MS. DIANE REHMThanks for joining us. I'm Diane Rehm. A new government report found that universities failed to tell parents that an oxygen study of their premature babies could cause blindness or even death. Joining me in the studio to talk about the study and how best to balance risks and benefits of clinical trials, Dr. Michael Carome of Public Citizen and Dr. Kevin Donovan of the Georgetown University Medical Center.

MS. DIANE REHMJoining us by phone is Dr. Edward Bell of the University of Iowa, one of the 23 universities that participated in this study. If you'd like to join us, call us on 800-433-8850, send us an email to drshow@wamu.org, follow us on Facebook or send us a tweet. Good morning to all of you.

DR. MICHAEL CAROMEGood morning, Diane.

DR. KEVIN DONOVANGood morning, Diane.

DR. EDWARD BELLGood morning.

REHMDr. Bell, I understand you're in the midst of a hailstorm.

BELLThat's right. So you may hear some thunder in the background.

REHMOkay. Well, we'll certainly tolerate that. Dr. Carome, let me start with you and ask you to tell us about this study, what it was intended to do and who participated?

CAROMEThe study goes by the name Support Study and it is an experiment that was funded NIH, conducted by 23 institutions that make up what's called the Neonatal Research Network, which are major, prestigious research institutions across the country. and it involved an experiment with extremely premature infants, in which the infants were randomized to get either or a high or a low oxygen and both of those segments or bands of oxygen level were within what infants get who are premature but what they did experimentally was try to, they ran only assigned half the babies to the low end of that range and half the babies to the high end.

REHMAnd the parents were not informed of which?

CAROMESo the parents were informed that this was a study comparing these two oxygen targets for the blood levels in the babies but according to a letter issued by the office of Human Research Protections, the parents were not informed about several important things, most importantly they weren't informed about the serious substantial risk of this research which included the possibility of blindness, the possibility of brain injury and the possibility of death depending upon which group the babies were assigned.

CAROMEThe consent form that the parents signed did offer the possibility of benefit, that the babies might not experience blindness and a type of eye disease that's associated with premature babies. But it didn't say the opposite, they could get, had a higher chance of getting that eye disease depending upon which group they were in.

CAROMEIn addition, when you read the quotes from the consent forms that were approved by the ethics committee at the lead site, it's clear that other things weren't disclosed to these parents. The true purpose of the research wasn't actually explained.

CAROMEFor example, one purpose was defined was to find out if babies were more likely to die if they were put in the low oxygen group or the high oxygen group. That was not disclosed. Also the true nature of the research describing what standard of care truly would be versus how this experiment would change the oxygen treatment was not explained to the parents.

REHMSo what were the outcomes?

CAROMEWell, we had two experimental groups and in one group, the low oxygen group, there was a higher statistically significant death rate. It was 20 percent in the low oxygen versus 16 percent in the high oxygen group. And then for the severe eye disease that they were measuring, it occurred in nine percent in the low oxygen group and 18 percent, so twice the level, in the high oxygen group.

CAROMESo basically opposite results for death and eye disease between the two groups. With respect to the eye disease there was substantial research previously known that this would likely be the result and there was certainly reason to suspect that the mortality data was predictable as well.

REHMSo now turning to you, Dr. Donovan, how typical was this research?

DONOVANWell, I think when we talk about this research we need to distinguish it from the other types of research that occur and this particular study had other things that led up to it and these are different types of studies. The first thing that sometimes happens is there is basically no research at all.

DONOVANIt's kind of the treat and hope approach where we hope that what we're doing for the patients is good. That actually back in the '50s is how oxygen started being used in premature infants in the first place.

REHMI see, I see.

DONOVANThey assumed that, you know, these children were having trouble oxygenating because of their immature lungs, let's give them oxygen. And of course, more is always better.

REHMAnd did the experiments that began in the '50s yield good results?

DONOVANWell, the interesting thing about it was initially they weren't experiments. They just treated and hoped. Then they started realizing that the treatment itself can have some serious problems. This is where the eye problems associated with oxygen in premature babies were discovered.

REHMSo you're saying that that was a known and a given before this experimentation began?

DONOVANBefore the Support Study, as you were hearing from Dr. Carome, many of these things had been recognized. Now, some other studies that address the same issue were different designs. They had retrospective studies, which you could say is, you know, was what we did, did it turn out good, where you don't actually study it ahead of time or while you're doing it but just go back and look and say well this is what we were doing, we thought it would be good. How'd that turn out?

DONOVANThat's useful information, that's how some of the information about eye damage and mortality and oxygen toxicity was developed. Then another type of study is an observational study, saying, you know, is what we were doing looking good? The Support Study was different from all of these okay and it actually followed what researchers would call the gold standard.

DONOVANBecause the gold standard for research studies is referred to as a double blind randomized study. Now, unfortunately the term double blind shouldn't be misinterpreted in a study where there's eye damage. We're talking about neither the people who are being treated nor the ones who doing the treatments involved in this research study know exactly where you are and that's ideal.

DONOVANThe other problem these previous studies had is they had conflicting results and one of the reasons they had conflicting results is not only they were purely observational or retrospective but almost always these were very small numbers.

REHMAll right. and now I want to turn to Dr. Edward Bell, since your University of Iowa was one of the 23 institutions participating, how was the coordination of that study to be conducted, what kind of communication were the individual institutions having with each other?

BELLHi, Diane. First I want to say what a pleasure and a honor it is for me to be on your show, I've been a fan for years.

REHMWell, thank you so much.

BELLIt's nice also to meet the other guests at least by phone. The Support trial as the previous speaker said is, was a randomized clinical trial which is the most powerful and most rigorous form of scientific investigation. And it was really the first such study to be done that addressed the particular issue of how do we know how much oxygen we should be giving to very premature babies for the very reasons that Dr. Carome outlined.

BELLAnd the major point of contention is that we didn't inform the parents adequately about the risks of the study and we the investigators happen to believe that we did a pretty good job of that. The OHRP or the office of Human Research Protection had objection to deficiencies in the place where the information was presented in the consent forms.

BELLI tried to explain that the Neonatal Research Network is a network of university academic units that take care of very premature babies, that cooperate to do clinical research that requires more subjects than can be enrolled in a single center.

REHMAnd I gather you had as many as 1,000 subjects in this study, is that correct?

BELLYes, there were 1,316 subjects which was the number that we calculated in advance would be necessary to examine the main question of the study, which was which of these two target ranges of oxygen would provide the lower risk of the primary outcome which was severe eye problems or death and I have to include...

REHMAnd Dr. Bell, I want to interrupt you for just a moment to apologize to our listeners for the quality of sound of your phone out there in Iowa. They are experiencing a hailstorm and I fear that that is getting in the way of the quality of the sound. We will do our very best throughout the hour to improve that quality but I hope you'll bear with us because Dr. Bell is an extremely voice to be heard this morning.

REHMWe're going to take just a short break here and when we come back, we'll talk further with all our guests and one more.

REHMAnd welcome back. I am hoping that joining us shortly is Dr. Arthur Caplan. he's director of the division of medical ethics at NYU Langone Medical Center. Good morning, Dr. Caplan.

DR. ARTHUR CAPLANHi, Diane. How are you?

REHMI'm fine, thank you. Always good to have you with us. I want our listeners to know, we have an email from an individual who identified herself as the reporter who first wrote about the support study, Theresa Defino, and gave a letter to Public Citizen. She wanted to know why there isn't a government speaker on this program. Unfortunately, this entire hour came about starting at 8:30 this morning when we learned that our scheduled second-hour guest could not be with us. And this program was put together at the last minute.

REHMWe did reach out to NIH. We were told that on such short notice the subject matter expert was not available. Now, to you Dr. Caplan, why do you say that the frightening headlines about this study were not justified?

CAPLANI say that, Diane, because we are in a very difficult area of research. We're not studying a new treatment for babies. We're not studying a new drug or a new device. That sort of study is certainly something that doctors do and should do. And it carries very great risks because you're not sure what the new innovation, the new drug or device or vaccine is really going to do. The study we're talking about now, the support study is trying to evaluate a existing set of treatments. In other words, doctors are trying to treat babies with underdeveloped lungs. They need oxygen. That is known to have side effects.

CAPLANThey try to do it but there's differences in practice. So you're not studying a new thing. You're trying to figure out among the things that doctors do, a range of standard treatments, what really is the best. And that is a little less risky than just testing a new thing for the first time in babies.

REHMOn the other hand, isn't it fair to say that parents participating in this study of their newborns should have been fully, fully, fully apprised of the dangers of either a lesser amount or a greater amount of oxygen intake for that child?

CAPLANYes, absolutely. And the government position has been, you still -- even though you're studying treatments, the treatments do carry risk. We're not sure which of the accepted approaches is best so the parents need to hear that. And let me add, we're focusing a lot on the consent form because that's the document we have. But crucial to this is the communication, the conversation that the doctor has with these parents.

REHMExactly.

CAROMEBecause, you know Diane, they've got a preemie baby. They're often overwhelmed by that.

REHMOf course.

CAROMEAnd to be honest, I'm not sure how carefully they'll read a somewhat complicated consent form. So the conversation is crucial. That's really what you want to see. And for this type of thing it would be useful in my view if we could ask them some questions, document their answers almost like a quiz, just to show that they really understood.

REHMExactly. All right. And I want to bring Dr. Bell back in on this conversation. Dr. Bell, how much oral communication was there with parents, not just signatures on forms?

BELLThis is a big part of the process and really the most important part. We believe that the key elements of (unintelligible) in the forms. In some cases they weren't located specifically in the risk section because we had said upfront that the purpose of the study was to look at how these different oxygen targets affected the risk of retinopathy in prematurity. We feel quite comfortable that the process was adequate and the parents were well informed.

BELLIf, in fact, OHRP would like us to explain as risks the possible consequences of using two different ranges within the normal standard of practice, that's going to change the way all consent forms are done for clinical research. It has widespread implications. And perhaps that's -- to do it but it's not the way it's traditionally been done.

REHMAll right. Dr. Carome.

CAROMEWell, I must vigorously disagree with some of the points Dr. Caplan made. I agree that a consent is more than a form but it's critically important that that form include key elements about risks and benefits and nature of the research. There is reason for alarm for the way this study was conducted with inadequate consent. We think the lack of adequate consent made the study unethical. And it's really misleading to say that for all these babies, getting oxygen anywhere between 85 and 95 percent represented the standard of care in that trying to test one versus the other did impose risk to these infants.

CAROMEWe know from the results of the study that you could change mortality, you could change eye outcomes by manipulating the oxygen. So they found differences between the two key endpoints that they were seeking -- you know, that they were trying to find out if there were going to be differences. And if you just look at, for instance, how the risk was described in the consent form, it basically implied there was no risk. Let me read from the possible risk section that OHRP quoted.

CAROME"There is no known risk to your baby from monitoring with the pulse oximeter, the device that's used to measure blood oxygen levels, used for this study. The possible risk of skin breakdown at the site will be minimized by your baby's nurse moving the oximeter, the probe that measures the oxygen, to another arm or leg a couple of times a day." I mean, first of all, these -- it's ironic that that was a disclosed risk because these babies have pulse oximeters whether they're in research or not. And so that wasn't a risk at all.

CAROMEThe risk that should've been described -- and I'll ask Dr. Bell if it was -- did it say that the risks of being in this research include the possibility of death, brain injury and blindness based upon the group assignment they went to?

REHMDr. Bell.

BELLYes. Well, those risks were present from the very outset of these babies' births as very premature babies. And the issue is whether participation in the study increased those risks. In fact, there was no increased risk of death. And the mortality in both groups, as you know, was lower than what we had seen previously in the network space and also what was seen in the babies who were eligible for this study but whose parents declined.

BELLNow I know you're aware that there was some differences in the risk factors between those groups. And we're in the process of sorting that out now. But we feel rather strongly that the babies were not placed at increased risk of eye problems or death simply because of their participation in the study.

REHMAll right. Dr. Carome.

CAROMESo I believe Dr. Bell is referring to a study published by the same group in which they looked at clinical characteristics of the 1300 babies who were in the support study with 3,000 other babies at the same locations who were eligible, who would've qualified for the study, but for various reasons weren't involved. And for those babies who were in the study, which he's now trying to represent perhaps...

REHMWho were not in the study.

CAROME...who were not in the study who he's trying to represent now perhaps as a control group. There were significant differences in the baseline characteristics of those babies. They clearly were a sicker population and different from the 1300 in the study. And indeed they had a higher death rate but that's completely to be expected. And to suggest -- to make these comparisons is simply scientifically in valid. You can't compare that different group to the babies in the research.

CAROMEAnd the purpose of the article, the Scientific Journal article that describes these results was to make the point they were different. And therefore we can't generalize our support study results to all babies who are premature.

REHMDr. Caplan, considering the outcomes and considering the questions raised about the outcomes of this study, would you believe that the approach to parents needs to be different, more specific, outlining more carefully the pluses, the minuses, the cost and the risk?

CAPLANI think in studying existing range of treatments where there are lots of variations, whether it's preemie babies or going to the dentist or how someone treats a heart attack, the practices that are out there in our hospitals and on the part of doctors still vary even though they're within a range of accepted practice. We have to figure out within that range what's best. Parents need to understand that type of study. They need to really understand that what's being proposed is to refine and determine within existing treatments what's best.

CAPLANI don't think -- and I'll stand by my statement -- that that is as risky or as dangerous as trying a new thing for the first time. But it does, because of the nature of the assignment of people to one treatment or another potentially shift the risk ratio. But parents need to understand it. And the way we're going to get that done -- I'm going to say this again, particularly for parents of preemies, is good conversation and quizzing them. I know lawyers and others love those forms. I know bureaucrats love those informed consent forms. I've been around long enough to know that a parent of a preemie baby is going to listen to what the doctor says, not so much what's written on the paper.

REHMAll right. And Dr. Bell, to what extent have the parents of the infants who were part of that study been, now that the study is completed, have they been told that their child was at risk?

BELLBefore I answer that, Diane, I'd like to just address one point that Dr. Carome made that I think is very important. And he said the fact that the study found a difference in mortality proves that we should have informed the parents of that risk. As a matter of fact, there was no reason for us to suspect that there would be a difference in mortality, and that was a surprising finding. We felt that, of course, it was very important to publish that result, and it has in fact changed practice for the benefit of very many babies

BELLWith regard to your question about...

REHM...informing...

BELL...whether the parents have been informed subsequently, we are quite worried that this debate is alarming parents and raising great concerns that their children may have been harmed by participation in this study. And most of the participating centers are considering the best way to communicate to parents to make sure that they understand that our motives were only to help babies. That's the reason that we're here and that's the focus of our job and the research that we do. And to offer to try to answer any questions that they might have.

REHMDoctor.

BELLWe're also interested in improving the consent process. And we are looking for advice from families as well as from government agencies about the best way to do that.

REHMDr. Edward Bell. He's a neonatologist at the University of Iowa, one of the 23 institutions that participated in the study. And you're listening to "The Diane Rehm Show." I'm going to open the phones. I want to hear what our listeners have to say. First let's go to Veronica. She's in Houston, Texas. Hello, Veronica. You're on the air.

VERONICAThank you for taking my call, Diane.

REHMSure.

VERONICALongtime fan, first time caller.

REHMI'm glad to have you with us.

VERONICAI'm a doctorally-prepared registered nurse. I have done research myself and I have sat on a number of institutional review boards. And my biggest question is, where was the IRB? This is terrifying that you have people on an institutional review board that are not asking the right kinds of questions and providing specific oversight and supervision of these kinds of pieces of research.

REHMDr. Kevin Donovan.

DONOVANI think that's a very reasonable question. I think there's some points that need to be made about this study and the response to it. I think perhaps the first point that shouldn't be lost is the study was worth doing. We didn't know the outcome in advance. There were reasons to suspect some of these outcomes but some of them were surprising, including the fact that the data suggested there was an additional death for every two cases of severe retinopathy that was prevented. That's going to change the practice in the treatment of premature babies.

DONOVANHowever, as not only a bioethicist but a pediatrician and a former IRB chairman, I will tell you that studies in children need to be approached differently. The protection for children is built in to be much greater, and it should be. The children themselves are not doing the consenting. Their parents are. And involvement in the study should have some prospect of benefit.

DONOVANI would also disagree with my colleague Dr. Art Caplan about the consent form just being a bureaucratic piece of paper. We were very particular about consent forms when we ran the IRB simply because this is what people have to take home with them. They can listen to the doctors but they may or may not remember what they've heard. This is the doctor's only documentation of what was said and what was done.

DONOVANAnd I think in this case that the criticism is well founded, that the consent form didn't really make it clear what exactly was being studied and to the extent that there might be additional risks because of the study. I think that the worst thing that could happen from this is to say, you know, the alarmist's view that this was a huge breach and great problems for the involved babies. And as a result we see less involvement, fewer studies done because then we're working blind and we don't know how to help kids in the future.

REHMSo you're saying that you believe that the studies have been necessary, but you're also suggesting that parents have not been sufficiently informed of what those risks actually are.

DONOVANI think that really is the issue at hand. And I think that Public Citizen, as Dr. Carome has said, and the OHRP have both made this point, I'm afraid the point's valid.

REHMAll right. And we'll take a short break here. Dr. Caplan, I do hope you can stay with us as we take more calls, 800-433-8850.

REHMAnd welcome back. Dr. Caplan, I understand you can stay with us for just a few moments, so I'll take this call from Jo Helen in Durham, N.C. and get you to respond. Good morning, Jo Helen.

JO HELENGood morning, Diane. Thank you so much for taking my call.

REHMSure.

HELENI love your show.

REHMThank you.

HELENI just wanted to comment. I am a mother of micro preemies. I had twin girls born at 26.1 weeks at Duke University due to twin to twin transfusion syndrome. And I just would like to agree with some of the comments that we as parents in those situations, we have no idea what's happening. We are so overwhelmed and concerned and just overwhelmed. And I do agree that the verbal conversations the doctors have are so much more impressive upon our decision making process than maybe reading through all that legalese of a contract.

REHMI'm sure. Dr. Caplan.

CAPLANWell, I couldn't agree more. I know we've been arguing a little bit about the importance of the written informed consent form. It is important. It should be accurate. Again, I'm going to say in my own experience and looking at studies, what the caller says is 100 percent true. The conversation, the communication particularly with parents facing a catastrophe with their baby is key. If we're going to make sure that parents understand when we're studying different approaches to treatment, that they're being asked to be in a study of those treatments, we have to quiz them.

CAPLANWe have to use things like videotape. We have to tell them to tape-record when they come into the unit so they can remember things. I hate to say this, but the forms and the IRB's these days are captured very much by lawyers. If we want to empower parents and really make them understand and be able to participate, it's that conversation and then going back and forth to establish comprehension that's key. And, Diane, this is true not just for this preemie study. We're going to see this issue come up for many, many other attempts to study differences in treatment.

REHMAll right.

CAPLANThat's what we're talking about here is studies, so-called comparative effectiveness research. There are a lot of things out there for many different diseases where there's a range of approaches to what's best to do. We want to narrow that and figure out really what is best to do, so making sure that we all understand when that option is put before us clearly. This particular preemie study has a lesson that's going to be much larger for many of us in terms of where future research is going.

REHMArthur Caplan, director of the division of medical ethics at NYU Langone Medical Center. Thanks for joining us, Dr. Caplan.

CAPLANThank you.

REHMAnd to you, Dr. Carome.

CAROMEJust a couple of points. Dr. Bell and Dr. Caplan keep saying we have this range of treatment and we don't know where is the right place to try to treat people along this span of oxygen levels. And they imply that for all comers, for all premature babies anywhere within the range might be good and there's a same good spot for every baby. And that's -- we have spoken to intensive care physicians, neonatologists, pediatricians who are not -- who are not affiliated with this study.

CAROMEAnd they explain that typically the oxygen level for babies is adjusted and customized based upon individual clinical factors, as well as considering the wishes of the parents who may want to take risk of high oxygen at the expense of some other benefit. And let me just give you an example of -- that might be better understood by your listeners. Suppose we talk about education and my kids in high school. And they take math classes. And there are three levels of math classes, all considered within normal education practice. We have advanced placement college level courses. We have college prep courses, high school level. And then we have remedial match courses.

CAROMEThat's the range of normal mathematical education. Suppose we took -- and there are reasons why students are in the AP class and other students in the remedial class, because they have different skill sets, different backgrounds and different cognitive development that allows them to perform at a certain level. Suppose we took all the children at my son's high school and randomly assigned them to the remedial education math program or to the high program. Some children are going to be hurt. Some children in both groups aren't going to do as well as they would if they got the individual math class that they should've been in. And that's a better -- an analogy of what failed in this study.

REHMAnd here's an email from Lisa in Indianapolis. "Were the doctors prohibited from adjusting oxygen levels based on their best judgment for what the patients needed? If not, wouldn't many of these infants have suffered blindness or death regardless of enrollment in this study?" Dr. Bell.

BELLWell, the medical teams caring for these children were permitted to provide additional oxygen or restrict the oxygen as needed. And I agree with one thing that Dr. Carome said, which is the ideal would be to have individualized care plans so that each baby could have the oxygen saturation target that we know is the best for that baby. Unfortunately we don’t have the information to make those judgments. And the support trial is one of the first that provides us with some basis to make those decisions with the participation of the parents, letting the parents weigh the risks of higher versus lower oxygen saturations.

BELLBut it's misleading if neonatologists have told you that they are able to individualize oxygen therapy, because we just don't have the scientific evidence to support that.

REHMAll right. And here's an email from Gail in San Antonio who says she was born in 1953, went to school with a blind boy who was also born prematurely. He was blinded during his time in the incubator from the oxygen levels. If it was known in the 1950s that high oxygen levels cause blindness in these infants, what was the level of, quote, "high oxygen" used in this research compared to that of the 1950s? Why would this research include high levels of oxygen if this risk was known 60 years ago? Dr. Bell.

BELLThat's a very important question and we've had a number of people claim that we should've anticipated this based on the findings from the 1950s. But those were very different days. Those were not carefully designed experiments. Babies were place in incubators when incubators first became available. We had a way to deliver high oxygen. And the oxygen was just pumped in at 10 liters per minute or more. And there was no way to measure the oxygen in the baby's bloodstream the way we can now.

BELLSo to say that we should've known from that experience that using oxygen saturation targets carefully controlled within a narrow range would place babies at increased risk is just delusional.

REHMInteresting. To Sheila in Jacksonville, Fla. You're on the air. Sheila, are you there? Sheila?

SHEILAYes, hi. How are you?

REHMI'm fine. Thanks. Please go right head.

SHEILAWell, I had a few questions. My first one was -- is that first one of your guests earlier said that more oxygen -- he said obviously more oxygen is better. So my question is would the protocol be that the sicker the preemie, would that preemie automatically be placed in the more oxygen group or the less oxygen group?

REHMDr. Bell.

BELLI didn't hear all of that.

REHMWell, she said that it had been said on this program, I'm not sure exactly how, and maybe Dr. Donovan, you want to...

BELLOh, that more oxygen is better.

REHMRight.

BELLYeah, that was -- that was an assumption in the 1950s. We now know that too much oxygen is harmful, just as too little oxygen is harmful.

REHMAll right.

BELLAnd we're trying to learn where is the best place between those extremes to manage premature babies.

REHMSo the second part of her question, the sicker the preemie, the more oxygen likely, Dr. Donovan?

DONOVANI was the one who said that as an example of the rough thinking about this type of problem that was available 50, 60 years ago. Clearly we have learned much more as Dr. Bell just pointed out, that more is not always better. And, in fact, trying to customize, as it was said, the treatment for an individual patient can't be done without good data. That's what this support study was trying to do was to fine tune the data so that we really would know the break point between too much oxygen and too little oxygen in terms of increased risk of death or an increased risk of eye damage.

REHMAll right. And to you, Dr. Carome, is your primary issue with this entire subject the consent form or the degree of conversation with parents informing them of the risks? Or is it the study itself?

CAROMEI think we have -- our concerns are two-fold. We certainly have concerns about the consent form and process being inadequate and not fully informing these parents so they didn't make an informed judgment about whether to put their babies in the research. Dr. Bell has said nothing to suggest that the way the benefits and risks were described were any different than the form. Secondly, beyond the consent form and process, we have fundamental concerns about whether this research as designed was even ethical in the first place.

CAROMEIt gets to the lack of a control group that got usual care without manipulating it experimentally. We only had two experimental arms. So we don't know for the population as a whole and for subsets of the population whether indeed they've established what's the ideal way to manage oxygen in premature babies.

REHMSo, Dr. Bell, how do you respond?

BELLWell, for babies who are not in the study, there is a very rough understanding of what the oxygen levels should be. And there is tremendous variation depending on the baby, the physician and the hospital. By doing the research, we just bring some (unintelligible) to that variation so that we can learn something from it. All of the consent forms, in fact, explained quite clearly that the focus -- the main focus of the study was to see which of these ranges of oxygen might help protect the eyes more from retinal prematurity.

BELLAnd it seems quite obvious and our parents seem to understand that that was the purpose of the study. And so the rate of eye problems might be higher in one group or the other, but that was not stated specifically in the paragraph on risks in some of our consent forms. And going forward I think we'll be careful to do that.

REHMAnd go ahead, Dr. Donovan.

DONOVANI think I would like to echo that point. I would differ from Dr. Carome on this. I think there were problems in the consent form as we have heard it described, at least in some of the institutions that used it. I think that the criticisms are aimed at the consent process, not at the study. The Office for Human Research Protections to the federal government nor I find that the study itself was unethical. We think that the study was worth doing, but the risks and benefits have to be clearly delineated for informed consent.

REHMDr. Kevin Donovan, he's director of the Pellegrino Center for Clinical Bioethics at Georgetown University. You're listening to "The Diane Rehm Show." Would you agree with that, Dr. Carome, that it was the form, the consent process, but that the results of the study do have some worth?

CAROMEDr. Donovan is correct. The Office for Human Research Protections only cited problems related to the consent form. The concerns about the design of the study possibly being unethical, I'm raising those on behalf of our group. And we have spoken to other experts in study design who share our concerns about the ethics of the study design. So I don't think there's unanimity that this study was well designed, ethical, and perhaps it shouldn't have been conducted the way it was.

REHMDr. Bell, if you had a preemie, would you have put that child into this study?

BELLAbsolutely. And, in fact, I would argue that it would be more unethical to continue the willy-nilly treatment of babies without any knowledge of what we're doing, that that would be more unethical than to try to find the answers.

REHMSo what is your conclusion from this study as far as going forward in the treatment of preemies is concerned?

BELLWell, as was explained earlier, the babies whose oxygen levels were kept in the lower part of the normal range we found had less eye problems, but to our surprise had a higher risk of death. And when we examined these babies at one and a half to two years of age, there was no longer a difference in the eye problems. And so we've taken this to mean that the higher oxygen saturation range, while it may cause more eye problems in the short term, it also saves lives. And most of our hospitals have adopted the use of the 91 to 95 percent range as our target.

REHMDr. Carome.

CAROMEAnd as I think his comments imply, there was a higher death rate overall by the time they measured outcomes to approximately two years. So in addition -- no difference in the blindness outcome, but more deaths in the group that got lower oxygen.

REHMDr. Donovan.

DONOVANI would agree with that. I think that also gives us reason to believe the study needed to be done, because it changed what had been the traditional practice, the normal interventions with a much wider range, now a narrower range of oxygen will save lives. This was information that surprised the investigators to some extent because it wasn't known with certainty beforehand. This is, to me, exactly why studies like this need to be done, but need to be done carefully.

REHMDr. Carome.

CAROMEThe study could've been designed better had it had an appropriate control group and would've been more informative. I think ultimately, you know, we've called on the secretary to take action that goes beyond what always should be called for. We think the parents of these babies deserve a personal apology and a full disclosure of the information they were not provided with. We're also concerned that there are several other ongoing high risk studies in the same types of populations by the same group of researchers.

CAROMEWe have demanded that the secretary make the protocols and consent documents for those transparent publicly available on a website so that independent experts can look at that. And we plan this afternoon to write to OHRP to take emergency action to make that happen.

REHMDr. Michael Carome, he's with the health research group at Public Citizen. That is, of course, a consumer advocacy organization. Dr. Kevin Donovan, director of the Pellegrino Center for Clinical Bioethics at Georgetown University Medical Center. And Dr. Bell, he is professor of pediatrics at the University of Iowa. Thank you all so much.

BELLThank you, Diane.

REHMAnd thanks for listening.

CAROMEYou're welcome.

REHMI'm Diane Rehm.