. If you can to go one 12 IRB instead of 200, you're more likely to get a 13 consistent answer. There are still some issues with 14 studies. We might not approve all of the 15 investigators that end up being submitted for that 16 particular protocol, but, you know, obviously, the 17 protocol is going to have much -- far fewer changes. 18 Generally, timely, not always. 19 We can provide consulting services. We 20 mentioned resources available for that. We can look 21 at protocols in advance and say yes, you know, if you 22 give us investigators, we'll prove this as it is 296 1 written. We can look at their consent forms in 2 advance and work out the hard issues, those types of 3 things. 4 A role we have been involved with a great 5 deal lately is providing IRB services for institutions 6 that have traditionally had their own IRBs, and this 7 started back in 1996 with the University of Rochester, 8 and we have had quite a bit of experience in this now 9 from 1996 to the present. And from the institutions 10 viewpoint, I think a lot of the same things apply. We 11 can provide timely review of research. We can free up 12 the internal IRB resources for other uses, such as 13 internal site visits, those types of things. 14 If the research is commercially funded, 15 usually the sponsor is willing to pay for our fees, so 16 it doesn't end up costing the institution anything in 17 that way. One thing that is an advantage for us is 18 that because we are outside the institution, we are 19 largely free from the politics, so we can apply 20 pressure where sometimes the internal IRBs or other 21 groups within the institution might have trouble. 22 And we have recently been providing a 297 1 service more and more where we provide review for 2 institutions when they have a conflict of interest at 3 an institutional level, ownership in the patents or 4 something along that type of institutional conflict. 5 So that's a role that we can provide. 6 There are a lot of issues to be worked 7 out. When you bring an external IRB into the system 8 of an institution, whether it be large or small, but 9 particularly the large ones. You've got to figure out 10 how this external IRB is going to mesh with the 11 bureaucracy and all the other committees, the IBC, 12 etcetera. Who goes first, who is going to coordinate 13 the review of which IRB or which projects will be 14 going to the outside IRB, how much control is the 15 institution going to have? Are they just going to let 16 the institutions submit directly? Are they going to 17 go through a central, some type of control point using 18 the IRB office internally? 19 How will the protocols be assigned? Is it 20 going to be commercially funded? Is it going to be 21 multi-site, including the NIH funded? Is it going to 22 be a certain category of research based on disease? 298 1 There is a lot of different ways that can be done. 2 And, of course, what you want is some means of 3 attributing the protocols where you don't end up with 4 gaming the system or internal IRB shopping. So I 5 don't think it's really appropriate to just say well, 6 the investigator can go wherever they want. I think 7 it's more appropriate to classify it by types of 8 research. 9 And then if you are going to use an 10 independent IRB for non-commercially funded research, 11 who is going to pay for it? Sometimes the institution 12 takes that and puts it in the Research Office budget. 13 Other times they require individual departments to 14 decide whether or not to do it. 15 Sometimes they just leave it up to the 16 investigator and say you want to do your own funded 17 research, you have to pay this fee. About the only 18 time that we provide free IRB services is for use in 19 a treatment IND type of situation, compassionate use, 20 so to speak. Although, we do have the occasional high 21 school student or somebody that makes such a good plea 22 to us that we say okay. 299 1 Other factors have to be worked out. How 2 is communication to be conducted? You have to set up 3 a system whereby all of the necessary information on 4 a regular scheduled basis gets to the other party. 5 And if you don't do that, things will start to fall 6 through the cracks. So that's something that has to 7 be set up. And how will noncompliance be 8 investigated? Sometimes some of the institutions that 9 we work with just said we want you all to do our 10 noncompliance. You're outside, you can do it better. 11 Others have said we don't want you doing 12 our noncompliance investigations. Give us the 13 information you found and we will do it ourselves and 14 keep you informed. And most often we kind of work it 15 out and just go along, you know, who has got more 16 resources available at the time? Who found the 17 problem originally? Those types of issues. And, of 18 course, how do we learn about community attitudes and 19 local conditions? I'm going to talk more about that 20 in a little bit, but that's something that has to be 21 worked out for each relationship. 22 So those are the issues that we face every 300 1 time we look at a new institution. So again, it's not 2 a simple thing. It's not as easy as assigning a 3 contract. You need to start putting the protocols in 4 the mail. One issue that comes up all the time is the 5 idea of responsibility and liability and people -- 6 I've heard both things. I've heard people say that 7 institutions absolutely relieve themselves of 8 liability and that they increase their liability. 9 Well, the independent IRB has liability 10 and responsibility. We can be sued. I think we have 11 a moral responsibility to protect subjects and to 12 enforce the regulations. The institution doesn't 13 relieve themselves of either the responsibility or the 14 liability by outsourcing. If we do things wrong, they 15 are probably on the hook. I put in this last bullet 16 that an institution may reduce its liability by 17 outsourcing just based on a comment by an in-house 18 counsel that came to visit recently, and they were 19 asked, you know, by one of the other members of the 20 team that was with them, you know, could you write me 21 a little memo on how much it is going to increase our 22 liability if we outsource our research. 301 1 And the attorney shot back, I can write 2 the opposite paper much easier. So that's hot off the 3 case. Though, I mean, you can outsource to an 4 independent IRB that's low quality. You can have a 5 great system in-house, so, you know, I put that in 6 there based on one comment that caught my ear. 7 Probably the thing I have been asked about 8 most in presentations in the last eight years is how 9 do you assess local attitudes with community attitudes 10 when you are sitting out in Olympia, Washington? And 11 there is really two aspects of this I found or that in 12 my mind I have divided out. One is the community 13 attitudes of subjects themselves. What do subjects 14 think? What are the cultural group attributes that 15 are common in an area that might contribute to a 16 different view of a research project? 17 And the other side of that is what is 18 local condition in the environment in which it is 19 performed? So I split them out. How do we ask about 20 community attitudes? We ask our investigators about 21 community attitudes. Now, you could say that's kind 22 of a biased source, but we do ask them on a regular 302 1 basis. We ask our institutions to tell us what are 2 the demographics? What are the local attitudes you 3 know about? 4 We try and get local representatives for 5 many of our community, where there is a specific 6 community in mind that we don't know much about. You 7 can have local representative board members. Site 8 visits provide us a great deal of information about 9 local attitudes and we did over 2,000 site visits or 10 we will do well over 2,000 site visits by the end of 11 this year and I think we'll go back up again next 12 year. We have been higher in the past times. 13 You can get a lot of information by having 14 -- each time you do a site visit inquiring about local 15 attitudes, also asking other people, taxi drivers, 16 etcetera. We also now have a database where we can 17 stockpile all of this. Now, it's not scientific, but 18 we're getting a lot of information from certain areas. 19 You can have interviews with community members, which 20 can be very effective. As I mentioned taxi drivers, 21 etcetera. And we also do a lot of work with the 22 Internet and, for instance, we have Westlaw and we 303 1 have various media searches up for IRB, certain types 2 of research phrases. 3 You can get articles about research that 4 way. What is interesting is to see an article come 5 out in one paper and then three days later it is in 20 6 papers. And you get all those hits and you can kind 7 of start to get an idea about the attitude towards 8 research, if you think that media reports are an 9 accurate reflection. I tend to think they are. And 10 what tended to happen back when the institutions were 11 sometimes being shut down by the agencies is an 12 institution would be shut down and we would get a 13 flurry of reports from investigators and the media in 14 the area. You know, you get this kind of local 15 reaction regarding research. 16 Local conditions. Again, start off with 17 asking a lot of questions. On our application form we 18 ask the investigator not just research that we are 19 reviewing for you, but for all research. You know, 20 how many research are you conducting right now? How 21 many staff do you have? How many studies? What kind 22 of on-site emergency procedures do you have? Those 304 1 types of questions that get at do you have the 2 provisions in place and the systems in place and the 3 resources and time to do research. 4 Again, site visits help that a great deal 5 to go in and actually see if they have what they say 6 they have and just confirm that things look good from 7 a visit. We get reports from our institutions, which 8 help. We ask the institutions on an annual basis have 9 you changed anything? Have you added any new 10 departments? Has there been a change in funding, 11 demographics, anything like that? And as part of our 12 contractual requirements, we often require the sharing 13 of certain information. 14 That doesn't always get done, you know. 15 There is nothing worse for us or for me, and I had 16 this experience once very vividly recently, where I 17 look at the citation letters from OHRP and lo and 18 behold there is one of our institutions that they 19 never bothered to tell us that they were having this 20 ongoing site issue. That was disturbing, even though 21 the contract required them to report those kinds of 22 things to us. So there's a lot of ways to get 305 1 information. It doesn't mean you always get it, 2 unfortunately. 3 Conflict of interest is the other question 4 I always get asked about, and it's very 5 straightforward conflict of interest. We are paid by 6 sponsor's investigators to protect subjects from 7 sponsor's investigators. That's about as straight a 8 conflict of interest as you can get. And I don't want 9 to try and underplay that or say that it's not a 10 concern. It is a concern and you have to be very 11 careful in addressing it, especially when your source 12 of income is as straightforward as ours. 13 It's not really dissipated by grants or 14 going through an institution or anything else. It 15 really comes straight to us and it's on a fee for 16 service basis, which I'll talk about later. So how do 17 you address conflict of interest? Kind of a common 18 framework for looking at conflict of interest as a 19 disclosure to parties that are interested, is often 20 used as an approach. Various other controls, such as 21 limiting who can get informed consent, those kinds of 22 things. 306 1 For IRBs I will talk about how we put 2 those controls into place or you just eliminate the 3 situation. And you know, the problems in Alberta up 4 in Canada, they just said there will be independent 5 IRBs. They eliminated the conflict. So that's one 6 way to do it. What I think are good controls for 7 independent IRBs that are operating on a for profit 8 basis or even a direct pay basis is that all, 9 preferably all of the board members would not be 10 employees of the corporation. 11 Now, we tried that and we just found that 12 we had to have doctors and nurses time for adverse 13 event review, those types of things. You just need 14 their time. Expedited review, etcetera. So 20 15 percent of our Board Members are also employees. It 16 would be nice to get rid of that, but we can't. You 17 can't have any equity holding corporation employees on 18 the board, and that's something that OHRP came up with 19 very early on in '97 or so and very appropriate. 20 We have what we call a firewall between 21 the board and the staff. Now, of course, there are 22 those 20 people, you know, like 20 percent that 307 1 operate on both, but we very much stress to our Board 2 Members, the community board members, do not be 3 concerned about the business. Do not be concerned if 4 you turn down every protocol that is on today's 5 agenda. We want you to protect subjects. And we 6 don't what you to worry about the business or even ask 7 about the business. And keep enforcing that in 8 education. 9 Then also set up a system whereby you 10 don't send memos to the board that talk about business 11 issues. You don't -- you know, sometimes it is hard 12 to separate those two. If a site goes bankrupt and, 13 you know, they know, you know, you're not going to get 14 paid for your services for that site from now on, but 15 they still have research subjects in open studies. 16 The board is going to realize that when you tell them 17 there is a problem that hey, we've got a bankrupt 18 site. 19 And so one of the things that we did with 20 that a long time ago is we just made the decision that 21 we're going to keep providing IRB services even if 22 they don't pay us on those cases to avoid having to 308 1 bring a business issue to the board. So if they ask, 2 we can just say, yes, we're still going to do it, you 3 know. We'll keep providing those services. 4 Likewise, you have to keep educating your 5 staff not to bring business issues to the board, 6 because it is very easy for somebody down in rather 7 the lower level on the staff to just think that this 8 is an appropriate item for the board or to 9 inadvertently slip it into the packet for that day's 10 board meeting, something like that. So that's kind of 11 like there is a continuing education process on that 12 side that has to take place. 13 We think that we should keep our payment 14 to Board Members low. Their payment should allow them 15 to buy cigarettes, not cars. All right. So we want 16 people that are interested in this for more than just 17 having a job. So we get a lot of retired people. We 18 get a lot of people with jobs that allow them to stand 19 a day a week out doing something else, and we think 20 that's pretty important. I think it is very useful to 21 also have as well as this division between the board 22 and the corporation or staff to have a compliance 309 1 process on the staff side as well. 2 That's a little self-serving given that is 3 my role, but that I think is important. And then, of 4 course, virtue ethics. If you can get the right 5 people on both the board and the staff, that will take 6 care of the problem. How you go about that is a much, 7 much more difficult thing. Easy to say, hard to do. 8 But you do really want to get the right people 9 involved. And again, because of the conflict of 10 interest, you know, again, I want to say that you come 11 up to tough decisions when you have this type of an 12 IRB structure. And you want to make sure that the 13 people make the right decision at the right time. 14 Now, IRBs in institutions face another 15 series of conflicts that have been mentioned a little 16 bit. The IRB Members might be involved in bringing 17 the research to the institution. They may report to 18 individuals submitting research, such as their friends 19 and colleagues and institutional officials can control 20 those priorities at times. We are largely free of 21 that, largely. I mean, the institutional officials 22 can still control priorities as far as whatever gets 310 1 to us in the first place or maybe whether we get 2 chosen for a compliance investigation, something like 3 that. 4 But I know we had a rather startling turn 5 of events at one point when one of the institutions 6 that had had an agency stop its research and was told 7 to re-review all of its research. And they decided 8 that they needed to use us for that function. We said 9 okay, we're going to start with the must vulnerable 10 subjects first and we're going to categorize by A, B, 11 C and A will be, for instance, pediatric oncology, 12 HIV, etcetera. We want to go through that order and 13 that's how we want to do it. 14 The institution came back and said no, no, 15 no. We're starting with IS funded. You know, people 16 that bring the most money in to the institution, 17 regardless of what it is. We were the only game in 18 town and so we won, but it was a very interesting 19 series of events. If they could have gone anywhere 20 else, at that point, I think they would have. 21 So the type of independent IRB I have is 22 for profit. Some are non-profit that are out there. 311 1 And we operate on a fee for service structure. Now, 2 the bad thing about that is the conflict of interest. 3 The good thing is our resources keep up with our work. 4 And it's kind of interesting, one of the things that 5 occasionally sponsors will come to us and say we want 6 to sign a contract with you for the review of this 7 protocol, but you've got to put a cap of say $30,000 8 on your fees. 9 And we say no, because that's going to 10 depend solely on how many sites you submit to us, how 11 many amendments you make the research and we charge 12 for those types of activities. We're not going to, 13 you know, cut and then just provide free services to 14 you if you keep giving us requests for service. So 15 that's a nice thing about having the type of fee for 16 service structure. 17 There is economy of scale once you get 18 large and I think that's important. It's not as much 19 of an economy of scale as people think and I want to-- 20 oh, before I go into that. I think once you get 21 large, along with that economy of scale comes certain 22 things that you can do. We can develop some pretty 312 1 deep infrastructure, record keeping systems, 2 computerized systems, staff. You can develop pretty 3 good expertise. For instance, the FDA has device 4 requirements, I'm pretty comfortable with now, you 5 know. Every few months I get a new one. 6 We can afford those high maintenance sites 7 that need -- you know, cost us far more on site visits 8 and compliance oversight and they pay us, and that's 9 not a problem. We can afford a Proactive Site Visit 10 Program, which I think is very valuable. And quite 11 frankly, any large client can walk away from us. It's 12 not really going to hurt us. We will wonder about it. 13 We'll wonder what we did wrong, but you know if Merck 14 or Johns Hopkins leaves, so they leave. It's not 15 going to put a big dent in what we are doing. 16 What I do want to stress, I said there 17 were economies of scale. There are not economies of 18 complexity. It doesn't get simpler the bigger you 19 get. And it is just incredibly complex to run a large 20 IRB like we're running. And I think the main reason 21 is you have to treat each study as it's own research 22 study separately, just like a local institution does. 313 1 So at initial review you've got to consider all the 2 same things that any IRB would have to consider, 3 including the FDA compliance history, the 4 investigator, State Medical Board history, emergency 5 care capabilities, conflicts of interest they may 6 have, stock ownership, etcetera, local vulnerable 7 populations. 8 You get a cancer study and they say we 9 want to enroll employees. Well, is this oncology out 10 in the community, an oncologist out in the community 11 where he has an employee who actually has this answer 12 or is this a large university in a small town, like 13 the University of Florida in Gainesville where half 14 the people in the town are employees or students and 15 that's why they marked it? You know, there is a lot 16 of difference between those two reports of we want to 17 enroll employees. 18 Local attitudes, state laws, you need 19 consent requirements, SMO requirements where the CRO 20 is involved. How they have handled HIPAA, embedded 21 authorization, separate authorization, all of that has 22 to be done on a study-by-study basis. At annual 314 1 review, you have to look at, for that particular 2 study, progress reports, the number of subjects 3 enrolled, subject complaints, adverse events, 4 correspondence, it all has to be reviewed, and so that 5 again study-by-study. 6 And then finally, I'll close up with 7 another concern that has come out a great deal over 8 the years is IRB shopping. And, you know, the 9 argument for a long time has been that if there 10 weren't independent IRBs, there wouldn't be IRB 11 shopping, because some investigator can go to four or 12 five IRBs if they don't like the answer, decide which 13 one to use once they finally get something approved 14 and go with that. 15 But, you know, the reality is it's a lot 16 broader than IRBs. This is a quote from a VP at 17 Genentech, and this was a Congressional Hearing, and 18 I use this slide all the time. I don't put his name 19 on it any more. But, you know, he basically just said 20 "One of the things that goes into our equation is how 21 difficult the IRB is." We don't use the University of 22 California. Steve left, he would have enjoyed that. 315 1 You know, you can shop whole systems. 2 And here is a question. Is this IRB 3 shopping? This happens all the time. Let's say in a 4 20 site protocol, three of the IRBs disapprove it and 5 the sponsor just says well, skip those three sites. 6 We can get enough subjects in the other 17. Is it IRB 7 shopping? It's the same question. It's the same 8 issue. So what can you do about IRB shopping? I 9 should also mention there can be legitimate reasons, 10 cost quality, etcetera. Be wary of any IRB that is 11 charging and charges really low. I mean, you do need 12 resources to do the work. And I see some pretty low 13 costs out there. 14 How do you address IRB shopping? FDA did 15 put in the Emergency Consent Provisions 50.24(e), 16 Waiver of Consent, I should say, that "Any one IRB 17 disapproval has to be reported to the others," and 18 that has been out recently with the Artificial Blood 19 Products Study that is going on. There are certain 20 Congressional Bills that require it. You can ask a 21 question on the IRB application forms. The ones that 22 shop are likely to be the ones that lie to you 316 1 anyways. And informal communication. So that's it. 2 Thank you. 3 (Applause) 4 CHAIR PRENTICE: Thank you, David. How 5 many staff do you have in your IRB? 6 MR. FORSTER: 255. 7 CHAIR PRENTICE: Our next presenter is Dr. 8 Lowell Schnipper. He is the Theodore W. and Evelyn 9 Berenson Professor of Medicine at Harvard Medical 10 School and also Chief of the Division of HEM-ONC at 11 the Beth Israel Deaconess Medical Center. Thank you 12 very much for agreeing to address SACHRP. We look 13 forward to your comments. 14 DR. SCHNIPPER: Thank you. Thank you very 15 much for inviting us down. I speak to you not really 16 as a member of the Harvard Medical School faculty, but 17 as Chair of the Task Force assembled by the American 18 Society of Clinical Oncology or ASCO. As most of you 19 may know, ASCO is the largest organization of academic 20 and clinical oncologists, be they surgical, medical, 21 pediatric or radiation oncologists in the U.S. with a 22 very large worldwide membership as well. 317 1 ASCO lives, breathes and dies for clinical 2 research as an agent of progress against the cancer 3 and in that context, about four or five years ago, the 4 organization assembled a task force to look at issues 5 pertaining to oversight of clinical research as it 6 relates to the cancer problem. We assembled a task 7 force of some 20 or 25 people composed of IRB Members, 8 a few were IRB Chairs at major medical centers. We 9 had a number of patient advocates, individuals who 10 themselves participated in clinical trials. We had 11 several representatives from industry, and we had a 12 number of members of the Panel who were active 13 clinical investigators. 14 This is a sampling of the adversity of the 15 individuals and from whence they came. The specific 16 charge was rather broad and as the task force began to 17 focus down, several areas of great concern popped 18 right out. One of them is the whole review process 19 for clinical trials, and the other turned out to be 20 conflicts of interest. And both of these issues 21 result in policy statements, one of which, the 22 oversight policy statement, is actually included among 318 1 the handouts today, if you would like to review that. 2 It is broader than just the IRB question. 3 In the process of developing our 4 perspectives on research oversight, we, of course, 5 reviewed relevant literature, but actually went out 6 into the field and I wouldn't say we did an exhaustive 7 survey, but we interviewed and visited with IRBs 8 across the country representing some major academic 9 medical centers, small community hospitals and an 10 independent IRB, the one from which you just heard. 11 We visited Olympia, Washington. 12 We then came together and specifically 13 focused on a few particularly important to us issues. 14 One had to do, of course, with IRBs. What we learned 15 and this again a relatively small sample size when you 16 consider there probably are several thousand 17 institutions participating in clinical trials across 18 the country, but the people we spoke with were 19 overwhelmed with work. They often felt under-funded. 20 When we weren't speaking to academic medical centers, 21 they indicated that they felt that there was quite a 22 variability in expertise in reviewing the types of 319 1 trials that we were most interested in, specifically 2 cancer prevention and treatment trials. 3 Funding was an issue at many of the 4 institutions. They often had difficulty identifying 5 members to sit on the committee and participate. Some 6 of them, the smaller ones, also felt that there were 7 problems implementing the trial itself and monitoring 8 that implementation. The policy statement we came out 9 with again touched a number of areas, but I 10 highlighted in white the few that really pertain to 11 the clinical trials review process. 12 Specifically, our goal was to assure the 13 safety of the participants in clinical trials period 14 and parallel with that to assure the integrity of the 15 research process to the extent possible. We, of 16 course, advocated a system that promotes 17 accountability and responsibility among all those 18 involved in the clinical trials process and mindful of 19 the fact that most medical centers, particularly 20 academic medical centers, small hospitals are strapped 21 for funds these days. 22 We felt that we needed to begin to think 320 1 about a system that would protect the public, protect 2 the integrity of science and also be cost effective. 3 Our policy recommendations, the most central of them 4 is, in fact, advocacy for the development of a central 5 review process for cancer clinical trials. We also 6 paid lots of attention to the issue of educating our 7 own membership about the ethics and integrity of the 8 clinical trial process, the informed consent process. 9 We paid some attention to the confusing 10 standards at times that exist from the federal 11 oversight agencies and, of course, we devoted a fair 12 amount of attention to conflict of interest, which I 13 will not address now. From the point of view of 14 centralization of review, and I might add that by no 15 means was our task force uniform in its sentiment 16 about this issue, in fact, we had many, many heated 17 discussions and still do to some extent. But the 18 policy statement endorsing this approach, in fact, 19 went to the board of directors at the organization, 20 and after two or three submissions and debates was 21 incorporated as you see it in that document. 22 But the appeal of centralized review is 321 1 the concentration of expertise, the possible promotion 2 of efficiency. I'll show you a little bit of data 3 that may speak to that in a while. We were not out to 4 endorse a particular model. We're aware that there 5 are a number of different models. In fact, some of us 6 thought a regional approach to this might well be more 7 wise than a single or central -- single central IRB, 8 from the point of view of practicality. 9 We certainly were enthusiastic about 10 cancer, at least, initiating such a process with the 11 Phase Three trials. These are comparative trials as 12 you know. They are often in the context of the 13 clinical oncologists treatment trials and because they 14 are done in multiple sites throughout the country 15 numbering in the many hundreds, the thousands, 16 centralized review mechanism helps to get the research 17 done more quickly. 18 What we are aware of right now is even in 19 those experiments that we tip our hat to because we 20 think that, indeed, centralization of the review 21 process has a future, we're finding that there is a 22 tremendous multiplicity of reviews and that is because 322 1 many institutions are not participating in it. And so 2 you have some that do, some that don't and in effect 3 the net effect is that there is almost less efficiency 4 from a time perspective in activating and enrolling 5 and thereby completing a trial than in the standard 6 old fashioned way. 7 We think then in the context of 8 centralized review, the local IRB has major roles, but 9 here it is a major role in overseeing the resources 10 and implementation of the study, the qualifications of 11 the investigator, for example. At the outset, if we 12 ever move towards this as a large scale experiment 13 throughout the nation, we certainly feel that only 14 those trials that have undergone rigorous scientific 15 review, such as those sponsored by the NCI system 16 should be admitted to a system such as this. 17 In particular, at a recent meeting with 18 officials from NIH, the idea emerged that, of course, 19 orphan disease, and it's so obvious, represents a 20 wonderful candidate population for the clinical trials 21 process in which a single review might thereby 22 activate a trial throughout the country. 323 1 Right now from our few visits with IRBs 2 across the U.S. there is tremendous variability in the 3 quality of the review. It's timeliness. The 4 centralized process would, in fact, eliminate that 5 variability to a large degree. We think that there 6 must be economies of scale available and I'm sure that 7 Mr. Foster could speak to it better than I, if one has 8 uniformity in the operational characteristics, such as 9 protocol forms, etcetera. A dedicated group of IRB 10 Members are very likely to become more and more 11 experienced with time and we think will, therefore, 12 provide the greatest level of expertise. 13 What about the cost? I don't know much 14 about this. I think we don't know much about this. 15 There are a few suggestions that are worth mentioning. 16 I may have lost a slide somewhere, but no problem. In 17 a brief letter in the annuals of internal medicine, 18 Humphreys, et al, provided an interesting comment. 19 This is a group of investigators that were sponsoring 20 a trial, a multi-institutional, nine institution trial 21 on drug abuse around the Palo Alto area around 22 Stanford. 324 1 The central or primary institution, if you 2 will, for which the trial emanated offered to be the 3 primary IRB for the entire group. That offer was not 4 accepted and each of the nine institutions then 5 reviewed the trial through their own IRB mechanism 6 using data that they have referenced in their paper 7 with respect to cost of evaluating a new submission 8 and adverse event this or that. They suggested that 9 an amount, almost 17 percent, of the budget of the 10 entire trial was consumed with the review process and 11 in their opinion, there were no additional reviews 12 that had an impact on the conduct of the trial. 13 Yes, there were modifications in the 14 consent form that met specific institutional mandates, 15 but they felt there were not substantive changes. A 16 much larger study, a little bit older, comes from the 17 Eastern Cooperative Oncology Group. This is one of 18 the NCI cooperative groups, as you know, that spans 19 hundreds of medical centers around the country. Bob 20 Catalano, in their administrative offices, 21 essentially, estimated in 1996 dollars, what the cost 22 percentage was for enrolling patients on a Phase Three 325 1 randomized trial. 2 He came up with a figure of $9,000 per 3 institution, itemizing specific action by specific 4 action required of the IRB and a total cost of about 5 $2 million for the review process. How real this is 6 in today's dollars, if we went through to a 7 centralized mechanism, how much of this would we save, 8 because for sure we would need to have local on-site 9 review of aspects of the trial's conduct in any case, 10 I don't know. But one senses that there is an 11 opportunity here to create greater economic 12 efficiency. 13 It's important to preserve the role of the 14 local IRB, in our opinion. How best to do that? 15 Really, I'm not sure, but I think what better site 16 than the local site is there to assure the 17 qualifications of the investigator leading the trial, 18 that the resources are there to conduct the trial 19 properly and, indeed, that the trial is being 20 conducted properly. 21 In fact, what we feel is that if this were 22 a concept truly embraced by federal oversight 326 1 authorities, they could provide clear guidance as to 2 what the role of a central review mechanism might be, 3 as well as what the local roles ought to be. There 4 are dissenting views. The academic medical centers 5 were surveyed, largely because of the noise centering 6 around this issue over the past few years. 7 In Academic Medicine July '04, was a 8 publication in which 70 percent of those institutions 9 surveyed were reported upon. The majority of them had 10 never used a central review mechanism, about a quarter 11 had. When asked why not, for those that had not, they 12 basically indicated that the quality of the review was 13 higher, they felt, than would otherwise be the case if 14 it was farmed out to a central review mechanism. 15 There were litigation concerns that were expressed. 16 And, of course, the issue of local concerns came up as 17 in we know our population, we know our scientists, 18 this is the best place to review the trial. 19 Interestingly, of those that did use a 20 central IRB, almost all were pleased by it. They did 21 note that the time it took for turnaround, for 22 basically submission and getting the protocol approved 327 1 and activated was much, much shorter. There have been 2 some proponents of the central IRB context who say 3 industry is going to look most favorably on those 4 institutions that, in fact, can turn around a trial, 5 open it up very quickly, get the accrual, close it, 6 get on to the next study. 7 The academic medical centers that did 8 respond felt that despite using the central IRB, they 9 weren't sure that it helped them attract additional 10 industry trials. I think that remains to be proven on 11 a larger scale than was looked at in that particular 12 study. And I hasten to add by talking about academic 13 medical centers, we're talking about the smallest 14 fraction of patients who are being put on clinical 15 trials. Probably the majority of industrial trials 16 are now being done outside of the academic medical 17 center and increasingly the clinical trials process on 18 these large Phase Three studies is moving away from 19 the academic medical complex. 20 So if we heard as we did earlier today and 21 we have over and again, OHRP and FDA tell us that it 22 is okay to use a central review mechanism. Why 328 1 doesn't it happen? And the legal concerns, obviously, 2 are always at the top of the list. A question to 3 which I don't really know the answer is how real is 4 this and if it's real, how can it be overcome? 5 What about local commitment to reviewing 6 trials and how important is that? Yes, there are 7 local constituent concerns in every institution for 8 sure. As a cancer physician, I can tell you that 9 cancer is an equal opportunity dreadful disease. And 10 I can't imagine that treatment trials that are deemed 11 reasonable by a panel of experts would find very much 12 that would be disputed upon from one local environment 13 to another, although that's absolutely possible. 14 Of course, how do we pay for this? Big 15 question. These are more questions than we have 16 answers to. We recognize that OHRP and FDA have 17 endorsed this approach. Our question is whether or 18 not this can be made even more emphatic. By having 19 the situation that we have now where institutions 20 choose whether they do or don't wish to use a central 21 mechanism leaves all of the other potential barriers 22 still in place not being addressed. 329 1 One wonders whether for certain classes of 2 trials there may even be the statement made that this 3 is a preferred way to go citing specific reasons. I'm 4 sure that is an arguable point. But I come back to 5 our notion that for the cancer community, for complex 6 diseases in general, multi-institution trials are 7 critical to getting useful answers. So the existing 8 approaches we have already heard about and I don't 9 want to be redundant. The NCI pilot study is accruing 10 institutions and is currently under way and will now 11 be expanded or extended to the Children's Oncology 12 Group, as Susan Weiner opened up this Panel. 13 Whether a central IRB that is single or 14 multiple regional IRBs would represent a more 15 practical approach, I'm not sure. How to integrate a 16 proprietary organization such as WIRB, who do their 17 work, I think, quite well, into this system remains an 18 open question. At ASCO, I think, we firmly believe 19 that while there are ambiguities and lots of 20 subtleties in this issue, if judiciously applied, this 21 type of approach to certain kinds of trials, certainly 22 Phase Three multi-institutional trials, trials 330 1 involving orphan diseases seems to be the way to go. 2 Thank you. 3 (Applause) 4 CHAIR PRENTICE: Thank you very much. 5 Would all of the Panelists, please, assemble at the 6 table now? Okay. I would like to begin by thanking 7 all the Panel Members for their presentations, and I 8 would like to make a few remarks before I ask the 9 first question. The landscape of clinical research in 10 this country has changed dramatically since the early 11 1980s and for those of us who have been in this 12 business since then, we have seen remarkable changes 13 accompanied by challenges. 14 As Dr. Schnipper mentioned, a great deal 15 of clinical research now is conducted out in the 16 private sector by private practitioners. The figures 17 are somewhere in the 70s, 75, 78 percent. The amount 18 of clinical research sponsored by pharmaceutical 19 companies compared to the private sector has decreased 20 at academic health science centers, which is alarming 21 to some academic health science centers. Independent 22 IRB reviews have increased. We're seeing more 331 1 independent IRBs and the volume of business that these 2 independent IRBs, like Western and like Chesapeake and 3 like Shulman, the volume of business has increased 4 dramatically, because they are fulfilling a need. 5 And all of the panelists pointed to one of 6 the problems relative to local IRB review, but 7 something that Bob Levine said really struck me, and 8 I'm looking at his conclusions slide where he said 9 "There is cause for concern about the well-being of a 10 local IRB" and we all recognize that. I certainly 11 recognize that in my institution. We suffer from all 12 of the problems that Bob mentioned, but something else 13 struck me even more. He said "But something of 14 immeasurable value will be lost," because the whole 15 concept of IRB review was really founded upon the 16 local IRB, but that may not be a viable model anymore. 17 So my first question goes to Bob, and the 18 reason is what you said and also the fact that you 19 have probably been thinking about the whole area of 20 research ethics far longer than any of us have in this 21 room. I don't want to disparage your age, because you 22 look pretty good, but I can remember going to many of 332 1 your workshops back in the old days, so I know you 2 have been thinking about all of these issues for a 3 very long time. 4 So this is my question to you. What do 5 you think the IRB or the HRPP system in this country 6 will look like in 10 years time at an academic health 7 science center? I want to make that the focus. And 8 what would you like to see it look like? It may be 9 the same thing, but I want to find out, you know, what 10 you think and then we can open it up for the rest of 11 the Panel to comment. 12 MR. LEVINE: Well, thanks very much. I 13 want to clarify my last comment on that last slide. 14 It might give the impression that I'm opposed to non- 15 institutional or central review boards and I'm not. 16 When I say something of immeasurable value will be 17 lost, what I mean is if you totally replace the local 18 IRB with external agencies, it's very valuable to an 19 academic environment to have people sitting in the 20 middle of it who are worried about research ethics, or 21 at least some of them, as their primary preoccupation 22 and who will go out into the community within and 333 1 without the institution and talk about research 2 ethics. 3 I think I will not retreat from my 4 statement that it's very important to get the message 5 out within the university community that these are the 6 values of the university. That we're -- too many IRBs 7 throw away their credibility by saying don't blame us, 8 this is all the feds' fault. You know, then if they 9 ever come up with the opportunity to make a decision 10 that's not explicitly covered in the regulations, they 11 have lost their credibility. They can't say well, 12 this one's on us, this one is our initiative. 13 What do I think it's going to look like in 14 the future? I think that we are going to see a 15 continuance or an ever increasing problem getting 16 people to serve on IRBs and on other committees that 17 consume a lot of time, admissions, curriculum. I 18 think the primary reason for this is the economic 19 squeeze on the academic medical centers and on the 20 universities generally, and the pressure on medical 21 school faculty to generate income. I think something 22 will be lost and I just hope we don't lose any more 334 1 than is necessary. 2 I'm very much in favor of turning over the 3 responsibility for primary review of multi-center 4 clinical trials to either regional groups or groups 5 like Western IRB or the consortia that Susan Kornetsky 6 mentioned, the two university initiatives. But you 7 know, I think a lot of things are going. We're here 8 to talk about IRBs today, but I also think that we're 9 losing something immeasurable as we turn more and more 10 of our oversight to managed care organizations, but 11 this is also beyond my reach. 12 We tell the students you have got to make 13 your decisions based upon an intimate knowledge of 14 individual patients, but they are not stupid. They 15 notice then that some case manager who lives 1,000 16 miles away will dictate to the doctor what treatment 17 will be acceptable for that patient. And so, you 18 know, they are usually not so impolite as to say 19 gotcha, but something is lost there, and I will just 20 quit at that point. Thank you. 21 CHAIR PRENTICE: Any of the other 22 Panelists care to comment on that particular question 335 1 or shall we move on? Okay. Before I turn it over to 2 the rest of the SACHRP Members, I would like all of 3 you to think about what sort of advice you would 4 provide SACHRP in terms of trying to resolve the 5 problem that has brought us all together today to talk 6 about, you know, this particular issue of various 7 models of IRB review, because we want to find out what 8 we should do as a Committee if we can do something. 9 That's the whole purpose of us being here. 10 Okay. Others? That's the question. Do 11 you want to start with that then? Okay. Let's just 12 open it up with that particular question and have each 13 of you provide us with some feedback. 14 MR. FORSTER: Could you repeat the 15 question? 16 CHAIR PRENTICE: Okay. The primary role 17 of SACHRP is to try to facilitate the resolution of 18 various problems that are brought to our attention. 19 This particular problem was not necessarily within our 20 charter, but that doesn't mean we can't go outside of 21 the charter given to us by the Secretary. We have 22 talked about the issue of central IRBs for quite 336 1 sometime and ASCO sent us a letter, as you know, 2 asking us to consider central IRB issues and this is 3 what we have done today. 4 Now, we want to know what is the next step 5 that we should undertake? What would be your advice 6 to us as to what should, could be done? Yes? 7 DR. SCHNIPPER: Since ASCO was, to some 8 extent, a stimulant for this discussion, I guess I 9 feel we ought to say something. If there is a way to 10 categorize some types of clinical trials, and that I 11 would submit remains a subject for some discussion, 12 but Phase Three treatment trials for cancer, for 13 example, is one that is obvious. I spoke of it during 14 my remarks. I would like to see the appropriate 15 oversight agencies endorse emphatically and indicate 16 that this is the preferred way to go. That's just a 17 suggestion, clearly arguable. 18 CHAIR PRENTICE: Bob? 19 MR. LEVINE: I think that I would like to 20 see some things done to address the excessive burdens 21 that there are on the IRB system, whether the IRB is 22 within an institution or outside. I touched on some 337 1 of these excessive burdens earlier, but let me mention 2 just two or three of them now. 3 I think, first, if and when it ever 4 becomes necessary to sanction a research institution, 5 great care ought to be directed to making it clear or 6 first, I think, the best thing to do is to try to 7 settle these things quietly, not with so much 8 visibility, but to tell people, you know, in certain 9 ways to shape up. But then if it comes to the point 10 where it's necessary to close down part or all of a 11 research operation, be very, very careful that the 12 public announcement focuses on what the real reason 13 was for this, rather than have a journalist copy a 14 list of 28 observations and nobody out there can 15 figure out well, you know, they see these things all 16 the time, what really got them to close this operation 17 down, because the effect that has had on the outside, 18 is that the upper echelons of university 19 administration, who used to be blissfully ignorant of 20 the IRB, all of a sudden got very interested and took 21 the most conservative postures. 22 They said well, look here, they closed. 338 1 One of the things they mention, that this place 2 weren't conducting their annual reapproval at convened 3 meetings. The first thing, evidence we had of a 4 response to that at Yale is they bought us two 5 shopping carts to bring all the protocols in for 6 annual reapproval to the convened meeting. The 7 adverse event reports ought to be turned over to the 8 Data Monitoring Committee, which actually knows what 9 the denominators are under all of these numbers. I 10 think also that I would -- well, that's enough, just 11 lighten the burden there. 12 The other thing is to see to it that the 13 other side of the story is told. When reports come 14 out of various governmental agencies that are 15 interpreted, perhaps interpreted extravagantly as 16 evidence, that there is something terribly wrong. One 17 Congressman stood up and he said we have evidence now 18 that the IRB system doesn't even know how many human 19 subjects there were. They know exactly how many 20 animals they have used in research. Why do they care 21 more about animals than about people? And I responded 22 well, first, no one ever told us that counting human 339 1 subjects would contribute to the protection of their 2 rights and welfare. 3 If you really want to see marvelous 4 documentation of numbers and demography of research 5 subjects, look at the records of the Nazi 6 concentration camps. Why do they know the numbers of 7 animals? Because they use NIH money to buy them. And 8 at the end of the year, the NIH says what did you 9 spend our money on? It's a different form of 10 recruitment than we have for human subjects. 11 And the next day a journalist who was 12 covering this episode quoted everything he said and 13 not one word that I said, and the newspaper story 14 looked like the whole system doesn't care about 15 anything, doesn't care about humans as much as they do 16 about animals. That's what I mean by making sure the 17 other side of the story is told. Thank you. 18 MR. FORSTER: I guess what I would advise 19 SACHRP to look at is one of the things, one of the 20 deep roots of causing this move toward central IRBs or 21 talk about it is funding and resources. I think that 22 is really the root of this issue and what perhaps 340 1 SACHRP should be looking at for human subjects 2 protection systems. 3 I think I would also warn about 4 centralization of power. There are centralized IRBs 5 in some European countries. They become very 6 bureaucratic, very slow and all powerful and that may 7 not be a better situation than we have now. Something 8 to think about. And I guess I would ask SACHRP to 9 look at how well whatever system they recommend 10 protects subjects and balance that against the 11 regulatory burden. 12 And finally, not to forget that there are 13 going to be investigator initiated studies and other 14 things that there's not going to be a lot of resources 15 available and how are we going to give those resources 16 and do you want to really think of a central IRB 17 system for some guy cooking up a protocol, you know, 18 in his lab on his own? That's it. 19 CHAIR PRENTICE: David? 20 DR. LEPAY: Well, again, I don't know. Is 21 this on? I'm not quite sure how to respond to the 22 question of what should SACHRP do here, because, of 341 1 course, SACHRP is advisory to our department, which 2 means to OHRP as well as to FDA. So providing our 3 advice as advice on how to advise us may be a bit 4 circular, but certainly what we are interested in 5 hearing in this process are any concerns that are 6 being voiced about either the process of central or 7 local review or the processes that we have in place 8 within our own agency, the kind of at least advice 9 that we have been providing in the past and are 10 expecting to provide in the future as we develop 11 additional guidance in this area as we have committed 12 to do so. I think that is perhaps as much as I can 13 say unless it's more a direct question. 14 CHAIR PRENTICE: Okay. 15 DR. CAROME: And I would just note that 16 because the provisions of 46.114 are part of Subpart 17 A, perhaps the subcommittee that was established this 18 morning can focus part of its attention on this 19 provision and how best to implement it and what could 20 be done in the way of guidance or change, if 21 appropriate. 22 CHAIR PRENTICE: Okay. Mark? 342 1 MR. BARNES: Bern and I were actually at 2 this ASCO Meeting. There was an ASCO Round Table 3 Meeting a number of months ago. It was, I guess, 4 three or four months ago, maybe even longer now. And 5 I think that, you know, one of the conclusions at the 6 meeting was what has been stated, which is that a way 7 to do this is to do it in a step-wise fashion and that 8 is to go for kind of the low hanging fruit first and 9 the low hanging fruit is clearly the trials like the 10 pediatric oncology trials and the orphan disease 11 trials and the other Phase Three oncology trials. 12 In terms of what SACHRP could do, I mean, 13 if, in fact, that's what's being done and that is kind 14 of the overall plan among those who are supportive of 15 central IRBs is to do it in a step-wise fashion, then 16 what needs to be done, I'm not sure whether SACHRP can 17 do it, I don't think we could, although we could be 18 advisory to OHRP and/or the FDA or other agencies in 19 doing it, is that there needs to be some evaluation 20 along the way of how the central IRB process has 21 worked, how it has worked to better protect subjects 22 or to lessen protection for subjects. 343 1 My intuition is that probably, especially 2 in cases like pediatric oncology and in orphan disease 3 trials, that it's going to offer better protection for 4 subjects and it's going to offer better expediting, 5 better oversight of the research and better expediting 6 of the research and, certainly, better analysis of the 7 meaning of adverse events given that there will be an 8 actual denominator instead of an odd kind of 9 hypothetical denominator. 10 So I guess where that leaves me in terms 11 of what we can do, Ernie, is unless somebody else has 12 a better idea, is it seems to me that what we could do 13 is we could try through -- I don't think we need a 14 subcommittee to do it. We could possibly do it just 15 among ourselves as a committee of the whole or 16 whatever, but it would be to try to list out for OHRP 17 and for the FDA and for any other agency, NIH, whoever 18 wants to listen, all the adjunct or the ex-officio 19 members, I should say, of those agencies that are 20 around the table is to list out what we think the 21 variables are that ought to be looked at, what are the 22 factors that ought to be looked at in an analysis of 344 1 how a central IRB process is working for particular 2 areas. 3 I think it's beyond what we could do even 4 through a sophisticated subcommittee process to do the 5 analysis ourselves, but we could perhaps at least set 6 forth some of the concerns that we would like to be 7 addressed in an ongoing analysis of how a central IRB 8 process might be worked and might be tweaked or 9 adjusted, so that it could better protect subjects. 10 CHAIR PRENTICE: Felix? 11 DR. GYI: You know, as I sit here 12 listening, I feel conflicted because of my role on 13 this Committee and my role outside of the Committee, 14 and I just want to put that on the table that there is 15 a perceived or real conflict of interest that I feel. 16 You know, I cut my teeth on the academic 17 Institutional Review Board. I restructured and ran an 18 IRB for a number of years for a large teaching 19 institution, and my thinking was forged and formulated 20 by the people that are around this room, the likes of 21 Lou Lazania and the gentleman that wrote the book that 22 is sitting on the Panel, Ernie Prentice and Charles 345 1 McKay and Charles McCarthy. I mean, those are people 2 who have helped us to understand where the regulations 3 came from and how we might want to have that local 4 attitude and local position be fixed in the IRB 5 system. 6 But we have come a long way. Times have 7 changed a great deal. I left academia, because I felt 8 that the constraints that were placed on the IRB 9 system within that institutional umbrella were 10 contrary to what the IRB should be doing, and I felt 11 that I could and the environment could support a 12 better model outside, and I think we have demonstrated 13 that through the likes of the good, independent IRBs 14 that are represented in the community. 15 In my consulting travels in helping other 16 IRBs and, especially, institutional IRBs address their 17 constraints, I think that those people who do 18 consulting like Ada Sue and some of the other folks 19 will tell you that the institutional IRBs are facing 20 the problems, because there is lack of institutional 21 will to support the IRB Office in the way that it 22 needs to be supported. David's office has over 250 346 1 people. Mine has close to 60 people, I mean, and all 2 we do are IRB work. That is what we do. That's what 3 our commitment is. 4 And the models are out there, so there are 5 folks that are doing this already. So one of the 6 questions that I have to ask ASCO and folks is what is 7 stopping us from simply utilizing existing pathways 8 already? Why are we dedicating time and what is it 9 that we're supposed to be doing now? What is it that 10 SACHRP should be looking at if there are existing 11 models and pathways to explore for efficient conduct 12 and review of research? 13 And I would simply echo at the last point 14 David's point about centralization of power. We have 15 seen from our European colleagues good models and bad 16 models, and one of the bad models is what he talked 17 about in addition to having a two tiered system that 18 can really paralyze the research that we're trying to 19 make more efficient. 20 CHAIR PRENTICE: Tom? No? Okay. Okay. 21 Susan? 22 DR. WEINER: Okay. So let me respond to 347 1 Felix for a minute and to the Panelists about what we 2 have discussed and what our thoughts are about what 3 SACHRP might do. I think that the consensus has been 4 that there is not a deep enough understanding of why 5 these barriers exist, what the costs are of doing it 6 one way or another, how actually to reduce some of the 7 burdens we have talked about, how to really discuss 8 local concerns or define local concerns, so that they 9 are meaningful locally, they contribute to the 10 protection of subjects in research and also allow a 11 more, for better or for worse, centralized review of 12 some of the important issues that need to be covered 13 as a way of keeping in mind both protection and 14 efficiency. 15 And one of the ideas that has come up is, 16 it actually came from Amy Patterson, one of the ex- 17 officio members, that NIH might be willing to sponsor 18 a workshop that might really take, you know, spend a 19 day or a day and a half asking people to do the 20 analyses of the issues that we have talked about, you 21 know, so that if there were to be guidance that was 22 recommended from SACHRP, that it would be grounded in 348 1 more data and more evidence than we currently have. 2 CHAIR PRENTICE: Yes, Nancy? 3 DR. JONES: I think the one point I would 4 like to make is the ethos of what an IRB was initially 5 asked to do based on the definition that Bob gave us 6 beyond just new compliance, beyond just making sure of 7 all the numbers, but that was the education back to 8 the researchers and to the institution and providing 9 a culture for the underlying principles of what it 10 means to do ethical human subjects research. 11 And I guess my question is, you know, I 12 don't have any problem with a lot of the multi-centers 13 going to a centralized review, you know, although I 14 think some institutions ultimately, it might be easier 15 to put all their review towards a central body. So if 16 that became the major model, how would we get those 17 other things back into the institutions, the ethos 18 back, or how could you maintain, that is the question. 19 And maybe it's not being maintained right 20 now with IRBs locally if they actually do that 21 educational component, which I found amazing that 22 those were the two things that were pointed out by 349 1 Lowell that you just omitted, you know, the education 2 back for the researchers and for the culture. 3 MR. LEVINE: Are you asking me? 4 DR. JONES: You first or -- 5 MR. LEVINE: Okay. I don't think we can 6 turn it all around right now. I mean, the 7 introduction, the IRB was brought into the world of 8 research in a very much different time than we have 9 now. I think at that time, quite a number of members 10 of the faculty were curious about the research and 11 about the ethics of research. Some of them were 12 irritated by the whole idea of informed consent, and 13 people seemed to have more time to talk with one 14 another. It was a time when, for example, an academic 15 department would see itself as a small community and 16 they would socialize together and so on. And now, my 17 cardiologist friends, who are muscle cardiologists, 18 say when they invite a speaker to a cardiology 19 conference that the electrophysiologists don't show 20 up, because it's not in their field. 21 We have that degree of fragmentation and 22 I think it's happening across the board. The only 350 1 thing I hope in my comments today is to take away the 2 impediments to getting this work done that I think we 3 have some control over and I mentioned a few of them, 4 but I don't think we're going to reconstitute the 5 ethos of the -- I mean, maybe you don't remember the 6 1960s, probably you don't, but it was a time when 7 people in general were very concerned about values and 8 carrying on about, well, you hear about some of 9 Senator Kennedy's activities during that time, which 10 was something the country was very much concerned 11 about and it seems to be a different setting right 12 now. 13 DR. WEINER: I have one response for you, 14 Nancy, which is something that I would like to 15 disabuse from now on, which is that we're not talking 16 about alternatives, a single set of alternatives that 17 has two options. We're talking about a range of 18 possibilities and that it's never going to be what's 19 less resolved, all kinds of trials or all kinds of 20 diseases or phases. And I think that what at least 21 people in the oncology community are looking for, and 22 I include myself in this statement, is more 351 1 flexibility that would yield more efficiency for 2 patients and for the investigators, you know, there 3 are those. 4 DR. SCHNIPPER: Just to answer you, as 5 well, I passed over the education piece largely 6 because it wasn't the focus of today's remarks, but if 7 there is anything that ASCO feels it's capable of 8 doing it's, in fact, educating its membership about 9 the proper conduct of clinical trials and clinical 10 research in general, and that particular theme 11 occupies a very, very large segment of almost every 12 national and regional meeting that we have. 13 I think that the question that Susan 14 raised of what are the barriers continues to confuse 15 us because of the apparent regulatory permission 16 that's already been out there and, yet, not 17 necessarily adopted by most people. You know, at the 18 risk of putting her on the spot, there is somebody in 19 the audience involved with the NCI Program which, in 20 fact, has over time been accruing institutional 21 adherence. 22 However, we hear from NCI and, Jaci, I 352 1 think, can comment specifically if she chooses, that 2 for any one institution that says yes, we're going to 3 buy into this central review, they might send one 4 trial through and, yet, other trials sponsored by the 5 same cooperative group mechanism are rendered, you 6 know, back to the institutional IRB not using the 7 central IRB, which has left me feeling somewhat 8 uncertain about what are the barriers to using a 9 perfectly usable mechanism? Is it financial? Is it 10 local issues or what is it? 11 Jaci, I mean, I hate to put you on the 12 spot, but I am. 13 MS. GOLDBERG: I think -- 14 PARTICIPANT: Come to the microphone. 15 Here's one. Here's one right here. 16 MS. GOLDBERG: Thank you. I think the 17 obstacles are varied and let me say at the outset 18 that, you know, from the NCI perspective, this project 19 is going well and central review is going well. The 20 local sites are utilizing it and we have not opened it 21 up. There is a potential poll in the NCI model of 22 possibly 2,000 sites that could eventually 353 1 participate, and we have been concentrating our 2 efforts on getting the systems to work, because I 3 remember David made a comment awhile ago about, you 4 know, central review and IRBs are very complex. So we 5 have been of late putting our energy into that and not 6 so much expansion. 7 But I think that one of the variables that 8 I bring up when I talk in other environments is IRB 9 culture, and so that has been one hurdle that we have 10 had to overcome. So while the investigators may be 11 interested, IRB, the turf issues and the kinds of 12 things and the fact the regulations are set up with, 13 you know, they have been in place for I don't know how 14 many years now and they are focused on -- you know, 15 people have just become in many ways set in their 16 ways. So on some other occasion or something, I could 17 give you more reflective comments about it, and if 18 somebody wants to ask me something more specifically, 19 I can address that, too. 20 I mean, you know, I don't know if numbers 21 really mean anything in this context, but we have got 22 at the moment 435 sites in this initiative. We have 354 1 got 184 IRBs representing 434 sites. We have got a 2 menu now of 59 or so approved protocols. They are 3 mostly -- I guess, at this point, they are all Phase 4 Three cooperative group trials. 5 And the mechanism we have set up, I maybe 6 should explain that for a second, is that the National 7 Board, the National IRB reviews the protocol before 8 it's distributed to these participating sites, and 9 they can do either a subcommittee or a single review, 10 but they don't need to do full board review, so the 11 process takes place very quickly, in a matter of days, 12 at the local site, and we're calling that facilitative 13 review and it has occurred now 1,100 times. So that 14 is just sort of a quick overview. 15 And I suppose, too, one of the obstacles 16 as a practical matter that hasn't come up as 17 frequently as it did in the beginning were liability 18 issues. I mean, what you were saying about some 19 strong endorsement on the part of the regulatory 20 agencies, but I know the regulatory agencies feel they 21 are sort of in a bind about that aspect of their job, 22 but if the word could get out more to more places that 355 1 this model is acceptable, I think that would help this 2 a lot. 3 And also in this project, we have a 4 mixture of comprehensive cancer centers, which are 5 major academic medical centers and local community 6 hospitals. So if there is anything else anybody wants 7 to know, I would be happy to answer it. 8 CHAIR PRENTICE: Thank you. 9 MS. GOLDBERG: Sorry for the sidetrack. 10 CHAIR PRENTICE: Okay. We belong to the 11 NCI IRB. My Co-Chair is the Chair of the Pediatric 12 New IRB, so we have thought a lot about this 13 particular model and how it works or how it doesn't 14 work. And in considering all of the issues we have 15 talked about today and why there is a reluctance to 16 accept central IRB review, I think one point has 17 already been made and I would like to emphasize that, 18 and that is that OHRP and FDA have not issued any 19 clear guidance that says that they accept central IRB. 20 I mean, it's evident if you read the regulations, but 21 I'm unaware of any particular guidance that says 22 specifically yes, this is fine. Okay. So some IRBs 356 1 may not be totally aware of that as they perhaps 2 should be. 3 But a second point I want to make is 4 something related to what Mike Carome said and I think 5 David said the same thing. The regulations require a 6 joint oversight. You can't just simply, you know, 7 farm out your responsibility to a central IRB. You 8 have to assume some responsibility at the local level, 9 and I don't think a lot of IRBs understand exactly 10 what that does entail. What does that mean? 11 Does it mean that you enter a protocol 12 into a database, so now you know that the protocol is 13 ongoing? Does it mean that you simply make sure that 14 the consent form conforms to your template and 15 includes some of the standard language? Is that what 16 is means? Does it mean that if there is an adverse 17 event that occurs at your institution, that the IRB 18 has to review that? There is no guidance on that. 19 Does it mean that you have some responsibilities in 20 terms of monitoring on-site, some responsibilities for 21 training, some responsibilities for continuing review? 22 You know, we have thought a lot about all 357 1 of these and we have what we think are reasonable 2 mechanisms in place to accommodate what we view to be 3 our oversight responsibilities, but I am unaware of 4 any guidance that OHRP or FDA have issued in that 5 regard and correct me if I'm wrong, but is that not 6 right? Mike, David? I don't think that -- 7 DR. CAROME: That's correct. 8 CHAIR PRENTICE: Okay. 9 DR. CAROME: We have not provided a formal 10 guidance document on central IRBs or cooperative 11 review. 12 CHAIR PRENTICE: Okay. And it may be that 13 if such guidance were to be forthcoming, you would 14 find more institutions perhaps receptive to the idea 15 of going with a central IRB. You know, if you look at 16 the advantages of central IRBs relative to multi-site 17 trials, you can't change the design of the protocol, 18 because it's a multi-center trial. You can ask 19 questions, but you can't change the design. About all 20 you can do is review the consent document and that's 21 about it. So I think that institutions really need to 22 avail themselves of, you know, these kinds of 358 1 mechanisms, particularly, you know, for co-op 2 sponsored multi-center clinical trials. Mark, do you 3 want to say -- 4 DR. CAROME: I would note that for the NCI 5 Project, we did issue a letter that I think I shared 6 with all of our sites, which outlined how you do this 7 and what your responsibilities are, and the lack of 8 participation isn't because they don't understand they 9 can do it and how they should be doing it. It's many 10 of the other reasons that Dave Forster went over and 11 some of the things have shown up in the AAMC Survey 12 not because they don't know about it or because they 13 think the regulations don't permit it. 14 CHAIR PRENTICE: Bob? 15 MR. LEVINE: Yes. When I said earlier 16 that OHRP could contribute by seeing to it that the 17 other side of the stories are told, for example, 18 several of our panel said that the IRBs don't want to 19 let central IRBs do this, because they are afraid of 20 litigation, liability. You might say name a case. 21 Name a case in which an IRB was found liable for 22 damages or, for that matter, an investigator when 359 1 working according to an IRB approved protocol. 2 The things that hit the newspapers are 3 dazzling, zillions of dollars, the Hutchinson case. 4 We all read about that and all the awful things they 5 did. Did anyone see a story on how that case came 6 out? No. The jury decided that, found against the 7 plaintiffs on every point, except the one that the 8 Hutchinson Clinic conceded and that was when they 9 broke the tube, the bone barrow tube broke in the 10 centrifuge. They said all right, we screwed up there. 11 Everyone read about this awful story about 12 the UCLA schizophrenia placebo study. Anyone ever 13 read a story, I mean, banner headline? Anyone read a 14 story on how it came out? No. They settled for 15 $199,000. The plaintiffs got nothing. The University 16 of California figured it would cost them that much to 17 litigate the case and win and the estimate was that 18 the plaintiff's attorneys sustained a net loss of $2 19 million. Where are these stories? How come the IRBs, 20 the universities, the IRBs are not hearing these 21 stories? 22 CHAIR PRENTICE: Mark? 360 1 MR. BARNES: Just a thought. You know, 2 there is something short of OHRP or the FDA issuing an 3 official guidance document about this, is that -- that 4 can have some impact is actually the input of this 5 particular Committee, and an example is in regard to 6 third parties in research and the follow-up from the 7 Virginia Commonwealth case. When the predecessor of 8 this Committee, NHRPAC, was empaneled, then we did a 9 statement. 10 We did a lot of group discussion in public 11 about what should be the treatment by IRBs of third 12 parties, what is the appropriate treatment by IRBs of 13 third parties in research, and we issued that. And 14 actually, that document, although not an official 15 document of OHRP, has been looked at and has provided 16 some guidance, I understand, to many IRBs around the 17 country. 18 So there is one thing that we could do or 19 a couple of things that we could do as a Committee, I 20 guess. One is that I think that if there is an actual 21 workshop or something that's done by the NIH, I guess 22 I would recommend that we send, you know, a couple of 361 1 our members from SACHRP to that to either participate 2 or at least to listen to the proceedings and to try to 3 gather some more evidence. But it also may be that 4 what we should be headed toward is some kind of 5 Committee statement about the acceptability and some 6 of the appropriate considerations in an institution 7 utilizing a central IRB establishing just as a 8 Committee statement, not as an OHRP statement, that it 9 is acceptable to do this, that there is no regulatory 10 bar that has been called to our attention that would 11 forbid it. 12 In fact, there is public policy in favor 13 of it, that these are some of the considerations and 14 that sort of thing. That might be a step-wise fashion 15 before kind of step-wise act that might be appropriate 16 prior to our sort of formally asking OHRP or FDA to 17 issue a statement that may, in fact, in the end not 18 even be necessary. And what we want to do is break 19 the log jam and that is what ASCO is calling to our 20 attention. There are ways of doing it by exercising, 21 you know, moral suasion rather than the issuance of a 22 regulatory document. 362 1 PARTICIPANT: On a public platform that 2 this offers. 3 DR. WEINER: I think that is an excellent 4 idea and I think that, you know, we have what amounts 5 to multi-step action steps that can be taken, you 6 know, to at least clarify the issues, to gather some 7 data, to issue some papers perhaps and then to see 8 whether the Committee would make, at that point, a 9 public statement about it. 10 CHAIR PRENTICE: Okay. I take it that 11 there is not a proposal yet, but certainly a 12 suggestion of one, that under the auspices of NIH 13 and/or OHRP, that some kind of a workshop be convened 14 and the objective of the workshop would be to, once 15 again, discuss some of the issues that we have already 16 talked about here, but ultimately the objective would 17 be to provide some kind of a series of recommendations 18 or a guidance document that SACHRP could perhaps use 19 to issue a position paper, so to speak. Is that where 20 you're going on this, Mark? 21 MR. BARNES: Yes, that is where I'm going. 22 That is where I'm going, yes. 363 1 CHAIR PRENTICE: Okay. Any other comments 2 about that? Well, not yet. I haven't asked for a 3 motion yet. 4 DR. GYI: Mark will give you that motion 5 when you ask for it. 6 PARTICIPANT: It will take him 10 minutes. 7 CHAIR PRENTICE: Okay. 8 DR. GYI: I can live with that process, 9 certainly, but I wonder if we are opening ourselves to 10 some criticism that says, you know, there are existing 11 models now. ASCO has some processes in place to do 12 the training and has an infrastructure to reach out to 13 its investigators, that WIRB is already out there, so 14 how come WIRB is not being used or how come NCI's CIRB 15 is not being utilized more? Why now, why ASCO when, 16 you know, there are other groups that may say from a 17 commercial side how come you didn't endorse our plea 18 for having some efficiencies in conducting research? 19 I'm not disagreeing with this process. 20 I'm just throwing out a cautionary statement that, you 21 know, why now? 22 DR. SCHNIPPER: Well, just to clarify. I 364 1 think ASCO is not interested in specifically one 2 mechanism or another as opposed to the generic concept 3 that for these multi-institutional cross transnational 4 trials, we think that centralized review makes the 5 most sense, because it will enable opening of the 6 trial to occur as promptly as humanly possible, 7 accrual to be completed and the answer to be known. 8 So it's really the efficiency of the process that we 9 are interested in and we're not wedded to any one 10 model. I think that may be beyond our, you know, 11 wisdom. 12 The other question I would ask, if SACHRP 13 and NIH are willing to consider this further, is 14 whether and as part of a subsequent discussion we 15 might well try to get as much data as we can about the 16 economic implications of either business as usual or 17 operating under a different context. 18 MR. BARNES: Ernie, can I say one thing 19 just in regard to what Felix said and Dr. Schnipper? 20 I think what I'm talking about, and I believe what 21 Susan is also talking about, is not at all to the 22 exclusion of the idea of the central IRBs as one of a 365 1 panoply of mechanisms to achieve central review. So 2 I would think that our analysis, our discussions, our 3 consideration and, ultimately, if we decide to do it, 4 our issuance of a Committee statement or a Committee 5 policy paper would not be to the exclusion of that. 6 That would be among the items and the possibilities 7 that would be discussed and contemplated. So I will 8 make it as a motion if you want me to. 9 CHAIR PRENTICE: Yes, it's time for a 10 motion. 11 MR. BARNES: Okay. My motion is this. My 12 motion is that we as a Committee, number one, 13 recommend to NIH, OHRP and the FDA that they consider 14 holding some sort of coordinated workshop in regard to 15 the use of central review mechanisms for research, for 16 human subjects research, number one. Number two, that 17 representatives of SACHRP attend such a meeting if and 18 when it is held and, number three, that we be 19 gathering information through that process and through 20 other processes, as well, in order to try to move 21 toward the idea of adopting in the future some kind of 22 Committee statement, some kind of consensus statement 366 1 from the Committee in regard to the prudence, the 2 wisdom, the availability, the considerations for 3 centralized review mechanisms. How's that? 4 DR. WEINER: I'll second. 5 CHAIR PRENTICE: Okay. Any further 6 discussion? All right. We have not established any 7 time line. We're saying consider. It will be 8 considered and you're saying considered by OHRP, FDA, 9 NIH, you know, some of the stakeholders in this. All 10 right. And I just want to make sure I understand 11 where we're going on this. 12 MR. BARNES: Okay. 13 CHAIR PRENTICE: Okay. And then SACHRP 14 Members, a number of them, would attend and ultimately 15 something would come out of this where we would look 16 at this and then decide what the next step would be, 17 right? It might be a position paper from SACHRP, it 18 might not be, right? 19 MR. BARNES: It might, it might not be, 20 but I think that the conceptual kind of gist of the 21 motion is that it would be a good idea for us to try 22 to endeavor to move towards some kind of Committee 367 1 position paper on this issue. 2 CHAIR PRENTICE: Okay. Pardon? All 3 right. Ready for a vote? Tom? 4 MR. ADAMS: I think I am actually more 5 comfortable with what Ernie articulated just from the 6 standpoint that we're not knowledge-based at this 7 point. We have heard some good, strong opinions from 8 other organizations today, but the Committee hasn't 9 looked at it. So I think to prejudge where we might 10 want to go would be premature, so I would move to 11 amend your motion to delete your third item and add in 12 the comments that Ernie made. 13 CHAIR PRENTICE: What did I say? 14 MR. ADAMS: I thought you would ask that. 15 Basically, you said that then, at some point, it would 16 come back to the Committee and the Committee would 17 take a look at it and make a determination. 18 CHAIR PRENTICE: Yes. 19 MR. ADAMS: As to whether or not we wanted 20 to do a paper or whatever. 21 CHAIR PRENTICE: Yes. Mark, acceptable? 22 MR. BARNES: That's fine. 368 1 CHAIR PRENTICE: That's fine. Okay. All 2 right. Are we ready now? Okay. All those in favor? 3 ALL: Aye. 4 CHAIR PRENTICE: Any opposed? Any 5 abstentions? The motion carries. Thanks for all your 6 hard work. The meeting is adjourned. Have a safe 7 trip home. 8 (Whereupon, the meeting was concluded at 9 5:02 p.m.) 10 11 12 13 14 15 16 17 18 19 20 21 22