Oral argument — Svindland v. Nemours Foundation

May it please the Court, my name is Frank McClellan, along with Rick Appel who is at the Council table. We represent the appellants in the case of Svindland v. Nemours and Norwood, Dr. Norwood. Also present in the courtroom are Mr. and Mrs. Svindland. I would like to start out by emphasizing to the Court that the rulings that we seek to appeal today and reverse affect not only the Svindlands but 12 other cases pending. Would you keep your voice up, Mr. McClellan? That's much better. Okay, thank you, Judge Gorham. I would like to reserve five minutes for rebuttal, if that's permissible. Okay. And during my primary argument, I want to address two issues. One related to the suppression and the CHOP data, and the second one that relates to the mortality data. With respect to the CHOP data, we submit that under the standard of abuse of discretion that the trial court should be reversed. The standard that we're going to argue from today is set forth rather There's three parts to it. One, that there's a clearly erroneous finding of fact. Secondly, that there's an errant conclusion of law. Now, with the CHOP data, are you talking about suppressing the question of subpoena or the deliberate issue, which would be OD? Both, Your Honor. Okay, okay. And third is the improper application of law to facts. And what we submit to the Court is that Judge Schiller and his rulings on both of these issues abuse discretion in all three ways. All right. Tell me what the ruling was that you're complaining about now. Plaintiff sought, in this case, to recover raw data in the possession of researchers at CHOP. That raw data was relevant to the question, a key question in the case, which is, does the manner in which Dr. Norwood coons increased the risk of harm to children? We have 13 families who have suffered either death or personal brain injuries, and this data that is available was based upon performance of these surgeries by Dr. Norwood in a particular way at the CHOP hospital between I think it was the period that goes up to 1992 to 1997. But the raw data was destroyed, is that right? No, Your Honor. The raw data that we think still exists. Okay, it was the CFRs that were destroyed. Right, yes, the CFRs were destroyed. And there's no evidence to suggest that whatever the results of the analysis that would come out of the raw data, whatever that analysis would show, there's nothing to suggest that Dr. Norwood knew of those results. Is that right? That may be correct, Your Honor, although that's not really developed on the record in this case. That's been asserted. Wouldn't he have an obligation to show it? Because unless he knew of the data, the results of the analysis, and let's assume that you could bring the data in and the data showed that if you eliminated all the other risks and all the other variables here, and I've got another problem with that that I'll get back to you because of the multivariate nature of the analysis that has to be done but wasn't done. Assume you could eliminate all the other variables and focus only on the rate of cooling and the cooling time. And let's assume that that cooling time suggested that by reducing the cooling time down below a certain threshold, whatever that is, 12 minutes, you increase the risk in a significant manner of mortality. Unless you can show that he knew that and did not adjust his cooling time to the results of that analysis, why would it be relevant? It's relevant, Your Honor, to causation. It's relevant to informed consent. Now, how is it relevant to causation? I don't understand that because as I understand it, the jury came back. They found that Dr. Norwood had breached this duty that he owed to the clients, so that so far as negligence was concerned, that was established by the jury verdict. And then the jury came back and said, but there is no, we do not find causation. How does that relate to causation, particularly since Dr. Hannum had testified, and this was before the jury, in his report, that it was more likely than not that the death occurred because of one of three variables, air getting into the ductus, the cooling, and then the third one, which I can't at the moment remember, but I'll get before we get to it. So how does the exclusion of that evidence pertain to causation? Judge Garth, if we start with the understanding that Norwood is looking to keep out evidence that's not reliable, the second step is that once you are able to introduce that evidence, then it becomes a matter of persuasion. We have two experts conflicting opinions. One says, I think reduced cooling, fast cooling time causes harm. The other says, I don't believe that. Dr. Norwood said, I don't believe that, and it's never been proven. Now hold it a moment, Mr. McClellan. As I understand the motion in limine, it was directed only to negligence. I have the motion here, and the motion that was made by your adversaries said preclude this evidence because on the issue of negligence. At no time was there ever a motion made to preclude the evidence on the issue of causation, and I went through the whole record and could not find anything where the attorney objected saying, now wait a minute. You can exclude that matter on the issue of negligence, which I think is a problem in itself, as you point out. But at no time was there an objection made. You have never ruled, Judge, that the evidence must be excluded on the issue of causation. Judge Garth, if you look at the colloquy of the argument at what was held with respect to Daubert, the plaintiffs argued the causation issue. You may be correct in saying that the defendants did not in their own writings assert that, but certainly argument was entertained by Judge Schiller. If you look at the letter that we wrote to Judge Schiller arguing as to why it would be an error of law and abuse of discretion to quash the subpoena, there's a heavy emphasis placed on causation. He never quashed it on the issue of causation. He quashed it only on the issue of negligence. That's what the record shows. I have the ‑‑ I'm trying to get the motion. You're not talking at all about Dr. Hannum now. I thought initially you were talking about Dr. Hannum. You're talking about Dr. Golan and Chopp. Yes, we're talking about Chopp now, as I understand it. And one of the problems, Your Honors, that we face in this case is that the trial judge is not writing an opinion. That's part of the mechanism. And one of the things when I read that, there were a few papers that were published out of that analysis, on which I guess Dr. Newell was at fault. But as I read the analysis that came out of the Chopp study, it suggested that there are a lot of variables that can affect mortality of this kind of procedure, but the thing that does not apparently affect it, one of the things that can affect it, and they tested for a lot of things, sexually transmitted disease, genetic history, drug abuse, maternal drug abuse, but they concluded that one thing that did not seem to correlate to risk was cooling time. It would seem to me that the defense would want that evidence more than the plaintiff, and I wasn't quite sure what you were trying to get at when the studies that you're trying to get at through the subpoena suggest that cooling time is not one of those factors that correlates to risk. Your Honor, they did not analyze cooling time. If you look at Appendix 778, what they assert is their bold conclusion, they're experts, that the fact that they did not analyze cooling time meant in their mind that it was not relevant. The data was there, they acknowledge. No one disputes the fact that data was collected. This is the only data available within this kind of surgery being done by Dr. Newell because he's the only one and his assistants who do it this way. This data was collected and it was available. They say we did not analyze it. Now, I would submit that it's clearly premature at the discovery stage for a trial court to accept the position of one party without having an opportunity to examine and give the plaintiffs a chance to test the credibility of that expert. What about your argument that the cooling times that would be relevant to this case are skewed by the data because there's not enough of the data, there are not enough instances where the cooling time in the data is less than 15 minutes? Again, Your Honor, that's a question of credibility. You actually showed reliability, though, and if you only have three instances in all of this data where the cooling time is less than 15 minutes in the tens of thousands of data sets, why would that be enough to get you over the threshold of reliability in Dalbert? Well, Your Honor, there are, in fact, ways statistically, even with small amounts of data, for experts to present reliable testimony with respect to projections, but that was not the fact of this case. They're not even asserting there were only three. What they are asserting is that they looked at cooling times in less than 20 and found a small number of cases, but that is not proven. And at the discovery stage, we should have an opportunity to get the data. The standard is, will it lead to what may be admissible evidence? If we got the data, our expert then was able to say, here's what I found in terms of it, and they then wanted to move at trial time to not have it admitted because it lacked credibility, that would be one thing. This is at the discovery stage. We're trying to develop a case, and they have been saying, and here's a hospital, Your Honor, if I might, that has been sued on the very theory in the 90s that we are asserting in this case. They have three doctors, including one who has been a defendant, who is a part of this study, who says, we don't need to look at cooling time. That on its face raises questions of credibility. Why would you collect the data and never analyze it? Mr. McClellan, I have a very definite view of what the record shows, and the record shows that the issues you're talking about are the issues of breach of the duty, which the jury found Dr. Norwood had breached his duty. What I want to know is how you get around the causation aspect of his having breached the duty when the expert's testimony was before the jury, and he gave three different variables as to how the child could have died, and the jury came back not finding one of them as having been the cause of his breach of duty. So it does no good to talk to me about the breach of duty because I agree with you, and the jury agreed with you. They said he was negligent, but that negligence has got to be translated into causation. Did it cause the death of the child? And the jury had all of that information before. I was looking at the charge that the court gave. There was no exception to the charge at all, and I don't understand how you get to causation to correct, if you can, the jury's verdict that there was no causation in what Dr. Norwood did that caused the baby's death. If you can explain that to me and tell me what evidence bears on the issue of causation, you will do a good deal better so far as I'm concerned. Thank you. May I try, Your Honor? First, we don't know what theory persuaded the jury that he violated the standard of care. They may have decided he violated the standard of care because there was a failure to close the ductus. Yes, they could have. If they had found that and then they decided, but we're not convinced that that causation evidence in connection with failure to close the ductus was the cause of the death, they then find negligence and they find no causation. Well, they found negligence. They found negligence. It was a general verdict. It was not an interrogatory verdict. They came back and said no causation. And no causation. Assume Judge Garza, the jury, deliberates and says, we agree there was negligence and failure to close the ductus. Now let's decide whether or not there was negligence with respect to cooling. That wasn't asked. Those questions were not asked in the interrogatory. Now I assume that the Svendland's attorney could have asked to have an interrogatory type of verdict. I assume that. If he had said, do you find it negligent, do you find Dr. Norwood was negligent in cooling for only eight, six seconds, eight seconds instead of 20 seconds, they could have answered that yes or no. Do you find that he was negligent in having air get into the ductus? They could have answered that yes or no. Do they find he was negligent in any other respect, yes or no? But that wasn't asked. And, therefore, they have a general verdict of negligence in which he breached his duty, but they say that breach of duty did not result in the baby's death. Well, Judge Garza, let's assume that you don't accept the first argument, that we have to bear the burden because we did not submit the interrogatory. However, we should not be prejudiced if the trial court views this question excluding the evidence. Now let me just say this. Assume that the jury now finds negligence in their deliberation with respect to cooling. They now have before them- But they didn't. They didn't. They didn't. And that assumption goes at the root of the case. We don't know, Judge Garza. We don't know what they found. We don't know what they found. So I'm trying to just go through the reasoning that would lead us to the relevance of this evidence to causation. Assume they found negligence in fast cooling. And then they debated, was it a causation? What they had before them was testimony unimpeached of Dr. Millward, that I've done thousands of these, and it doesn't cause a problem. We're not allowed to cross-examine him. Remember, what about these 13 cases where it did cause a problem? We're not allowed to cross-examine him with respect to what about this CHOP data. We're not allowed to cross-examine him with respect to what about the mortality data. All of that is kept out. That's my problem with your argument, because you're assuming that the data will allow you to focus on one on the causation. There are so many variables in that data. And your own expert, Dr. Hannon, admitted that you had to do a risk adjustment to try to focus it on cooling time, and that he didn't do it. Oh, he did do it, Your Honor. I thought he said he did not. I know, Your Honor. He says very explicitly, what I did was to use the RACS method. And that's not even disputed. What the defendants and police dispute is whether or not the way in which he did the RACS analysis was appropriate. And the RACS analysis is published. It's peer-reviewed. It's published. It's adopted by the best experts in the field. And here's what he did. He said there are statistics that I have that they were required by law to file with the Department of Health in Delaware. I took all of those, which would be the gross mortality data. I then selected from those VSD surgeries which would match this case. From that, I found that there were six deaths with the kind of a surgery that should produce zero deaths in comparable institutions. Does that include the medical patient's history? When you say surgeries that would match this case, does that include an identical medical history in terms of other diseases other than the VSD? The problem I had with it is that there's no uniform system of coding so that if a VSD surgery is combined, as I understand it, and you know better than me, but if the VSD procedure is combined with another procedure and the surgery results end up, the procedure, the overall procedure may be coded as something other than a VSD because of the coexistence of another surgical procedure along with the VSD once the surgeon is in there. My problem is, again, the fit and the reliability in making sure that we're comparing apples to apples. Judge McKee, if we could get the court to read the RACS articles that are in the appendix, we would end this argument on that basis because the very purpose of the RACS system was to take into account those variables. What they decided was, bringing all these experts together, that one way in which we could compare apples to apples within an institution and apples and apples in other institutions is to focus on the procedure. Focusing on that procedure will allow us to come out with not necessarily prediction on an individual case but overall predictions as to what the rates are because if one is on a VSD procedure, it's very different from repairing an aortic arch. It's very different from HLM. My problem is let's assume that infant A has VSD and also a pneumonia at the time. Assuming you wouldn't operate if the kid had pneumonia, but just assume that. And you code that as a VSD and patient B just has the VSD without the pneumonia. If you look at them both as being VSD procedures and comparing the mortality, you're ignoring the fact that one infant had pneumonia at the time of the procedure and had a much higher risk and a much higher likelihood of an uncurable outcome. The assumption under the RACS system which they found reliable and persuasive as experts is that each institution will have a mix of those kinds of cases of VSD people. If one institution can do 200 operations of VSDs and have no mortality, the other one ends up with 10 mortalities, that tells us something meaningful. And you know, one other myth. Don't you then have to have a high enough number of inquiries to make that statistically? Let's assume, I know this is in the record, let's assume that one institution has just a single VSD procedure and has an uncurable outcome and compare that against another data set with 1,000 VSD procedures. You don't have any system that I can see, and I don't pretend to be a statistician, but that doesn't give you any kind of data set that you can compare because you don't have enough data in the first set to allow you to have a statistically meaningful inquiry. There's no ability to get to a standard deviation. There's no way of determining whether or not the one situation you're looking at is typical of the universe that you're saying results in the kind of risk that you're arguing. I'm sorry, Your Honor. The data that's collected under that RAC system, was there any effort to determine whether or not it was possible to analyze and remove risk factors that would not be relevant to the issue? If you look at what Dr. Hannum did, and he testified, and it's in the appendix, that's exactly what he did. He removed all of those surgeries with VSD that had concomitant kinds of risk that would not fit into this and make it a fair comparison. That's what he did. The other point I would make is as long as we're talking about Dr. Hannum, his report, which was before the jury and on which they could cross-examine him and on which they did cross-examine him, stated that Dr. Norwood's practices, and I'm reading from appendix page 46, placed Ian Svendron at unreasonable risk for neurological damage, hypoperfusion injury, and air embolism injury. He has stated in his letter that given Ian's post-operative clinical course, the most likely cause of Ian's decreased heart function and failure is air embolism introduced through an open ductus, that is the failure to litigate the ductus. Now all of those matters were before the jury. They were before the jury, I suggest to you, on the issue of negligence, and the jury found that there was negligence. But he ascribed three different possibilities leading to the death of the child, which is a dreadful thing, and I can't tell you how much this panel feels sympathy toward the parents of this child, but we're dealing with the record here. And the record shows that these three things, which Dr. Norwood did, breached his duty of care. But there is nothing that I can find in the record that says that anyone, particularly the cooler, was the causation of the child's death. It may have been not litigating the ductus. It may have been an infection. It may have been neurological damage. All of these things were before the jury, and the jury didn't answer any of them except to say no causation. Now how it got to that conclusion, I don't know, because they found them to be negligent, but they came to that conclusion. And you're asking us to review a jury's determination of no causation when they had all of the evidence before it on causation and came out and said there was none. Judge Garth, they did not have the critical evidence that the trial judge excluded. What critical evidence? The chart data, which would have enabled the plaintiff's expert to say. Let's take them one at a time. Judge McKee asked you about, and Judge Rodriguez asked you about, the elimination of risk factors on each of the children that the VSD operation was performed upon. And as I understand Judge Schiller, he refused to have many trials within this one trial of having every child looked at to see whether it was an infection, error in the ambulance cooling that caused the child's death or that didn't cause it. That's what I understand. Respectfully, Judge Garth, they're separate issues. The RACS data and the argument about the VSD relates only to the mortality data. The information of the raw data underlying the CHOP study, which was done earlier, is there. It's published. And so, therefore, it's all a matter of are we entitled to look at it, to have our expert. Now, he's allowed to look at it. What good is the raw data without the CFRs? Well, the raw data, they don't even contend, Your Honor, that the CFRs are essential for purposes of trying to prove, to assess the cooling part. They argue that that would be helpful because they would allow us to get more information about the cases. But isn't that a question of credibility? They can cross-examine our expert and say that isn't what the effect of this have. I would point out the best, the strongest argument we could make, is in the defendant's own expert's letter on this RACS system. If you look at the appendix at 549, and here's what he said. 549. Yes, Your Honor. And here's what he wrote to them before he was converted into an expert to testify at this trial. They asked him, what system should we use to analyze mortality data? He said, don't use crude mortality data. And then he said, I suggest you should use the RACS 1 method, understanding that it's not a perfect system and has some drawbacks. And then he said, talks about what the drawbacks are. But then he said, I think the RACS method has better published statistical validity than the Aristotle method. It's also important, and the final point is, there's no reliable proven method to convert it into other than this RACS system. So we're using an accepted scientific methodology, Dr. Hannon is. I'm now back on the mortality data. And he's been going through and getting the VSDs using the RACS system. That is the best that we can do. And at a premature level, we're being precluded from ever getting that evidence to support us. And if we had that evidence, what we could have said to the jury, you've heard Dr. Hannon testify. There is a causal link. You've heard the other experts testify. There isn't. You now have before you evidence that shows when they cooled these babies with VSDs, this is simply a patch, a little hole they got a patch. Six of them died. Why did that happen? Do you believe Dr. Hannon in terms of his expert opinion, or do you believe their expert? Similarly, if we would have been able to get the top data, and we say here's a situation where you have all these cases, 200, 300 cases, and the cooling data shows you X, do you believe our expert or do you believe their expert? Finally, Judge Garth, on this causation, I know you're taking the view that because we did not take a special interrogatory, that we should bear the burden. Not just the interrogatories. Objections made to the charge, supplying a charge to the jury, any of those matters for which the preclusion of evidence was not directed, the preclusion of evidence was directed at negligence and negligence only. I'm reading from page 31 of the record. It says we're talking about mortality rates. Additionally, this evidence, mortality rates, is irrelevant and has no bearing on whether Dr. Norwood was negligent in this case. That's the motion that was made. Again, it's repeated on the following page, 32. However, if this court permits such testimony, the causes of action should be bifurcated, and the evidence should be admitted, precluded in any claim directly against Dr. Norwood or the hospital for negligence. Negligence, that's what this whole mortality issue was about, not causation. Your Honor, we fought hard with Judge Shiller to recognize that it wasn't just negligence but causation. Where does it show in the record? Because I couldn't find it. I looked for it. There's a letter from counsel, and there's also colloquy. Where does it say that the mortality exclusion of evidence is directed to causation? Because I want to see that. I couldn't find it, but perhaps I didn't read carefully enough. I know I've used up my rebuttal time, but I can find this. Well, you haven't used up your rebuttal time, but you can look at it during your opposing colleague's argument. If you can't find it, you can submit a 28-J letter pointing it. I would like to know before we leave today, and perhaps Mr. McClellan can find it during the recess period, during the rebuttal period. Yes, Your Honor. There's a letter that we submitted, a second letter from Ms. Blanco to the judge saying, Judge, this is not just about negligence, it's about causation. That I would like to see. I have never seen that. Okay. You'll give me the appendix reference to it. I will, Your Honor. I will, Your Honor. Your chances are your colleagues will have a little more time than the 15 minutes allotted since you did, and you might have to be able to dig it out before your rebuttal. Well, this is an important case. It is. Well, they're all important, but this is particularly tragic. Your Honor, the final point that I would make is that in addition to precluding the plaintiffs from getting this evidence, it set up a trial where the defendant, Mayerwood, had a safety zone to testify. In other words, he knew before the trial started that we would not be able to impeach his testimony, that he's done thousands of these and it works well, because we could not come in with data from CHOP. We could not come in with mortality data. We could not even cross-examine him about cases in which he was a defendant and, in fact, had two people in less than 10 minutes and they died or had brain damage. So that's not a fair trial. Well, I will be anxious to see that letter when the court received it, because if, in fact, it came in prior to the trial and it did direct the court to evidence pertaining to causation, you might have a point, Mr. McClellan. It's that point that really has hung me up on this case. I understand, Your Honor, and I will find that before I get back up again. Thank you. Thank you. Ms. Petrosky? Good morning. Good morning. May it please the Court, my name is Sarah Lynn Petrosky, and I have the privilege of representing Dr. Norwood and the Nemours Foundation. Today we ask this Court to affirm the lower court's decisions as there was no abuse of discretion and a new trial is not warranted. Can you tell me where, if your study of the record was more thorough than mine, as it should have been, there is any direction with respect to the exclusion of the evidence Mr. McClellan has pointed out pertaining to causation? I believe that at times, and in the letter that Mr. McClellan is referring to, which came in late in the sequence of things. What do you mean late? There were a number of hearings that this judge had on this topic. Did it come in before the trial? It did come in before the trial, Your Honor, and I do believe that in a series of a number of things that Mr. McClellan is right We believe that doesn't change the ruling that this Court should have affirming Judge Schiller because he did not abuse his discretion, and I would like the opportunity to explain why. There's so many broader issues and question issues here. When you say not abuse of discretion, can you kind of make me understand what you're talking about? The self-subpoena, Dr. Hannum, Dr. Reichert, what you're talking about? Yes, and that's how I'm prepared to go through the argument, if you will, Your Honor. There are four issues that the Finlands have raised on appeal. Mr. Dreyer is here on behalf of Children's Hospital of Philadelphia, and he has reserved some time in my argument to address the child subpoena, although I will address the fact that there was no relevance of that data to the case. So I'll go through the four issues in the way that they were raised in the brief. I think because there's so many data sets from so many sources, that may be prudent. Judge Schiller sits as a discovery court judge on the Finland case, as well as a dozen cases brought by the Finland's lawyers with the same theories and the same discovery requests. He was armed with – is that better, Judge Garza? A little bit higher. Higher. He was armed with extensive briefs, medical literature, deposition transcripts, an argument zealously fought but civilly fought by counsel on all of these points. And yet the judge made decisions that we believe should be affirmed on appeal because there was no abuse of discretion. At times, Judge Schiller made decisions that were favorable and unfavorable to both sides. Mr. Dreyer is arguing that basically he was put in a position where his hands were tied, that the client, Dr. Nohru, was able to get on his hands and talk about his history of performing these operations. And I said it from Professor McClellan. Mr. McClellan was not able to cross-examine Dr. Nohru on the bad results that he, the professor, and Mr. McClellan knew were out there. Dr. Nohru testified that he had performed 8,000 surgeries in his history, and it would be remarkable to have a medical malpractice case against the descendant doctor and not ask the physician where he went to medical school, where he learned to do heart surgery, and how many he had done. Beyond that, Dr. Nohru did not testify. There was no testimony that X percentage of his patients died or didn't die or what the causes were. And certainly a doctor like Dr. Nohru, who was at the forefront of cardiothoracic surgery in this country, had death and had death experience. We didn't go into that at trial. There was no comparison. What the Simlins wanted to do was take raw data that they got from the Delaware Health Statistics Department, derive a percentage, and Dr. Hannan, their expert, did that. And he wanted to compare that with a website that Boston Children's Hospital has up on the Internet that said out of 89 surgeries, they had 99% survival. I know that's been the case, but I thought Mr. McClellan was specifically talking about Hannan's attempt to compare the Delaware databank with Dr. Norwood's individual, not the Boston databank. The Delaware databank, he wanted to come up with a number. His number was 6.74%. He had to do something with that number. You can't just say, oh, here's a number. It's meaningless to the jury without comparison. The one thing they could compare it with was a hearsay website from Boston Children's that wasn't even mathematically correct. Did they compare it with Dr. Norwood's specific practice and his individualized cases? What he attempted to do, and we believe it's unreliable, was to come up with a number that would show how many patient deaths Dr. Norwood had of VSD surgeries while he was at DuPont. The data he used was billing codes. It wasn't separated by surgeons, so he couldn't tell us which surgeries were done by Dr. Norwood, which were done by Dr. Pizarro, which were done by Dr. Mainwaring or other surgeons who had operated at DuPont. But would it matter? Assuming everything else was okay, would it matter when the law of probability operate to negate that one variable? Oh, that's so unreliable. You couldn't take a raw number like that. He was doing crude mortality rates, and the letter Mr. McClellan referred you to, by the way, is by Murray Pollack, who is a physician who is cited in the RAC's data as one of the leading experts on this. And he said you can't just take a numerator over a denominator and come up with a meaningful number. And there's a couple reasons why. One, again, is it didn't differentiate among doctors. One surgeon who's younger and junior and less experienced might have a different rate. One who's senior and more experienced might be taking patients that have comorbidities like some of the ones that Dr. Norwood operated on. Norwood and the RAC system, maybe I'm misunderstanding the RAC system. Mr. McClellan was saying that there's a way statistically of neutralizing all those variables and getting them out of the case so you can just compare apples to apples. The RAC system is on I think it's 105 of your appendix, not that you'll look now. But what you'll see is that it compiles surgeries based on the type of surgery. All the RAC system does is say a VSD is in Category 2, like ASD or some other sort of cardiac surgery. It doesn't tell you that the baby also has a genetic disease that's 95% fatal. It doesn't neutralize the medical history or anything of that sort? No. And RAC is a good system. There's a new system now, by the way, called Aristotle that hospitals are using. Was that the basis upon which the district court judge ruled that there were too many other risk factors? No. There were a variety of issues raised. One is that it was completely irrelevant. Mortality data never has been used in any case to establish standard of care or causation. So even if you could show that Dr. Norwood's mortality rate was higher, which we don't believe you can, but even if you can reliably establish it, it doesn't show what happened to Ian Svindlund and why Ian Svindlund died. It doesn't tell you what the standard of care was in 2003 when Ian was operated on, and it doesn't tell you what happened to Ian Svindlund. The data that Dr. Hannan was using could not be adjusted for the complexity of the cases that Dr. Norwood handled. In addition, it comes from billing sources. Their own expert, Dr. Hannan, said it was invalid because it wasn't risk adjusted. Well, I asked Mr. McClendon. He specifically said that that's not true. That was my reading of the record, too, but he said that Hannan did not say that. Hannan conceded that it had to do with risk adjustment, and I thought Hannan conceded that he had not done an appropriate risk adjustment. I think we might be talking about two things when we're saying risk adjusted, and it didn't occur to me until I heard Mr. McClelland's argument. When he's talking about risk adjustment, I think he's talking about what RACS does, which is Categorized Congenital Heart Defect Surgery into Categories. There's something called hypoplastic left heart syndrome that's a baby born with half a heart. Those are Category 6 because we still have 80% kids surviving and 20% dying after those operations in this country. It's the best institution. VFDs are in Category 2. And I think when Mr. McClelland's referring to risk adjustment, he's referring to that. When we're talking about risk adjustment, and our expert was talking about risk adjustment, and it's clear in the letters that are in the appendix, they're talking about comorbidities, other problems that these babies have. You cannot isolate a patient just because they have a VFD. They don't come in neat little packages. One VFD isn't the same as another, for example. If you have the information and the records to analyze it and continue to remove what would be other risk factors. Not from the billing code data that Dr. Hannon used. Not even from the data that Judge Schiller required. We produced a separate set of data from the hospital that was much more complete. Instead of two comorbidities, it showed up to eight comorbidities. So if a baby, for example, is born prematurely, has a heart defect, and a genetic condition, all three of those are shown in our system. Under the system Dr. Hannon chose to look at, they weren't shown. By the way, by the time he was deposed, Dr. Hannon no longer had the data he looked at. He said it was on his computer and he deleted it within three days, and he was not able to explain how he arrived at his number. Furthermore, at trial, despite the fact that Judge Schiller entered an order saying you may not talk about mortality data, Dr. Hannon said that, one, he has no VFD death. Two, he tells his patients that absent lightning striking his equipment, they should not expect a VFD death. And three, that Dr. Norwood's seizure and death outcomes were ten times higher simply because of the way he cooled. All of that evidence got to the jury despite Judge Schiller's ruling. So if there was an abuse of discretion, it certainly didn't keep mortality rates out of the case and it didn't impact the jury decision. But if the doctor testified that he did some 800 surgeries, and I think there was a question as to whether or not there were complications with the 800 surgeries, the defense expert said that the silence in a record-keeping, whether or not it related to cooling, meant that they were favorable. Now, if there's no record, is it appropriate to put that concept in against 800 surgeries without allowing an individualized search of that specific doctor's records? Your Honor, you're referring to a report that was written by Dr. Warner, a defense expert, in which he said, in a report that didn't go to the jury, that one of the things that he thought was telling, it's a 20-page report, by the way, one of the things he thinks is telling is that certain articles were printed over a period of time, listing cooling times of 11 minutes, 10 minutes, and so on, and there's a way in medical literature for people to comment on that and respond to that when peer-reviewed literature comes out, and no one ever responded to those saying, oh, my goodness, there's shock and awe, you're cooling for 10 minutes or 11 minutes. So, first of all, it didn't go to the jury. The jury heard none of that. Secondly, it was written in a defense expert report that wasn't read, disclosed, or talked about at the time of trial. And thirdly, it was important for the fact that people didn't respond to those articles, not what was actually in those articles. Ms. Petrosky, as I understand your argument to this point, it all centers on the fact that Dr. Norwood's breach of duty, which the jury found had been breached. Where is the causation aspect of what you're speaking about? Where does the lack of mortality data fit into the causation aspect? Plaintiffs suggest that they could use the mortality data to show, if you cool one way and you have a death, it must be related. That's simply not the way science works, and that's why we have Daubert and other cases on junk science prohibiting this very kind of evidence. The trial judge was proper in saying, you cannot use mortality data to show what caused Ian Finlan's death. And not only did the- Now, where did he say that? Did he say that, or did he just enter an order? He entered an order based on arguments that this would be inappropriate. And again, that was- So the argument was made that mortality data was irrelevant. Irrelevant. And then he just entered an order saying, keep it out. Well, there were other arguments as well. It was irrelevant. It was not probative of what happened in this case. It would be more prejudicial. It was an attempt to use prior bad acts. Did he say keep it out as to negligence, or did he say keep it out as to causation, or did he say both? He said keep it out. The order is silent about that. Well, the motion that was made was predicated on negligence, was it not? The motion that was made was that the cooling data does not establish standard of care or tell what caused an injury in this case. It is not relevant. I didn't see the cause of the motion. So it wasn't a disjunctive, it was or? Pardon me? It was disjunctive? You're saying it was caused standard of care or? I believe all of those issues were raised before the trial court, before we received the final order. Well, what about the motion that was made in limine? I read that motion, motion in limine, to preclude mortality data under Daubert and rules 403, 404, and 702, and it was made, as I read it, only on the issue of negligence. Where does the causation appear? I believe it appeared later on. There were a number of. Later on in the motion? No, later on in the proceedings. It's a little bit disjointed because Judge Shiller said. Where does it appear in the record, Ms. Petrosky? We're bound by the record. Right. Where does it appear in the record that causation was that no mortality data could be induced with respect to causation? Because I couldn't find any. I believe that it's the same place Mr. McClellan was referring to, letters that were written subsequent to the motions being filed and in the colloquy with the court when we had argument. We had argument several times on this issue, and I believe it was in there, and it will be in the underlying record. Does the colloquy of the court, that has been transcribed? Yes, it has, Your Honor. Can you point to me where in the record it deals with causation? I mean, that's the thing that I'm hung up on, and I would think that that's what you might want to concentrate on, at least so far as I'm concerned. I know that my time is up, and I have other issues. Should I do the same thing that Mr. McClellan did and look at that and give you a page reference and then make sure I address the other arguments? Sure. Thank you, Your Honor. I appreciate your tolerance. Secondly, there was the issue of the – Your time is up, and you can't get me that. Perhaps you ought to reserve your rebuttal. She doesn't have a rebuttal, though. Well, I have – Mr. Dreyer also has his time. You're suggesting taking a break, recessing, yielding the floor, if you will. Okay. Having the intervening room up and then coming back up afterwards. Okay. I'm not suggesting that. Is that what you're arguing for? Well, I would be happy to find the page reference for Your Honor in whatever context or however you would prefer. Are you objecting to that? I just want to make sure I don't impede my ability to address the other issues. That would be great so that you at least have an opportunity to ask questions. I'm not objecting to it. I don't have any problem with it. If you want to just kind of cede your balance of your time is gone, but allow the intervener to address us, and then come back up and try to briefly make the other points you want to make. We're obviously not being locked into the red light anyhow. Right. Or should I address the three other arguments now? Why don't you just address them? Let me just address them now real briefly, because I don't know that there's much question about it. They're going to merge together. Right. Those two do merge together a little bit, and they merge together because the Simmons lawyers have, and they've launched these other cases, with an argument that has not been proven in science. I mean, part of the problem here and— What is the argument? The argument is that a short cooling duration is associated with bad outcomes, simply put. Their argument is if you cool too quickly, you'll have a bad result. There is no medical literature that they've uncovered that's peer-reviewed and reliable and that they can make that point. On the mortality data, would there be a way— and again, I don't have the statistical background to really answer this question, and maybe it came out in the argument. Would there be a way of conducting a multivariate analysis, which is sufficiently sophisticated, to just isolate cooling time as being the operative variable and conclude that beyond a certain time threshold, wherever that is, be it 15 minutes, 60 minutes, 20 minutes, once you get below that cutoff point, you're going to have outcomes that are significantly less favorable than you will have above that point. I assume that you could perform such analysis, and I say that's what he's asking. There's a question that you can't do it prospectively. In other words, you can't take kids and put them in two categories and say we're going to cool this group for 10 minutes and this group for 20 minutes and see who does better. No, I'm saying it can't be done. That obviously can't be done. But in the data that we have— To the extent it can, Children's Hospital did it in 2001 and 2003, and that's how we know about this article, those two articles at the Finland site, that say we looked at all these variables and we found no association between shorter cooling times and outcomes. Okay, but Mr. McClellan is saying if he can get a hold of the raw data for those studies, he could do his own analysis on that, and maybe there is some kind of multivariate analysis that would suggest that there is a causation relationship between cooling time and mortality. He's saying because they have a novel theory in this case that they can't prove with real science, that litigants ought to have underlying data that was used in an NIH grant and which was found not associated with that outcome to tweak it and find— Yes, Your Honor. Isn't it a question that should be pursued that if, in fact, someone cools for, say, four minutes, and there's some suggestion that the cooling took four minutes, whether that would be sufficient to cool the entire body when an operation has taken perhaps 33 minutes to reinstitute the circulation and the body is oxygen-deprived for a period of time where the four minutes may be an indication that it didn't cool sufficiently. So there is a point where the body's warmth demands oxygen. And it's not there if the body isn't cooled. So if you're looking at that, is there a way then to look at the record and see whether there's a bottom line that should not be crossed because when you do, then you have bad results? And would that be relevant? The answer is that that question should be raised, and it is raised, and it has been raised, but it's not to be raised in courts of law. It should be raised in science laboratories, hospitals, and places where they have raised it. And the Finns weren't prohibited from bringing textbook after textbook, journal article after journal article, and expert after expert to explain the concepts of cooling, as did the defendants. If anything, we had a week, a week and a half, of testimony by very, very smart people with the best scientific literature possible explaining these concepts of warmth and cooling and oxygenation. It should be done, and it has been done, but it shouldn't be done in the court of law. And what the Finns are trying to do is say that a litigant unhappy with the medical literature and a conclusion in a medical article ought to be able to get that raw data and use it for its own purposes because it can't find a legitimate scientific article to point to. And that's the real problem with the subpoena of Chopp and Goyne. The third issue that they raised on appeal had to do with Jack Rycheck, the physician who testified to an isolated area. You're talking about the HIPAA argument? The HIPAA argument. I'm not. Do you want to hear any testimony on that? No. There is a new case in December, Luna, that addresses whether HIPAA preempts. The judge speaks for me, too. Okay. Okay. Thank you, Your Honor. And then there was an issue about whether the judge abuses discretion in using the standard Delaware charge, and I assume you want no argument on that as well. No, we don't want to. I don't want to hear it. Okay. Thank you. So in summary, Your Honors, we believe that there was no abuse of discretion and certainly that no new trial is warranted. Can you answer me one question? At any time during the testimony of Dr. Hannum and Dr. Hannum's report, did he ever segregate the causes that could lead to a child's death among the three liabilities that he spoke about, hypofusion, litigating the ductus, and the cooling? Did he ever say the hypofusion was not the cause of the death, that the infection was not the cause of the death, that neurological damage wasn't the cause of the death, the only thing that caused the death was the time spent in cooling? No, he did not. And, of course, that's what makes this harmless error if it's error at all. Not only did he give those three issues, Judge Garth, Dr. Hannum and Dr. DuPlessis gave a myriad of reasons that could have caused Ian's death. Say that again. Both Dr. DuPlessis and Dr. Hannum gave a myriad of reasons that could have caused Ian's death. They said there was a hypointensive event six hours after surgery. Were they all before the jury? This is all before the jury. If you look at what is page 40 of DuPlessis and Hannum's deposition. The jury found that none of them caused the death. And none of them caused the death. Dr. Hannum actually said, well, maybe they didn't give enough heart medicine to stop the heart, called cardioplegia. Maybe it was the PDA wasn't ligated. Maybe his lactate level, he actually said, his lactate level was too high before the operation, so maybe he had hyperfusion problems then. And then, of course, he admitted on cross-examination that you can have a death from pediatric open-heart surgery even when you cool 15 to 20 minutes like they suggested. And, therefore, the jury did find that there was no causation. And if there was any problem with that, it wasn't the trial judge's orders, it was the stimulus testimony. Thank you very much. Thank you. We're here for your colleague, Mr. Dreyer. Is it Dreyer or Dreyer? Dreyer. May it please the Court, Alex Dreyer for the Children's Hospital of Philadelphia and Dr. James Gowen. Would you move that microphone so that you're talking into it? How's that, Your Honor? Much better. Thank you. Children's Hospital. You ought to tell counsel to have a big sign up there. Remember to speak into the microphone. Go ahead, sir. I should have known from your advice to other counsel, Your Honor. Children's Hospital and Dr. Gowen, of course, were not parties below and intervened here to defend the district court's decision. Help me understand the relationship between the CFRs and the raw data and why the raw data is kind of, from your position, not that significant with other CFRs. The case report forms are the forms that were used to collect the information, and they are really what's used to create the database. And it was Dr. Gowen's testimony. Dr. Gowen is the biostatistician who built this database. He wrote all the code for it. He's the person who really knows this database. And his testimony is that there's a tremendous obstacle to validating this data if you don't go back to those CRFs, that what a statistician does when he runs statistical analysis is he first needs to make sure that the data is validated. And it's been a number of years now since this data set was used for statistical analysis. It's been used for some other things, like going back to do follow-up with patients. But it's been years since those CRFs were destroyed over the course of time. And his testimony is that you would have a difficult time in identifying which field is which. There are codes, you know, 0, 1, and 2. It doesn't say seizure. It doesn't say death. You would have to go back field by field. Some of the fields are transparent. Many of the fields are opaque. And keep in mind, this is a data set that has thousands of fields. It's an enormous database. This was not just framing what was before the district judge here, and it is an abuse of discretion standard, and so we need to look at the record that was before Judge Schiller. This was an unusual kind of discovery request. First, of course, CHOP is not a party to the litigation, and so we're under Rule 45. So the standard there is whether they had a substantial need for the data. That's because Rule 45C talks about confidential research information and says they have to show a substantial need. So what Judge Schiller was trying to decide is, based on everything before him, and he had evidence about the lack of relevance, the fact that Dr. Norwood was not connected with this, wasn't involved in the collection of it or in the analysis of it. His name was only on the article because his patients participated. Based on the fact that what they were asking to do had already been done, that's undisputed testimony from Dr. Clancy, the principal investigator, who said they looked at all the variables. What they were trying to do here … Why was they not content with the analysis performed by CHOP as opposed to doing – assuming they could and understand the absence of the IG calling them CFRs for obvious reasons that lawyers will understand, but the absence of the CRS might negate the usefulness of the raw data. That's a separate issue. Why should they be content with your analysis as opposed to doing their own? Well, they're not content, but the question was whether their desire to get data where there was questionable usefulness, questionable at best relevance, where the analysis had already been done and published in a peer-reviewed article. This was not just the testimony of one expert. This was a peer-reviewed article that said they looked at other variables and the only three they found statistically significant were three others. The court had to look at all of that evidence based on this substantial need standard and based on the important interest in protecting the data, that these were several hundred patients who had no connection to the litigation who consented to have their data used for scientific research in the service of science and medicine but did not sign up to have it given over to an expert for a litigant to use for narrow litigation purposes. So some side of their hypocrisy. And so the judge was really looking at both sides of that equation, and he was looking at the case reporting form issue perhaps not as an absolute issue, can you do it or can you not, but as something to weigh in the analysis that it would be very difficult, that it would be difficult to validate, that it would be extremely burdensome to redact, to construct this back in a way that it could be useful. The fact that we had a published article that had considered this variable and found it not to be relevant and the limited relevance of this to the case before him, and he made a judgment based on that balancing that we would submit was well within the considerable discretion he had to manage discovery. Are you in a position to state whether or not the exclusion of that evidence, more on negligence or causation? As I understood the request from plaintiff's counsel, the issue was whether Dr. Norwood, their letter specifically says that they wanted to get at whether Dr. Norwood knew about this data. And, again, the undisputed evidence, the testimony of Dr. Clancy, was that Dr. Norwood didn't know anything about the data set, that he left CHOP in 1994. The data was collected between 92 and 97, and he didn't even have access to it after he left. I'm sorry, Judge? In any event, the standard of care, that's not causation. Norwood's knowledge would not go that far. That, to me, sounds like standard of care. I mean, on the question of causation, the article reports, again, many, many variables. And the conclusion of this article was that there are a lot of different variables that can explain adverse neurologic events. They found three that were significantly related, cardiac, anatomy, genetic predisposition, and DHDA cooling, a different kind of cooling. Do you understand how the jury came to its verdict of no causation once they found that Dr. Norwood had been negligent, no matter what the negligence consisted of? Do you understand that? It's really a question for Ms. Petrosky, Your Honor, because I wasn't involved in the trial. We only appeared for purposes of the subpoena. Well, you know the results of the subpoena. You know the results of the verdict. Based on what I know, Judge, I can't understand it. Well, I can understand it, I should say. I can understand the issue about it, because based on what I know, which is really just this article and the testimony of the principal investigator, what they essentially said was this was very complicated, and you could not identify. This is my interpretation, not his words. You could not identify a single but-for cause that there were a number of different variables that can explain these adverse neurologic events. I realize this may be beyond the scope of what you wanted to speak with us about, but how would you as an attorney prove causation in this case? What evidence would you want to prove causation in this case that you can garner from the evidence in this trial? I really have to defer. I don't try cases like this, and I'm well out of my league. Well, it's a question that has occurred to me as to what would be necessary. Just before your colleague sat down, she said, well, these are ticklish operations, and the jury obviously found that there was nothing in this operation that ---- I would think you would need experts testifying based on peer-reviewed literature, that you would not do it by seeking raw data and trying to do it on your own. Would you have a right to explore the contributions of one of the contributors to that peer-reviewed article? You could, Your Honor. I mean, it would depend, again, on a balancing under the discovery rules, under Rule 45. There's another issue, which is beyond what you're arguing here, of 404 billion knowledge and informed consent. Is there any way at least you could garner sufficient information? And if it's inadequate, it's certainly subject at a double hearing at that point to refuse it. But without even having an opportunity to see whether an argument can be made, do you think it's then appropriate simply to say, well, it's too difficult and you can't get at it? I think on these facts, and again, there were a number of facts that Judge Shiller had before him. And on these facts, it was a balancing. Clearly, there's always an interest of one party to discover anything that they think they might be able to turn to their favor. But the Federal Rules of Civil Procedure make it quite clear that the district judge needs to weigh that against the interest in protecting confidential research information. And I think the record shows that that is exactly what he did. All of this information about the confidentiality of this data, the fact that there was an NIH contract that guaranteed confidentiality, the fact that there was informed consent of these patients to use the data for research but not to turn it over in this context, all of that was before the judge here. And he also had their arguments on what they wanted to do with it. He also had the argument or the testimony that the data set here really wouldn't be amenable to the kind of comparison they wanted to do because, as was pointed out earlier, 75 percent of the patients in this article had cooling times of 13 minutes or lower. So some number greater than 75 percent had cooling times of 15 minutes or lower. And Dr. Clancy's evidence was it wouldn't be meaningful to compare those groups. What they wanted to do was to compare the longer cooling times to the shorter cooling times. So the judge had that before him. And he also had before him the fact that this was a third party, that CHOP is not a party here. Dr. Clancy, and there was a reference to other lawsuits and the credibility here. Dr. Clancy is a professor of neurology. There's absolutely no issue with his credibility. He had no incentive to fabricate this data, to misinterpret this data. It was a peer-reviewed article. So that's what the judge had before him. And he exercised his discretion in balancing those factors. I'm not too certain what it all shows, but if the data indicates that the majority of cases were 15 minutes or more, they are looking at the cooling time. Is there a way to look at that same data and say none of them go below four minutes, some of them go to five minutes, and at least have some starting point to determine whether or not there's a bottom line? It's a little over 15 minutes, and that's it. The majority were under 15 minutes. The vast majority were under 15 minutes, and there were only a very few that were in the category that the plaintiffs asserted was dangerous. And so the issue was that you couldn't make a valid comparison, according to Dr. Clancy, between those 15-plus or 20-plus and the 15-minus or 20-minus. Certainly you could do an analysis of cooling times, and that's what the article and that's what Dr. Clancy says they did, that they did look at cooling time. I think they looked at it in terms of quartiles, I believe, the bottom 25%, the next 25%, and so on, and they found no significant relationship. So in theory, could you take that data set and massage it in different ways and see what you found? Yes, probably so, if you could validate the data, if you could figure out what the different fields mean. But you would also have to control for all kinds of other variables. And this was, I think, Judge McKee's point, that it's not just a matter of lining up one variable, cooling time, with another outcome, and then seeing whether there's a relationship. A statistician would have to go through this and control for all kinds of other things. Who was the doctor? What was the genetic predisposition? Preoperative variables. It's a very sophisticated analysis, and that's why this is done in peer-reviewed articles by scientists. And in order to do that analysis with this data set, there would be an enormous task with the testimony to go through and identify what all these variables are. So the judge looked at that. That went to the burden, that went to the relevance of it, whether or not it had, in effect, already been done. And the decision he reached was, based on those facts, this data set, the interests that were involved, he didn't allow it. And I would note that the appellants have not cited a single case in which this court or any court of appeals has reversed a district court on an abuse of discretion on a subpoena issue like this. Rule 45, you mean? I don't think they've cited a single case of a reversal of a decision like this under the abuse of discretion standard. We understand your argument, and I do want to hear again from Mr. McClellan. Thank you. Thank you, Your Honors. I have a number of answers, starting first with Judge Garth's question with respect to record causation. In the letters that were written, if you look at the appendix, page 631. 641. 631, Your Honor. 631. 652. 652. And 660. And 660. And if you look at the colloquy of counsel in the argument on page 825. 825. Yes, Your Honor. And I would emphasize that Judge Shiller emphasized to us that he did not want these things re-argued or asked for reconsideration, and yet we persisted in continuing saying it's a causation issue, Judge, and we did it several times. With respect to the method that Dr. Hannum used, if you look at the appendix on page 47, he discusses the mortality data, and he shows where he starts with the Delaware data. He then says, I removed from them the above-listed defects, which he went carefully through, that would have confounded his analysis. He ended up with 89 VSD repairs. Of these, 89 fixed died. He then computes his rate. And I would suggest to you that if, in fact, the criticism being offered of this approach were used, the two published articles that these authors used would be disallowed. It would be unreliable. They published articles on the use of a drug and trying to assess the effect on children with congenital heart defect, which all 13 of our clients have, and they reached conclusions without trying to be as specific as they are insisting that we need to be. Well, but they did, in trying to find a loophole, I guess you're referring to, in each of those studies they did go in and try to find correlations, and as I read it what they concluded was that the primary correlation was to things like medical history, genetic predisposition, pulmonary problems, but they excluded from that conclusion the VSD, which is very contrary to the very conclusion that you need to get to the correlation. The cooling, Your Honor. They excluded the cooling. And very significantly, this shows the danger of the argument from a public policy point of view of a researcher's privilege, because basically they're saying, believe this fact because we say so. Science pursues knowledge by having other people evaluate and criticize what you've done. That's all we were attempting to do here. Now, the other point that's very interesting is the reason you could not do a prospective study, if you think about it, if you were on an IRB and somebody came to you and said, Dr. Norwood wants to cool babies in six minutes, other people want to do it in 16 minutes, let's do a double-blind study to see whether or not it increases the risk of harm, the IRB would probably say, there's enough known in the literature about the dangers of that. We can't ethically allow that to occur. Now, if that's true, that makes the need for this top data even more compelling, because it's the only data available. We can go back, get the information. If you look at what Dr. Groin says in his declaration, basically what he says is, you'll need my help. He says the original CFRs aren't there, but there's an alternative method of doing this. If you look on page 706, and you'll need my help to help you. I thought he said that because he's a biostatistician, and he was concerned that he'd bring somebody in who wasn't a biostatistician to do the analysis. He was concerned that he'd do the codes better than anybody else, and that he could do the most efficient job, but he said he was willing to do it. He also said that he could redact the names, but it would take 40 or 50 hours. He said, we'll pay for it. And in light of that, in light of the fact that we needed the information, it was available, not disputed, the court then still ruled that we could not get the information. Your Honor, the Hannum testimony, I submit, is still more on point in the way in which he went about it, because he used the RAC system. Yeah, but the RAC system, I was initially troubled by it. Now I'm both troubled and confused by it. There are two problems. One is comparing the Delaware mortality information against Dr. Millward. And the other problem, which I think is dramatically exponentially greater than that, is bringing in the website information from Boston, which I was even surprised to hear you making that argument, because there's nothing in the record that I can see to find out how the information went up on the website. It's not even really, you can argue it's a business record, but I doubt there's a foundation for showing it's a business record. What is it about the website information from Boston to suggest that it in any way can be used as significant comparison with Dr. Millward?