Dr. Tom Jaksic - Glimpsing a Bigger Picture; the Vermont Oxford Surgical Collaboration

I'll just check. Ladies and gentlemen, welcome to Grand Rounds who may wonder why I'm torching you again when I tortured you on December and the answer is that someone cancelled, so they put me here to talk again. And it's always a pleasure to present things to you. And what I'd like to talk about today is glimpsing a bigger picture, the Ramanox for the surgical collaboration, which is something that we've been doing for close to two decades. And the objectives of today's talk are first to discuss the problems with data acquisition and interpretation in neonatal surgery, tell you a bit about the Ramanox for network, discuss necrotizing intercollitis as an attempt to see the bigger picture, and finally conclude with some new studies and developments. This is Sir William Thompson or Lord Kelvin, as he's better known. He's known for defining absolute zero. He's a bit of a patron saint at MIT. And what he said is if you cannot measure it, you cannot improve it. And this is about a century before Dammling and his Toyota Acolytes adopted this in the commercial stream. Unfortunately, in neonatal surgery, we have a marked posity of quantitative data. This is the inspirational picture of William Lath and most of us are only two, three generation separated from Dr. Lath, who is the progenitor of academic pediatric surgery. And unfortunately, we all as pediatric surgeons suffer from several problems in data interpretation acquisition. The first is you can't see the wood for the trees. And that's an old English saying. And what it means in modern English is that you can't see the forest due to the trees. And as we go from challenging case to challenging case, we're going from tree to tree. And it's hard to rise above that and see exactly what's going on. The other problem in pediatric surgery is the tyranny of small numbers. Pediatric surgery is a compendium of rare diseases, particularly neonatal surgery. And HB, Steve and I were discussing the problem of billiardia treesia. The last two billiardia treesia is that I did, or at least helped with, HB pointed out at long-term drainage. And perhaps I should just do those from the Hiron Inn. Of course, that fails to take into account the five, six previous failures that I had. That's our problem. We only see these few cases. And it's more broad than just surgeon to surgeon. The US News and World Report, in fact, overweights rare cases in our results in rare cases. And again, that's the tyranny of small numbers. I'll say something nice about the US News and World Report at the end of the talk in case anyone's keeping a score. And the last thing that I think is the most subtle aspect is the consortium or registry trap. When we take a look at consortia or registries, by definition, they tend to be interested expert centers. As I was stilling out the multiple choices, quizzes that we do week to week in pediatric surgery for CME, they arose a question arose about how could a blast happen? And they pointed out that the COG results for hepatoblastoma are far superior to those of the general results of the United States. And that happens in every single disease that we generalize consortium or study data to the broader population. We take a look at studies how we should construct them at the top of the pyramid or two. And cohort studies, which should be prospectively collected, are extremely powerful. They tell us what's going on in the real world and the larger the cohort, the better. Randomized control trials are of course cohort studies where one to two variables are altered in a very specific fashion. And we get a binary answer, things of benefit or they're not of benefit. All randomized control trials should then be followed in a cohort to see what happens in the real world. This is Jerry Lucy. Dr. Lucy was an inatologist. He's passed away at the University of Vermont. He has many brilliant accomplishments to his credit, not the least of which is the establishment of the Vermont Oxford Network. The Vermont Oxford Network is a grouping of interested neonatal intensive care units that evolved organically over decades. There are now over 1,200 centers that report their data collected prospectively. The data are amalgamated by clinical staff. They're not administrative staff. And the data are then processed, validated, and analyzed in a central nonprofit foundation in Burlington, Vermont. One of the most amazing things about the Vermont Oxford Network is that it accumulates data on greater than 90% of all very low birth weight neonates born in the United States. Dr. Lucy parenthetically was the study partner of Dr. Hardy Hendron when they were both at Dartmouth College and trying to get into medical school. And I had a chance to meet Dr. Lucy and he was very supportive of surgeons. So what is the Vermont Oxford Network surgical study? Well, in 2003 I was invited to be a contact expert for a quality improvement initiative at the Vermont Oxford Network. And the CEO at that time and who continues to be the CEO is a neonatologist named Jeff Horbarr, who is brilliant. Another neonatologist Roger Saul is the scientific head. He said the reason that he wanted me involved is he wanted a surgeon with a good head for numbers. So at least I impressed him and I had a good head for numbers. And then since 2010 we've had the wonderful assistance of Bearing Modi who has been a stalwart in this effort. There have been multiple statisticians involved. The current statisticians are Erica Edwards and Kate Morrell who are also superb. And then there's a whole cadre of people and some dogs in the office at Inverlinct and where we do our work. All of this is based on prospective cohort data collection with surgical fields. The Vermont Oxford Network had made a decision to put surgical fields in and then asked us to validate those fields. At least vet them is a better word for it. So this is a little bit like farming. You put in the seeds of these fields and then collect the data many years later. There are three data collections that are done through the Vermont Oxford Network. First are very low birth weight neonates. This is almost a population based study. Perhaps the very best type of cohort study one wishes to have. Then there's a big baby database, the aspect of that that is unfortunate or fortunate depending how you look at it is of course that not all big babies go to neonatal intensive care units. Those data need to be looked at in that light. All babies in the Vermont Oxford Network are followed to discharge home even if they go out of Vermont Oxford hospital and they're considered to be survived or alive if they are alive in hospital in one year of age or if they're followed to discharge home. There's also a third data collection which is a subset of extremely low birth weight neonates. Those are neonates less than 1,000 grams that are evaluated at 18 and 24 months for near logic and other salient outcomes. So let's take a look at the model of necrotizing antarclitis. Jeff and I initially had a bit of a debate as to whether we should look at procedures or diagnoses and we settled on looking at diagnoses and the first diagnosis we want to look at was necrotizing antarclitis. It's the most common guess for intestinal surgical emergency in the neonate. Present softened with a morphous picture with a little bit of speckling and the right lower quadrant which is either stool or perhaps nemetosis. The next day you see a funny shadow in the middle of the abdomen and you think wow this could be the football sign. This could be perforation in fact. It is and then you know for sure that you're dealing with necrotizing antarclitis that requires surgery. This is something I wrote a while back and this is the way I view necrotizing antarclitis. Necrotizing antarclitis is actually a numbrellid diagnosis for a group of similarly presenting largely idiopathic diseases that affect the neonatal intestine and we need to talk about specific subsets of neck and we need to consider those separately. So how do we define neck in the Firmont-Oxford network? There are clinical and radiologic criteria as well as findings at the laparotomy in postmortem. So all of these are amalgamated. You have to have one or more of the following clinical sides, bilis gastric aspirate, amesis, abdominal distention, occult or gross fecal blood with noinatal fissures and one of these radiologic findings. So it's and so nemetosis intestinalis, portaldinas gas or pneumoparicinium. And if you have that amalgam, then you have necrotizing antarclitis. How do we disambiguate neck versus sip? And we'll talk a little bit more about that. There are two related diseases, some things that they are the same and the way we do it is at laparotomy. So when you see anything from the Firmont-Oxford network pertaining to sip, it's laparotomy proven sip versus laparotomy proven necrotizing antarclitis. This is necrotizing antarclitis. We all have seen it. Well, we can recognize it. Sip is a little different. Necrotomy antarclitis was first described in New York in the 1960s. That's when we started to save premature neonates on a quantitative scale. Sip was then described again in New York. It was in 1981. And these are localized perforations, usually on the antimesontaric, but not necessarily so portion of the bowel. By definition, they have to involve less than two centimeters of intestine. Again, some people consider this part of the continuum of neck, other consider it, consider it a separate disease. So when we look at neck subsets in the Firmont-Oxford network, the classic neck is the neck of very low birth weight neonates. It's an amalgam of neck and sip. The sip only group is subsequently cleaved from those patients who have laparotomy, and we can differentiate sip from neck. So one has to be careful when you look at the studies. There's also gastroskecis neck, which I like to call GAP, which is gastroskecis associated pneumatosis. It's not really neck, and I hope to convince you it has a totally different survivorship and sequelae. And finally, there's big baby neck, which is really a skemened bowel of a bigger child, and it's associated with serious cardiac abnormalities and chromosomal abnormalities. And I hope to prove that to you also. So the first study we did was a very simple one. We asked, what is the incidence and mortality of necrotizing intercollitis expressed by birth weight categories? In all Vermont-Oxford network studies, estimated just facial age is a less valid predictor than birth weight because it's estimated. The two are almost the same, but every single time when you look at regression analyses in terms of which predicts mortality better, it's actually a little birth weight. So this was a study that the first fellow that was involved in these studies with me, Shemife, its Gibbons, was the lead author on. And we looked at the incidence of necrotizing. And we were able to gather around 72,000 neonates. And then when we looked at the forest, it became pretty clear. Necot incidence drops linearly by 3% for every 250 gram increment above 500 grams. This goes 12, 9, 6, 3. And if you're comparing your center where you have fantastic results and you have a 3% incidence, well, it may be because you only have bigger, very low birth weight neonates. So let's take a look at mortality. This is the mortality of babies without necrotizing. This is the mortality of babies with necrotizing. And as you can see, the highest mortality is in the lowest birth weight. When you think of it, if you ask surgeons, what do you think of baby with necrotizing? They'll tend to describe a larger baby to you. Why? And the incidence, the reason they're odds ratio, we often think about signal to noise. You know, that's what we notice. So the babies we really notice as dying of necrotizing are the unexpected deaths, 1,000 to 1,250 grams, or even more egregiously, 1,200 to 1,500 grams. So those are the babies we notice. But the fact remains, the very highest mortality is with the lowest birth weight neonates. So these are some helpful numbers. The incidence of necrotizing by birth weight, 12, 9, 6, 3, mortality, 40, 30, 20, 15. If you take a look at all necrotizing, what this should convince you is that necrotizing in or colitis in very low birth weight neonates is, in fact, a horrible disease. How about the mortality of spontaneous intestinal perforation? So if you take a look at spontaneous intestinal perforation, there are certain things that occur. Gerois for lung disease, Peyton-Duktis arteriosus is more common than baseline in the methamphetamine administration is more common than baseline pda ligation is more common than baseline. Indomethasin and steroids are more common than baseline. For some reason, the latter finding was published in New England Journal of Medicine, although looking at these data, you'd say, yeah, that's the way it should be. However, when we look at babies with neck and babies with a sip, it's hard to differentiate. Although babies with sip tend to occur at a younger age, not just stationally age, perhaps, but age, and they, in our minds, are much less serious. They may, in fact, have a pretty significant mortality. So we decided to take a look at that. This just shows you the numbers in each bar. It's a little distracting when you take a look at it, but I just wanted to show you the relative numbers in this study of 804 babies who had laparotomy and had sip. You can see, again, the pattern that the highest mortality is in the lowest birth weight neonates. But then again, there is a significant mortality even in lower birth weights. But this is a disease where we most often see in 500 to 1,000 grand babies. If we compare it to neck, you can see that neck has a higher mortality for each bar, which we would expect. It's a more severe disease. But the most important thing I wanted to take away from this is that the overall mortality for spontaneous intestinal perforation is 19%. It's very large. So when you go to a patient and you say, oh, it's great news. It's just sip. It's not really that great news. It is a problem. The other things are pretty self-evident. The mortality of sip is significantly greater than baseline and the mortality of sip is significantly less than that of surgical mac. I love telling this story because I'm a Canadian and this is a laparotomy versus primary peritoneal drainage. Primary peritoneal drainage started in Canada. I'm sorry that Dr. Weldon isn't here because he tracks carefully every time I mention Canada in Grand Rounds. And there's a town in Ontario where I grew up in the promise of Ontario in Ottawa, which is very beautiful town. There's a somewhat less beautiful town called Hamilton, which is on Lake Ontario. And Hamilton is the Pittsburgh of Canada and certainly was in the 1970s when primary peritoneal drainage was started. It was a nexus of coal and iron ore. And there was a brand new medical school there at McMaster, the Beatrix surgeons were just making inroads. They were covering multiple hospitals and babies were starting to have neck at that time. So what they thought was, you know, we've got to stabilize these babies. Let us put just a quarter inch penrose drain in these babies abdomen. Get off pressure, perhaps improve physiology, stabilize them, transport them. And they actually never wrote this as a paper. They presented as an abstract at the Canadian Association of Beatrix surgeons. And it was picked up by one of the somewhat iconic plastic surgeons there, Ziggyhine, who I ended up really liking. Morris and attending as a fellow for him. But nonetheless, Ziggyhine was a wonderful man. And he started doing this. Now why did we do this in Toronto where we had this huge hospital? The reason was that there were three hospitals that gave us their babies. Toronto General Mount Sinai, which were connected by tunnel to the hospital for sick children. And then there was a third hospital called Women's College Hospital, which we called Women's Cottage Hospital, because it was around 500 yards away. In the Canadian winter, 500 yards is an awfully long way. And what we did in that hospital is that we put in drains. And Ziggyhine started to study this. And we had some rules if they were less than 1,000 grams. We put in a drain, tried to stabilize them. And lo and behold, some of those babies that we put in drains reestablished balcony and didn't need any other surgery. So the original thought for drains was a stabilization phenomena. The indications for drains became looser over time. And there were two randomized control trials. And there's part of this Cochrane review. And this is the bottom line. So evidence from two randomized control trials suggests no significant benefits or harms of peridineal drainage, overlap orotomy. However, due to the very small sample size, clinically significant differences may have been easily missed. No firm recommendations can be made for clinicians. So in this amorphous ambient, what did pediatric surgeons in the United States do? That was our next question. Just as a side, if you're looking at a baby, it would be really nice if you could say that baby's going to just continue with medical, medical mac, or it's going to progress to surgical mac and death. Unfortunately, in this study organized by Randy Moss, and we were contributors to lovely study sponsored by the laser network, a prospective cohort analysis of babies with definitive neck. We couldn't find a single useful criterion to determine which way the baby is going. So unfortunately for our fellows, you have to round every day and figure out which way things are going to go. So what is the mortality of surgical magnetizing and or colitis in very low birth weight of the N-8? So classic neck. And this was another prospective cohort study. And we decided to limit this to US centers. Almost all of the studies I'll show you are and have shown you are limited to US centers just because we tend to understand the chaotic healthcare system of the United States better than the chaotic healthcare systems of other countries. And we wanted also to assess utilization associated mortality of laparotomy and primary per-neil drainage. So in this study, there are about 215,000 neonates. The fidelity of the data collection and the Vermont Oxford is very good. You can see that only 40 of those 215,000 neonates did not have the data field of neck status filled in. So you can see that about 17,000 of them had neck by that. Rather, the stringent definition of neck, about half or medical neck, half surgical neck, and then you can see that there were large numbers in each group. So let's take a look at these numbers a little bit in a broader fashion. So 9% of all very low birth weight neonates were diagnosed with neck. About half or a coin flip evolved surgical neck. So they got a primary per-neil drainage or laparotomy. Two thirds at that time period of surgical neck neonates had laparotomy rather than primary per-neil drainage. So surgeons were certainly not in equipoids at that time. And I'd like you to remember that because I'll show you some later data. And 46% of the primary per-neil drainage group also had laparotomy. So about half were treated in the initial Ziggy-I in fashion of stabilization, then subsequent laparotomy when they were stable. This is the mortality of surgical versus medical neck. Medical neck is the gray bars. You can see that mortality drops for every increment in birth weight. Search for neck though doesn't continue to drop. It platose at about 30%. We look at these data a little more numerically. Surgical neck had a significantly higher mortality than medical neck overall. 35% versus 21%, which is really terrible. There's a mortality plateau of 30%. So even if you have a bigger baby who's still very low birth weight, their mortality for neck is 30%. When we did multi-variable logistic regression analysis, there were only three significant things that came out to predict mortality. Birth weight, which we knew, surgical neck, which is self-evident. If you have surgical neck, you have worse neck, you're more likely to die. And lastly, primary peritoneal drainage. Why was that? Well, when we look at today a little more carefully, the mortality of laparotomy versus primary peritoneal drainage with laparotomy, so the initial way you treat people thought we should treat these babies was similar. 31% versus 34% no significant difference. Peritoneal drainage alone was associated with significantly higher mortality. Now all of these data are risk stratified for all standard variables. But I will tell you, you can't risk stratify everything. And we know as pediatric surgeons, those patients that we put in drains are often those babies who can't be moved on the oscillator and we just try to stabilize them. So they are a more ill cohort. And it gets even more complex because 27% of neonates who receive primary peritoneal drainage survive needed no other surgery. So they, by definition, had SIP or minimal NAC. So what this tells you is that in the United States primary peritoneal drainage is applied to a dichotomous cohort, either the very ill or the very well. And in the middle, you tend to get laparotomy. Let's ask another different question. If we're designing a healthcare system, when we know we have this terrible disease, what is the effect of NICU and surgical center type and transfers on the mortality of surgical necrotizing and our collitis? So in the Vermont Oxford Network, we have three classes of centers, A, B, and C. They're classed according to, with A being the least sophisticated on the availability of manpower, the availability of sophisticated ventilation. And the differentiation between B and C centers is that although B and C centers have pediatric surgeons and have pediatric anesthesiologists, C centers also have cardiac bypass capability for children. This is not to imply that cardiac bypass for children is necessarily important for NAC, but it just is a way of looking at the sophistication of the center. It's extremely enough in this three year cohort study of 4,328 neonates. B centers versus C centers for inborns had a variant mortality. So the more sophisticated the center, the less the mortality. And the data we're quite convincing, these data have been replicated by another former fellow of mine who's not involved in the Vermont Oxford Network, but he started being involved with the California neonatal registry and the same thing was found there. So it does seem that if you're born at a more sophisticated center, you do better with NAC. I also would like to point out that that's not the case with every neonatal disease. Charles Hong, who is another one of our fellows, published a paper looking at gastroschesis, and that's definitely not the pattern that we saw there. How about NAC transferred versus transferred? Those NAC transferred at a higher mortality than those that were transferred. It's hard to know what that actually means because there are two things that are going on, right? You transfer the baby, they go to a potentially better center, maybe they live longer, or maybe they were just so sick you couldn't transfer them, they died in your center. So both of those things could be occurring there. But one congruent explanation and needs to be validated further is that if you have NAC, you want to go to the most sophisticated center possible that's close by and you want to be transferred there as promptly as possible. So let's take a look at the trends and incidents and mortality of medical and surgical NAC in very low birth weight neonates over the past 12 years. Let's ask the question, what's happening over time with necrotizing and or colitis? In this prospective cohort analysis, there were 890 reporting centers. They're all US centers. And this is an impress in the Journal of Pediatric Surgery with Sam Han as the lead author. And I'd like to show you some interesting things. If you look at year over year data, you can see that those babies with medical NAC seem to have a decreasing birth weight. Also, if you look at those with surgical NAC, they have a decreasing birth weight. So what do you think should be happening to incidents and mortality? You're having smaller and smaller babies. What did we show you with the first thing? They should go up, right? So incidents should be higher over time and mortality should be higher over time. Now, another thing, as surgeons, we think that we actually control mortality a lot of times. So surgical management actually did change over this. And you can see over time, the white bars and the black bars more or less equilibrate. The white bars are primary peritoneal drainage as the initial therapy and the black bars are laparotomy. But always be cautious as to whether you are determining mortality or not. So the only foreshadowing I'll give you here. The incidence of medical NAC is actually down year over year from 5.3% to 3%. The incidence of surgical NAC is also down from 3.4% to 3.1%. What happened to more... so when you look at this, you may ask yourself, is it happening just in one area of the country as somebody doing something special? Well, the answer is that regardless of which region you look at, the orange bars are the south, the blue bars are the north, the gray bars are the central, and the yellow bar is the west. They're decreasing everywhere in the United States. The incidence is decreasing everywhere. What's even more remarkable is if you look at the mortality. The mortality of medical NAC is down from 20.6% to 16.5% again, a significant year over year trans analysis. And mortality of surgical NAC is down from 36.4% to 31.3%. And a 5% drop in mortality is a big deal in neocortizing intercollitis. So you might think, gee, must be asus surgeons using more primary parodneal drains. The only problem with that is that the surgical mortality decline was almost identical in the primary parodneal drainage group as in the laparotomy group. And one very optimistic interpretation is that now we know exactly in whom we should do drains and exactly in whom we should do laparotomy, and we're lowering fluorine mortality. But those of you who are at the bedside know it's more like, gee, it's Wednesday, so-and-so's, the anesthesiologist, I think I'd be better off. I just put in a drain rather than some higher order thought that this baby really deserves a drain versus laparotomy. So why does classic NAC, a very low birth weight NAC, incidence of mortality seem to be declining? So cohort studies can't tell you, but associations which may not be causal have been noted by us in the Vermont Oxford database. So over time, there's an increased documentation of breast milk feeding at discharge, implying that there's more human milk being fed to babies. And secondly, there's a greater use of antinatal steroids. Antinatal steroids are done for a long development, but they may have a salutary effect on the intestine. The other interesting thing is that babies now are discharged from the NICU at increased weights. So they're kept longer at the NICUs, and it may be, or an nutrition may be just more effective, and that may also play into this improved survival. Now, the other interesting thing I'll tell you is that in the Vermont Oxford Network, almost all neonatal diseases seem to have decreased mortality over time. The sole exception being chronic lung disease. So that leads us to the next question. In Vaughn, why are nearly all neonatal diseases declining in incidence and mortality? And I'd like to posit the following, and this is the corollary of Lord Kelvin's doctrine that we've discussed before. That is, if you can measure it, you will improve it. And what happens when in the Vermont Oxford Network, as you get the data for the whole network, you get your center network, and then you're encouraged to do a PDSA cycles to improve the quality of your work, which may be different from center to center as to how you improve it, but does occur. So in theory, incrementally improved systems transpire, and that is exactly what we're seeing. So in medicine, as in Toyota, if you apply this system of measurement and try to incrementally improve things, you will decrease incidence and mortality. Let's ask a different question. Okay, now we're getting these, we're decreasing incidence, good, we're decreasing mortality, good. What happens to those surviving babies? Well, in the Vermont Oxford Network, I told you there was another data collection, which was looking at extremely low birth weight neonates with medical and surgical neck. And this is a study that was done by Brana Fullerton, for which she won the Fulkman Medal at the American Pediatric Search for Association. And this graph in a nutshell tells you the bad news. So if you take a look at ELBW babies, 17% of them at roughly two years have severe and your developmental disability. What severe in your developmental disability? That means two standard deviations away from norm on your belly scales and or cortical blindness and or inability to walk 10 steps. So these are really devastated babies. If you follow these babies out over long term, you know, they will actually become more functional and humans are quite malleable. But in fact, if you look at subtle neurologic signs, they're going to be a hugely increased proportion of these babies that are going to have subtle neurologic signs. So it's bad with all ELBW babies. It's significantly worse if you have medical neck and it's incredibly worse if you have surgical neck. It's almost 40%. And that's important for you to know also. And it's just a statistical analysis. How about gross morbidity? Why can we do about this? Well, one simple thing we may be able to do is improve nutrition. And if you take a look at anisopaper from Charles Hong at what happens at discharge, 56% of patients with medical neck and 61% of patient with surgical neck are less than the third percentile weight for age. Those without neck about 36%. So, you know, if you look at ELBW babies, perhaps one of the things that we can do is just give the better nutrition, grow their brains, and try to get them better. Now, so that finishes our talk about classic neck. How about what we call neck? But it's actually gastroskeces associated pneumatosis. It's not real neck. It doesn't occur in premature babies to the extent that we've seen in very low birth weight mandates. Gastroskeces is a disease of slightly premature neonates. And the neck doesn't occur until many weeks down the line. The incidence of neck, and this is a study, a broader study that we did about gastroskeces and it's published in the journal with pediatrics. But if you look carefully, there's a neck portion. The incidence of neck in gastroskeces is 5%. So much less than in very low birth weight neonates. The mortality of gastroskeces with neck is 5%. That is hugely elevated over the baseline mortality of gastroskeces, which is 2%. And the length of stay of these patients is much longer than those without pneumatosis, 84 days versus 37 days. But this is a completely different disease with completely different outcomes than the very low birth weight neck. And perhaps we shouldn't even call it necrotizing and or colitis. How about neck in big babies? Often we say, oh, that baby has neck. And it's a full-term baby. Well, it's not really neck. Big baby neck is is Schemek-Bowl. And if we take a look at 1,692 neonates from the Permanoxford Network, grade in 2,500 grams, median gestational age of about 37 weeks, we have to say that Schemek-Bowl disease is relatively rare. I didn't put in an incidence for a reason because not all big babies go to neonatal intensive care units. So that's not really a fair estimate of incidence. The mortality is large. It's 23% if you get pneumatosis as a big baby and surgery is needed. It's only 8% in medical cases. The most significant independent predictor of mortality is severe congenital heart disease and the only other one that's of significance is chromosomal defects. Independent variables, but sometimes they can occur together. Now, this is a completely different category than very low birth weight neonates who have congenital heart disease. Very low birth weight neonates with congenital heart disease. We wanted to study those because we thought, first of all, would be interested in the mortality and the incidence of neck in those babies, but also we were trying to find a clue as to what the possible hemodynamic causes of necroticine or colitis were. We failed miserably in the second regard and I'll tell you why. So congenital heart disease is significantly increases the risk of neck in very low birth weight neonates. So 13% versus 9%. The mortality of congenital heart disease and neck if you have the two together is huge. It's 55%. So significantly higher than either alone. The two things that are associated with neck significantly are AV canal, which is a bad deal. The reason it's a bad deal is not because it's only a bad deal for the baby, but it's a bad deal for the scientists because AV canals can have a myriad of hemodynamic effects. So it's almost impossible to call out what actually is going on. And the other independent risk factor, interestingly enough for neck, although AV canals occur and try to sell me 21, it's just try sell me 21 alone. And that's been shown in other studies also. So big baby neck, a schemic bowel cardiac disease is a very interesting story. It's a schemic bowel with very low birth weight neonates. It's actually the neck of very low birth with neonates as being accentuated by cardiac disease. And this is a very different disease than what we call big baby neck. Let me tell you about two new studies. And Biren is the senior author and both of these. I'm one of the et al authors. The first one is the is tracheostomy and neonates. And the next one is the outcome of drainage versus laparotomy on your developmental outcome. This is a study where Sam Han is the lead author and it's in press and pediatrics. Jeff Horbar who is the CEO of Vermont Oxford Network and I had a discussion. I kind of want initially that we should look at diseases rather than procedures. He's a little more interested than procedures. And this is finally Biren through a bone to Jeff and we decided to look at tracheostomy. And this is really the first study that looks at tracheostomy and very low birth weight neonates with these substantial numbers. So this is a 10 year study 796 North American centers around 463 thousand very low birth weight neonates. We can tell you what the incidence of tracheostomy is. It's 0.75 percent. Around 3442 received tracheostomies. Not surprisingly these had a doubling of mortality over baseline very low birth weight neonates. They also had four times length of stay. The independent risk factors associated with tracheostomy included all of the usual suspects, ELBW, chronic lung disease, and IVH. And sometimes you ask, well why do people do these studies? And the reason that we do these studies is sometimes we find things that are really unexpected. And one of the unexpected things is that an independent risk factor for tracheostomy in the United States is being a black non-Hispanic mum. And you have to ask why is that so? If we take a look at surgical disparities and care, it's often under treatment of minorities. Here we have a minority that's over treated. And there are various hypotheses as to why that is so. And we're taking a look at it, but it's a fascinating finding. And we have to figure out why. The next study is the impact of surgical treatment, drain versus lap on your developmental disability and home health care needs and ELBW survival. So surgical MAC again, here is the senior author. And Jamie Nell will be presenting this at APSA. I'm not going to tell you what the results are and spoil their talk. But I will tell you that the findings run quite counter to popularly held beliefs. So what are the new developments that have occurred regarding the surgical collaborative and the Vermont Oxford surgical collaborative specifically? So Nell Boston's children is a full member of Vaughn as of 2020. And Buren and I sat down with her neonatologists and we carefully reviewed this. And there were many arguments prone con. All level three neonatal intensive care units by mandate of the state are members of the Vermont Oxford network in Massachusetts. However, children as a level four was not a member of the Vermont Oxford network. A lot of our patients and almost all of the very low birth weight neonates were tracked through the Vermont Oxford network because they were born in Vermont Oxford network sites, bring them to women's and BI and South Shore. But nonetheless, it deprived us of putting in data about our bigger babies with interesting surgical diseases. We broke the ice by taking some philanthropic funds that I had and paying to join the Vermont Oxford network. So we did that as surgeons. Interestingly enough, I told you I was going to say one good thing about the US News and World Report. One of the data fields now in the US News and World Report is, are you a member of the Vermont Oxford network? So I'm hopeful that the hospital will start paying for a membership of Vermont Oxford as a consequence of this. And I think that's really what we should do. I'd like to thank all of the brilliant fellows starting with Schumann-Fitts-Gibbons and going through chronologically on this column and then going to this column the current position who helped and contributed greatly to these studies. So what have we said? So firstly, large prospective cohort analyses, preferably population-based, are extremely helpful in understanding neonatal surgical problems. I think the Vermont surgical collaboration is a good example of this, of course I'm prejudice. And I think accurately measuring our outcomes is a professional imperative. That's how we are going to improve outcomes overall by doing that and if we don't do it, we won't on a national scale. And I think this will also likely become an economic necessity over time. Every man, woman, child, baby and message shoots, it contributes $8,500 to the healthcare budget. Most of this goes to private insurance companies now, but that's a huge amount of money. And it's going to come a time when people are going to say, hey, what are we getting for this money? And it would be who was to have an answer for them. Thank you. Well, I heard whisper from the front here saying that was awesome and that's always kind of the bar for Dr. Essex-Toxin. And you know, he always has his style and surprised us with something. And so Dr. Schemberg and I looked at each other when we saw the word disambiguate. I looked it up, it is a word. It is a word. So thank you, we just need something early in the morning in addition to the medicine. And I have a question before we pass this on. You've made the point extremely well about the bigger the base. Of course, the population is possible of the COVID study, the better the information. But it seems newly impossible to do this. But they did it. To have over 90% of very little birth weight babies included in it says is stunning. Now, you just mentioned that even we weren't full members of the Monastery until recently. So how is the data collection from all of those hospitals without them being members? What is the motivation and the sort of operational methodology to get the data from people who aren't getting to publish the data? Well, that's an excellent question. First of all, the vast number of neonatal intensities in the United States are members of the Monastery. And it actually says a huge amount about neonatologists as a group. This is a specialty that had to show their results. There was tremendous change really from the 1970s to now. And the Vermont Oxford really took off in around 1990. And they wanted to benchmark their unit's data to the national benchmark. And that's what they did it. It was really altruism. And all of the data collection is done pro bono. There's nobody paying them. They actually pay to be members of the Vermont Oxford. And it's not a lot. It's around $8,500 a year. But if you multiply that by 1,200 centers, you're close to $10 million, which runs the central office, which is in Vermont and allows them to do quality improvement in issue. But who actually enters the data at all these hospitals that are. So the again, it varies. So there are either nurses, neonatologists, clinical research assistants. I don't know the single surgeon that has ever entered a data field, but there may be. Perfect. Good questions from everybody. Tom, very interesting talk. And at the world level, the work that you've been doing, and that Baron's been doing with the Vermont Oxford, is really a model for the world in terms of really excellent clinical research. Could you explain with all the challenges in large databases at the world level, what are the elements? Because there's multiple pieces to the puzzle. What are the elements that really get us the most impactful collaborative studies? And the most translational studies. Could you shut some light on that? Design phases and analysis presentation of the data. How do you put it all together? Most impactful studies. Well, that's a very difficult and broad question. But I think when you design a database in this day and age, it's very important to try to have a very large representative database where you have baseline data and then give blinded feedback to each of the centers as to how they're doing so that they can benchmark themselves. Everybody wants to do the best they can as physicians and as healthcare professionals, none of us in this room want to be mediocre. And that's an, you know, an Ocelarian sense, Sir William Osler would say, you know, this is our calling to try to help these children. And that's what we have to allow these centers to do. We have to have a sufficient and robust and reliable data collection and allow themselves to benchmark and do quality improvement. Well, terrific. I think it'd be a miracle if we could extend this. We are talking about these registries and collaborative and you talk about the inherent limitations that I created a long time ago, a registry for newborn infant hepatic mangeomas and the enticement was if you submit your patient to us, we will give you a 24 hour turnaround on a world class multist for expert review, somebody looking at the x-rays and all that. And it's great people do that. But then their motivation to give us the follow-up data is, it's really hard to do. And now I have these collaborative networks that are forming the Midwest clever and now we're forming Eastern collaborative. You know, one of the studies they want to do is looking at outcomes of long gap, the soft-gill adrida. And so we, you know, we're participating in that. But the vast majority of the data is going to come from one center and there comes all the questions of the competition for whose data is it, who gets to be the author and the paper and all that. So the fact that they need enough, how does it pull this off is something of a miracle. Yeah, I think it speaks very, very highly to the altruism of physicians, which we often forget. And we often forget that you should be proud to be part of a healthcare team and that should never be minimized. Well, time is exactly out in a minute. So thanks so much Tom for all of this finished.