 Ladies and gentlemen, it's a pleasure to welcome our guest today, Professor Alan Leikman from the University of Michigan. Dr. Leikman is a professor of medicine there and the Associate Director of the Transplant Center. His long-standing interests have included the allocation and use of organs, policy issues related to that, and living organ donor outcomes. Since 1989, Dr. Leikman has been the primary physician for the University of Michigan's Kidney and Pancreas Transplant programs and served as the medical director from 1989 to 2005. His work on the OPTN as chair of the OPTN Kidney and Pancreas Committee struggled with allocation issues years ago. In fact, I learned last night that one of their proposals resulted in death threats to Dr. Leikman, showing that these are not trivial matters. Today, Dr. Leikman is going to talk to, I should say that Dr. Leikman is one of the great health service researchers in the field of transplantation. His work over decades has given us tremendous new and primary information about transplantation and its outcomes. His topic today is challenges in living kidney donor outcomes research, as you see up on the board. Please join me in welcoming Alan Leikman. So I want to very much thank Mark and Lainey and Rich and Michelle for inviting me. I've been a visiting faculty member at a number of institutions twice, but this is the first institution that I've been invited to three times. So I am especially honored. I was going to talk about another topic and I changed it the last minute because of a challenge of living kidney donor outcomes research, namely we discovered that we had a file that had mis-assigned data. So that's another problem that I didn't add to the slides, but let me give you a little bit of background here about the renal and lung living donor evaluation study. So relive, the renal and lung living donor evaluation study is U01 Consortium and it is funded and managed through NIAID, but it's also funded to a large extent by both HRSA and the NHLBI. There are three study sites, the Mayo Clinic, the University of Alabama and the University of Minnesota, studying lung living donors and there are two centers, USC and WashU, that are examining outcomes in, I'm sorry, mis-smoke. Mayo, UAB, and Minnesota are renal and lung is USC and WashU. And the Data Coordinating Center for both of these sets of studies is through the University of Michigan and the Arbor Research Collaborative for Health. So relive has six studies. Two of them are lung that would be relive 02 and relive 05 and I'm not going to talk about the lung studies today, apologies, Ed. And four of them are kidney studies. Relive 01 looked at all donors at these three centers between 1964 when the first one was done at the University of Minnesota through December 31st, 2007 and chart reviews were performed at all the study sites to try to identify who the donors were and to assess the data that was available from the medical records, which for the most part because they're living donors, reflect only their pre and immediate postoperative periods. Then there was national database linkage for vital status, cause of death and ESRD and resolution of a considerable number of discrepancies. Relive 03, which I'm not going to talk about today, was a prospective study looking at informed consent and living kidney donors but in the very, very short form what Relive 03 showed was that better informed patients later reported better health outcomes and that better informed patients obviously also reported better informed consents. Cross-section study Relive 04 involved contact with patients or with donors and all donors through whom contact information was available and who were alive as of 6.30, 2005 were invited to participate. They were assayed for renal and cardiovascular morbidities in addition to their height, weight, blood pressure cratting and urinalysis was obtained and they underwent surveys regarding quality of life. Finally, Relive 06 measured glomerular filtration rates in white donors from Mayo and the University of Minnesota who had previously had post-donation glomerular filtration rates measured and black donors from the University of Alabama and there's now a new cohort of white donors from the University of Alabama who are going to be assayed because the black donors from the University of Alabama had GFRs that were surprisingly higher than those of the white donors and we're trying to figure out if that's a characteristic of being black or of having your test done at the University of Alabama. So the source of Relive data already mentioned that there was case reviews of all of the records that we could find at the three transplant centers. There's cross-sectional study data which means that we had data directly from patients and including GFR data on minority of the patients. There was prospective study data informed consent. We looked at the OP10 and the SRTR database for data on weight-listing, transplantation and death. We looked at the CMS database for dialysis service or end-stage brain disease service of any kind and death. We matched the Social Security Death Master File for death dates but not for cause of death. We matched the National Death Index for death dates cause of death and we review the death certificates in everyone who we identified as having died and we did this to search for evidence of kidney failure or of kidney disease that might not have been captured from the other sources. And they were in the process now that NHANES has recently been linked to SSDMF and is in the process of being linked to CMS data of using or building comparison groups from the NHANES data sets. So Relive, the kidney part of Relive has 8,951 donors, 2,300 from the Mayo Clinic, 2,900 from Alabama and 3,700 from the University of Minnesota. We collected more than a thousand variables that include demographics, medical, surgical, family and social history and there was information on the pre-donation experience, interoperative course, post-nation complications and mid-term and long-term outcomes and truly thousands, probably tens of thousands of data queries were resolved. So what are the difficulties with performing studies on this scale? Well one is it's just a lot of work. We're in the 70 year of Relive to date. There are actually 17 manuscripts in various stages of submission. We are in the midst of our last no cost extension which will end at the end of June of this year and we have just received an R21 to extend the scope of the Relive investigation so Relive will continue for at least two more years. Finding agreement among a large group of investigators and methodology and the interpretation of results is a challenge because we are they and they are we and there's a responder bias. So our so Relive 01 started with 8,951 subjects. We were able to find 7,029 individuals or almost 80 percent of the sample who were still alive and for whom addresses could be ascertained. Of those 7,000 people only 2,957 or 33 percent of the 01 sample or 42 percent of the eligible 04 participated. So you know we have the question of the people who didn't respond are they representative of the people who did respond is there something about the individual that respond that they have fewer or more problems that they have you know better or worse experiences that is really very very hard to resolve because there's no way to get a comparison group from the people who are silent. So let's talk about the topic which is problems. So one problem is that transplant center records are often incomplete around available. These individuals from the 60s if their data was available on microfiche that was great sometimes it simply couldn't be found sometimes there were people who were noted to have donated with no record of there ever having been a donor surgery there are people who are noted to have donated to recipients and there was no record of the recipient having ever received a kidney. I mean every flaw or foible that you can imagine in a medical record reflects on research when you're trying to do research using this as the source. People I think don't appreciate that when data collected for clinical not research purposes that the definitions of events change and they're not only different between centers but they change within the same center over time. So what might have been called a urinary tract infection in 1970 may not be called the urinary tract infection 1990 what might have been called you know a bleed or another complication in one year may be different in another year and when we looked in general there was for every intro and post-op event we looked at if we looked across the three centers at any given time point there was between a two and a tenfold incidence in the frequency in which the clinical record reflected this particular complication. So what does this mean these are the best transplant programs in the world these are best hospitals in the world. If you if one of these hospitals had reported to you their incidence of events post transplant and you had read that in a journal you would say my gosh well it's whichever hospital is and you know if it was the University of Michigan or Michigan would be the same you know with all those thousands of patients and with all those years of experience this must reflect the donor experience it doesn't just cancel it you know if they came out as three you know separate or five separate individual center reports you would say my god there's no way to figure this out. So I would tell you there's no way to figure this out. There also is incomplete naps and follow-ups of donors and collection of data. Most donors at most hospitals get seen once post transplant and unless they're involved in an organized follow-up that really ends their interaction with their transplant center and I'll talk a little bit about the SRTR data collection in a while but for practical purposes just about everyone in this cohort had one visit with their center post op and that is the end of their follow-up. The data is not audited from clinical records for quality and completeness and I kind of alluded to that but let me give you examples of the kinds of things that the consequences of this. 107 relive l4 subjects self-reported a different race than that reported by the centers in relive. 601 relive l4 subjects self-reported a different ethnicity than that was reported by the by the centers. Trust me this continues on for just about anything you can imagine. In addition we know that relive is an unrepresented cohort. There's few of African-Americans and they're almost all at one center and obviously we don't have prospectively matched controls although I think we'll do a reasonable job with the enhanced NHANES in generating a believable control group. So let's talk about administrative databases. There's the CMS ESRD database it includes all medical or enrollees where 65 years and older are all those on dialysis. Data is absent completely prior to 1972 it's incomplete prior to 1995 when it was mandated that all dialysis patients be included in the CMS database. Claims data are only available for Medicare beneficiaries and for those of you who don't do claims research you have to understand that the diagnosis in the claims is the same whether you have the condition or not. So if you have a heart cath you had a heart cath for coronary disease it doesn't matter whether the heart cath was negative or positive that's your claim okay. Social security death master found we'll talk about in a minute but it has most deaths of those with social security numbers organ procurement transplant network and the scientific registry of transplant recipients have databases and there's the national death index which not a lot of people use for research because it's expensive but it is the treasure trove. So SSDMF restrictions death records provided by the states are no longer included in the Social Security Death Master File. So effective November 1st 2011 if your Social Security Death Master File learns of your death from the states which report it then those deaths are not reported. The reason for that was that there was a sense in Congress that records on dead people were too easily available to criminals and that you know for reasons of homeland security and identity theft that it was important to protect individuals who died the data of individuals who died from being exploited. My belief is that it's a problem that virtually never occurred. It was a solution for a basically a non problem but it has really impacted the ability to do a health outcomes research. However records obtained from other sources family members funeral homes hospitals financial institutions are still including SSDMF. So what's been the consequence since November 2011? 4.2 million deaths were removed from the database but the database has 89 million records so it's still pretty robust and there'll be one million fewer deaths added annually which is about a third of the deaths. So right now they're both the AST and the ASTS and numerous other health professions are trying to negotiate with Congress to change these restrictions but at this point this is the law. All right OPT and SRT are living donor data collection. There's now data at all before 1987. Data collection began in 1987. It was voluntary prior to 1994 and some centers but only a minority notified the OP10 of their living donors. Some notified the OP10 of some of their living donors and most didn't notify the OP10 of any. And in terms of demographics in general many of those notifications were just notifications that we did the transplant but it didn't really have names, birth dates, social security number, the kind of stuff you'd need to trace these individuals and other databases. In 1994 social security numbers became a required field in the living donor registration form. In 1999 living donor follow-up forms at six months in one year were added and in 2008 living donor follow-up forms at two years were added. However the living donor follow-up form criteria really meant that you had to submit a form. It didn't have to have any data. So most of the data points are less than 40% complete and much of the data that was submitted was just the repeated data because again most of these donors were seen only one time post-op. So you know they asked the last date of visit and the last date of visit was one week post-op and what was the cratting. Most recently it was the cratting one week post-op that's in et cetera, et cetera. So as a database this is flawed. I wish it were better. I was one of the principal investigators in SRTR for a decade but it is what it is. So you can go to the National Death Index which I said is the treasure trove. It has the death dates and the cause of death from all the U.S. states, the District of Columbia and Puerto Rico. Death certificates are available. So you can get the death certificate of any individual who you have identified as having died and you can review that death certificate and there's lots of data on the death certificate. It tells you whether people have cancer, it tells you whether people have kidney disease, even if it tells you whether people have kidney failure. And this data is not available in multiple other sources at this level or with this reliability. But it can't be used for administrative purposes. You can only use it for research purposes. And the reason for that is the NDI has a separate contract with each state and with Puerto Rico and D.C. And those separate contracts stipulate this and those contracts are binding. So it's a conundrum for doing this kind of research because deaths that were identified through the NDI alone can't be reported to the OPTN but federal regulations require that every center report every death to the OPTN and deaths that are known to the transplant center therefore would to perform this and to the transplant center to become aware of a death through this mechanism put the transplant center at risk. So the D.C.C. actually wanted to wind up binding the transplant centers to their own deaths. So if you went to those centers and said tell me who donated and died they can only give you the partial list that they're aware of but they can't actually give you the whole list because we can't let them know. Wish it were different again. All right. So I think you've probably gotten the idea that you can't go to one data source and build a good clinical database. So additional ascertainment is needed not just from one source but from multiple sources to fully identify events that are not captured in clinical records and the donor identifying data in medical records is sadly often incomplete or incorrect which I mentioned but there's also limitations on matching algorithms and I'm going to talk about that in a little bit and all these databases the OPTN SRTR the CMS social death master filer subject to error and issue regular updates with additions and corrections and it was because of one of those errors and the release not till this week of the update with corrections that I had to change my topic to talk to you guys today. So what are common sources of impact perfect matches? Well there's names. Lots of people use their middle name as their first name and initial or a nickname can be used for first name. Last names often don't match because of marriage or divorce and that's more common for women and names are misspelled and so if you're matching on names and you get a name from a hospital record it just may not match in the other databases for these kinds of reasons. Digests are transposed. So you can have one or two digits that are wrong, threes are often mistaken for eights and vice versa, leading O's are missing because they can be dropped during data collection processes and in older data women will have their husband's social security number and children will have their parents social security number. So if you were a child in the 1970s and your father donated to you the donor and the child both had the same social security number. If either one of them died there's no way to tell which one was dead unless you can get a death record that shows a date of birth that's accurate which goes to date of birth. Date of birth is wrong or not known frequently. The month of birth is wrong you know August might have been transcribed is the ninth month instead of the eighth month. The year of birth especially for people who are older where the records may not have been accurate or may not exist of their actual births can be off by one or two years and days and months are transposed 04 and 104. So we have these incredibly sophisticated matching algorithms that we use to try and match databases. We being the Kidney Epidemiology and Cost Center and Arbor Research Collaborative for Health because this is what we do and we apply fuzzy logic and the matching algorithm first looks for exact matches and then it looks for alternative spellings and it looks for name order and it looks for transposed digits of social security and it looks for transposed digits in years and it looks for you know closest match and using these algorithms which use fuzzy logic then you get a probability of a match okay. So I'll show you how hard it is with this. So you need to be a little skeptical when you read manuscripts where someone says oh we looked at all our patients we matched with to the Social Security death master file and here's the number of events we found okay you can those manuscripts shouldn't be published anymore literally but they still are but you need to be very skeptical about beliefs and information that are based on those kinds of studies. So there were 650 deaths obtained from the Social Security death master file through at least one of the two matching algorithms that we used and the first matching algorithm we used is one that is used primarily for finding events among transplant recipients and the second algorithm is primarily for finding events among people with cancer. So between these two complicated algorithms 554 individuals were identified in both databases so that's 85 percent of the deaths. 87 deaths were identified with algorithm A but not with algorithm B that's 13 percent of deaths and nine deaths were identified with algorithm B but not with algorithm A. So algorithm A you know in this case is a bit more robust in algorithm B. Now there were 734 total deaths that were identified though so that 650 that were used through the matching you know represents only about 80 percent of the deaths. So the 734 deaths that were identified 432 reported by the center so the centers were aware of 60 percent of the deaths. 650 were found in the Social Security death master file so that's 89 percent of the deaths. 686 were found in the NDI which is 93 percent of the deaths. 77 were found in the NDI but not in the Social Security death master file and this was done before the November 11th restriction so this was when Social Security death master file was its most robust and it's only by aggregating all of these individuals that we came to understand that there were 734 total deaths in the donor population and not the 434 that were reported by the centers or that if a single center was trying to do research even though this research was done very very carefully that they would have had available to matched other public databases. Then Lane this is actually pertinent to our conversation last night I forgot about the slide so there were 63 donors with ESRD so 39 reported by the centers that's 62 percent of cases. 12 were found in the OPTN SRD SRT our candidate recipient database which is 19 percent. 13 were identified from NDI with were chronic maintenance dialysis patients who were identified chronic kidney disease and that was 21 percent of cases and 52 were found in CMS ESRD database. So the question you asked me last night of what percentage of the OPTN SRT of renal failure might be captured at least from these three centers it's 19 percent. So I think the bad news is that these three centers are probably pretty good respondents it's probably worse overall than that if you went to all centers. All right so then we've been struggling with the control group and relive is in the process of linking with the with NHANES to establish control rates of events in a matched population. So there's six NHANES cohorts the earliest dates back to 1971. The data elements that are collected in each of the cohorts are slightly different and the data definitions that is what's an event are different for each of the cohorts but you know it's it's no worse than what the transplant centers are doing. So you know in that respect you know there's sort of a comparability and the other thing is that NHANES has been linked now the SSDMF now it's the new regulation SSDMF so there's going to be missing deaths more missing deaths and to the CMS database which I think is going to be completed this month and we are going then to try to match the relive cohort to this augmented NHANES cohort and we start the process we start by is we're going to screen the NHANES cohort for conditions that would include donation. So people with cancer people who were very very old people with infections people with you know severe heart disease we're going to exclude those for the NHANES population then among those people who are healthy in the NHANES population we're going to match the donors based on sex race ethnicity and history of tobacco use and then we're going to use progressive radius matching for age with moving out to up to plus or minus five years BMI plus or minus two and systolic blood pressure up to plus or minus 15 and using this strategy you know we haven't gone down to match but playing with their data we think we'll be able to match about 95 percent of our relive population. One of the things is that the match has to perform physically at CDC so you can't they don't send you a database you can make an electronic connection with them and shadow what they're doing or you can send something physically down to sit with them to negotiate with what they're doing but you can't get the data and then you get a de-identified file back so which is which is very which is appropriate to protecting the NHANES individuals and actually it doesn't really harm the research if you have the resources to do that which blessedly we do. So conclusions death and ESRD information are available from national databases but the quality of the results depend on the use of multiple data sources the accuracy of the matching algorithm and the care used to resolve imperfect matches. Mid-term and long-term morbidity quality of life cycle social and socioeconomic data are not consistently available from transplant centers and they're not available at all from public sources. The absence of prospective match controls can be partially overcome through matching like we're proposing with NHANES but they contribute to uncertainty in evaluating donor outcomes. The value of the relive studies is they're comprehensive with medical, surgical, psychosocial and socioeconomic endpoints. The cross-sectional components capture mid-term morbidities which really you can't get other than by having contact with the patients. You have to somehow get this information from the patient either through their doctor as a surrogate or from them directly. The size, scope and duration of relive allows the high probability of accurately estimating frequency of common post-denation events of identifying uncommon or late events and describing the evolution of donor outcomes over time. The quality of life studies offer donors the chance to comment in their own words and many of those comments are poignant and one of the publications we are preparing is sort of a donor in their own word comment and you know it's touching. These were people, so many of these people did this in the most loving way and their lives were changed and they expressed that to us beautifully and we prospectively studied informed consent. So let's see where we are with time. You want to look at some relive data? I got some relive data. Okay this is just demographic data. So there were 8,951 individuals in the study. Early on the University of Minnesota was the largest center, there we go, and represented about 58% of the sample for the first decade. By the fourth decade of follow-up the three-setters male UAB and UMM were about the same size, but about 40% of all the study data came from Minnesota. Females are more likely to donate. They became progressively more likely to donate in each of the decades of the studies, so in the first decade it was 51% by the fourth decade it was 58%. Most donors are white and the number of the percentage rather of African American donors hasn't changed much in the last 30 years. The donors are getting older. The median age has increased from about four years from 37.1 years plus or minus 11.9 years to 40.5 years in the most recent decade. It's not that the percentage of older donors has increased much over the last 30 years. It was 3% between 75 and 85 and 4% between 96 and 96 and 3% again between 97 and 2007, but what you find is there are fewer younger donors. So if you look at donors under the age of 30, they've halved by a percentage basis from 36% 40 years ago to 18% in the most recent decade that we studied. I don't know that off the top, but we looked at people under 17 and it's a handful. But they're less likely to be donors now, but it's a handful. I mean, it's not many. I'm going to show you, this is called a Lois diagram. And in Lois diagrams, every point represents an individual, so if you could tell there's 8951 points on this. And the middle line, the green line, I'm going to show you a bunch of these, is the median, 50% above, 50% below. The yellow lines describe the 25th to 75th percentile. This is the blue lines are roughly the fifth to the 95th. They're not precise because they're actually come from a mathematical formula, but that's the gist of it. And these are the regression lines for the highest and the lowest value. So since they're a regression line, you can have some points above the regression line and some below, but this is the regression line. So if you look at age, you can see that the median is up a bit, but the 25th percentile is up quite a bit, reflecting the fact that there are fewer younger donors. As Laney just suggested, there's sort of a bottom barrier, so you can't have much change in the minimum. You just don't have many donors under the age of 17 or 18, but although the number of people over the age of 60, as I just showed you in the percent over the age of 60, hasn't changed much, you can see that our tolerance of people who are far over 60 has increased. So it's not that the number of older donors is going up on a percentage basis, but the range of acceptable ages among older donors has increased. All right, so no surprise that fewer of us smoke in the more recent decades than in the 60s, and more of us have never smoked in the more recent decades than in the 60s. There were a number of publications around 1988 demonstrating good results among biologically unrelated living donors, and so as you can see there were virtually none in the first 20 years of the study. In the 80s and early 90s when people started to appreciate that you could have a good result with living unrelated donor, this practice began to emerge, and between 97 and 2007 at these three centers it represented about a third of all donors, and I suspect if you assay now, because in the most recent years it's been more than 40% of donors that are living unrelated, that at these centers you'd see similar. All right, and just because we're a little late in time, this is sort of the same data again. All right, body mass index. Donors are getting heavier. So the mean VMI was 24.3 in the first decade of the study, and it rose to 27.3 by the fourth decade of the study, and that's a big change, or it had changed. Well the very obese on a percentage basis really didn't change at all. They, you know, in the last 20 years they're 4%, 5%, and the more massively obese, those with VMI's over 40, you know, they're 1% or less of this of the sample going forward, you know, really across all time. The change has been that people who are normal weight, people by at least the World Health Organization, people with VMI's less than 25, have reduced from 50% to 34%, and so the increases obviously are among the slightly overweight and the mildly obese. And again, here's the lowest diagram, and you know, it shows that, you know, there's a small increase in the mean, there's a small increase in the 20th percentile, and the thing that's changed has not been that there's a lot more donors on a percentage basis who are heavy. It's that we've allowed donors with VMI's, you know, 45 or higher in recent years, and it's just a few people, but it's just different than it was before. And again, when you look at the lowest diagrams, you have to realize that this is based on the maximum value, so one individual has a large effect on the curve, on the regression. Fasting blood sugars, they are a bit higher. They were, the mean was 86 in the first decade, and it's 93.5 in the most recent decade. If you look at people who would have been diagnosed as diabetic either under earlier criteria or more current criteria, it hasn't changed much, 2%, 1%, it's about the same. If you look at individuals who are borderline, it hasn't changed much, 1% or 2%, but if you look at people who had fasting blood sugars that are less than 100, our criteria are more strict. In fact, our FPSs as a fraction have increased. So I think if you talk to people in the community, they say, oh yeah, we're taking all these diabetics, and many more diabetics than ever did, and in fact, at least at these three centers, it's not the case, nor is it the case that we've relaxed criteria about the percentage of people who we require to be truly normal. To the contrary, it's been the other way around. And all this stuff about we're taking diabetic donors is urban myth. But again, as I showed you before, if you look the tolerance among those who are abnormal, we are grabbing more people who are more abnormal than we did before, even if the percentage of those people hasn't changed much. Clusterol. So we have better control of clusterol than we did a while ago, so it's a little better. Triglycerides are getting fatter, they're a little bit worse. Let me just skip through these so we don't get caught up. High blood pressure. So again, the urban myth is that we have a much higher percentage of hypertensive donors now than we used to, so it's seven or 8% through the whole 40-year cohort. If you look at systolic blood pressures, those with systolic blood pressures over 140, the percentage basis really hasn't changed. Look at diastolic blood pressures, those diastolic blood pressures over 90 really hasn't changed. Going to the lowest diagrams, again, there's just really no trends here that are meaningful except for the tolerance of more people who have blood pressures that are more high. But again, these points here represent dozens and dozens of people. These are single individuals. Again, it's not much different. And again, there's this urban myth that we're transplanting hundreds and hundreds of hypertensive patients, which isn't the case. The same is true with diastolic. We have a tolerance for some people with higher single diastolics than once we did. But again, there's really no difference in the population. And then if you look at the blood pressure among hypertensive donors, people with hypertension, you can see that they are perhaps a little less well treated in the modern era than in previous eras. And I find that as a surprise, but presumably, if the transplant center decided to transplant them, they also decided to transplant them with a trend to improve their hypertensiveness. All right, cratonines. The methodology changed here from measuring cratonines, so they ought to be down a bit. And they're down a bit. And I don't know if I have much to say about that. These are a little bit peculiar. And I would assume that there must have been some other data that the transplant center had when they accepted the patient, and that we just happened to have the data from their database, but that maybe there was a repeat on the outside or something like that. I find it hard to believe that people with cratons approximately two would be allowed to donate. And so some people had more than one condition. So among those under 60, 62% had neither obesity or hypertension or glucose tolerance. There were 13% who had obesity alone. There were 14% who had glucose intolerance alone. And there was a 5% overlap. There were 3% that had hypertension alone. And they had a 1% overlap with the obese. And they had a 1% overlap with the glucose intolerant. And then there was a magical 1%, which had all three criteria. If you look at those over 60, you would have 6% that were obese, 11% that were hypertensive. Again, the overlap between the obese and the hypertensive is 2%. That would be my own personal demographic group. Glucose intolerance, 23% of the older with 10% overlap with hypertension, with 17% overlap with obesity, and 4% had all three and only 40% had none. All right. So thank you. And I hope that this is not too discouraging. Thank you very much. We both go by our middle names and from the database confusion. I'm not sure if that means that we are already dead or never will die. Despite the limitations of the database, it's very interesting data. Questions from the group? Michelle? What do you think that the upcoming changes in the regulations with the need for centers to report up to two years on living donors is going to do for us with respect to missing data and also with respect to knowing what's going to happen to donors in the short run? All right. Well, that's actually a complicated question. So I think the data will get better. And I think one of the reasons the data will get better is because it won't be acceptable just to turn in the form part, mark that I turned in the form. So there'll have to be some contact with donors. Donors are hard to keep track of. They're healthy people who often live very far away and have no particular motivation to be studied. And so they prove an illusive population. We had, as I know, you know, because you were there, a national consensus conference on how to follow donors and national consensus conference recommended that transplant centers not follow donors, but instead that donors be followed by professional organizations that follow people. And the model for that is bone marrow donors who are followed by an independent organization and not by either the donors or the recipient center. And they have remarkable follow-up. I mean, it's, you know, they can find more than 90 percent of the people who have ever donated a bone marrow and they contact them regularly and they update their, you know, contact information. And, you know, so I think it's a model for success. I personally think that the new regulations are a model for failure and, you know, it'll be better than what we have and it's not good enough. Now, fortunately, there's actually a member of Congress who is at least discussing a bill that would mandate the recommendations of the consensus conference and that would be perhaps good if Congress passed anything. But it would be, you know, it could be as a cost savings because having one group of people who nothing but follow people is better than having 258 groups of people who don't know how to follow people, chase them down. It's got to be cheaper. So maybe they can pass it off as a cost saving. Oh, I'm sorry. Then the other thing is donors are screened. So nothing bad for practical purposes happens to donors in the first two years that the transplant center probably doesn't know about. So is it worth focusing all the energy on basically early outcomes? And I think that's not true. I think that the focus should be on midterm outcomes. You know, do they develop cardiovascular disease as a consequence of having only one kidney? Do you know, are they at risk for, you know, ESRD and all the consequences of ESRD? And that, that, sorry, that this new impetus obviously is an arrow aimed in exactly a different direction. That was great. Just a small question. Are there data of race donors? The aggregate data showed the 10% African-Americans. What was the UAB percentage since they had mostly African-American recipients? Yeah, I don't want to make up a number, but it was somewhere near 20%. I can't remember the exact number. Pardon? Well, I mean, still small, but, you know, 20 fold higher than the other centers. Alan, what do you think will be the impact of these type of data in terms of what you will utilize as information to get informed consent, which is at least part of the reason for finding out all these data? What can you do with them? And how is this going to change living donation? So I, my belief is not just this data, but other data that I, you know, haven't reported today is that we're going to find out that what we thought we knew was pretty much what we know. And, but it will, you know, give us a more authoritative source. And in the community, there are, you know, a cadre of prior donors who very much wish that they'd had more reliable and more robust information. And certainly government regulators are pushing the community to develop more robust data. In fact, the origins of relive really come from ACOT. There's something called the advisory committee on transplantation, which asked the secretary of health and human services to do a study that could be definitive. And the secretary of health and human services then required of the NIH that they fund such a study. And, you know, this study is that study. So I, you know, whether prospective donors feel that they need better information than the information that centers tell them, you know, something that, you know, I don't know. But there is, will be data about what kinds of information people thought they needed more of, and certainly the community could respond to that, you know, if they so chose. So the question is, how did the donors get segregated in a variety of different fashions? Yeah. So the question was about deceased donors. And, you know, that's all right. And all this data refers to living donors who actually donated. Well, except for a few who we don't know that they actually donated. Any other questions? So I didn't actually present this with the idea of discouraging people from doing this research. I think people need to appreciate that it's hard to do this research. Well, what I really want people to do is have skepticism about this flood of sort of poorly done, you know, quote unquote outcomes research that really is not interpretable. And as I said, I'm surprised these things still get published. But, you know, when you look at the literature, so much of it is sort of, you know, the methodology is casual. And I just think that it's a benefit to all of us to realize that you just can't go look in your clinical database from your hospital's billing records and go throw it up against the Social Security Death Master File and make a publication that reports to tell what any kind outcome is. And that's true for any disease. And so I would just urge people again to be aware as they read these manuscripts, what kinds of things to look for in the methodology that might give you confidence or reduce your confidence in, you know, that particular publication. What do you think the role of primary care physicians is in following donors, if at all? Well, you know, most donors don't live near the center where they donated, don't have a relationship under the donation with that center. It would be a burden for them to drive to that center for follow-up. So I think that follow-up of donors, you know, was most properly done within the community in which they live. People tell donors different things. You know, I tell our donors, I mean, some people tell their donors, and actually now there's a regulation that we have to tell donors that they should get a follow-up every year, which is, and so to comply with that, we now tell people that. But, you know, what I tell donors is that I will give you the same advice whether you have one kidney or two or whether you donate or not, which is I don't know what the future is going to hold. And down the road you may develop diabetes or high blood pressure or some other condition that might affect your kidneys. And that the quality of the care that you're provided is going to be variable depending on who you happen to be your physician. And it's important that you take responsibility and common sense to make sure that you get good care. So, you know, you know what normal blood pressure is. If you're seeing a doctor and he's giving you blood pressure medications, your blood pressure is not normal, you need to take responsibility to demand better blood pressure control. Or, you know, if you become diabetic, you need to take, you know, responsibility to demand, you know, optimal diabetes control. And I said in any other condition that you may develop in your health down the road, you know, it's important that you do that. So, can I push you on that point now? It's somewhere between 10 and 20 percent of living donors don't actually have health insurance at the time they donate. So, how do you expect them to get that health care? So, the first thing is, you know, we only say those words that you should get yearly follow up because the government mandates us to. You know, I don't expect any of them unless they already have high blood pressure or diabetes when they walk in the door to develop it soon. And if you have insurance now, or you may not have insurance when you develop this condition, and if you don't have insurance now, you may or may not have insurance when you develop that condition. And, you know, I'm a big fan of the Affordable Care Act for that reason. That was what Mark's comment was to the ACA. So, Alan, thank you very much. We're at the one o'clock hour. So, thank you, everyone, for showing up, Mark. Well, thank you so much.