Dr. Michael L. McManus - Concentration of Pediatric Hospital Care in America

. . Good morning, everyone. Thanks for being here for Grand Rounds. We're happy to have Dr. McManus here to share a story and some of his important work. Many of us know Dr. McManus very well from Seven South in the thank you. He's an amazing teacher and we really love working with him and I've learned actually that his career second sort of different phases and he's been a researcher in neuroanesthesia and of course an excellent teacher and he spent a number of years of South for hospitals were developing frontline pediatric care systems and and sort of look forward to hearing what I say and sort of thinking about some I don't think about very much which is really what it takes for a kid to get from their house being sick to in our HIKU and all the steps in between and where they could go we're not go which things are really important for the care that the children get so thank you Dr. McManus for being here. Morning everybody. It's nice to be here with my friends and be able to talk about different things. This is the stuff we've been working on for the last couple of years so your guinea pigs is the first time I've gone through this there's only a hundred slides so it should be fine. You know if you're like me you probably feel like the hospital is getting a little crowded that it's it was like this a long time ago but it's like this again it's really pretty pretty much a problem and up in the ICU when I'm carrying the MIKU beaver I feel like the cop at a traffic intersection with no rules and spend most of the day trying to get people through and the feeling that goes with this is a feeling of deja vu because we went through the same thing in the early 2000s and the late 90s early 2000s and then to about 2010. Spent a lot of time then being worried about us not being able to get transports in and we were turning away in in 2000 012345 we were turning away hundreds of transports so I would take the phone and we just didn't have room and that really bothered me I started working on the endless diversion problem and Massachusetts because we diverted through transport but they diverted through the emergency room so we worked on that did different things related that actually fixed the problem in Massachusetts we can talk about that another time and I started looking at flow through the unit to see how we could optimize that and see if we could get people into the unit and have access to critical services they you know weren't getting access to did that for a while and then even look to the OR a little bit see if we can figure out how to make the OR work better and over time I've come to realize that these were all just kind of tactical kind of responses to an outside problem and we needed more of a strategic kind of approach to what we're doing and the reason it's obvious that these are just kind of temporary solutions is all the same problems are coming up all over again this is from San Francisco when I was working in Massachusetts my good friend Scott Campbell was fixing the problem in San Francisco and had the same thing at a very interesting story there got the Amos diversion problem fixed and now just a few years later they count the hours in terms of how many hospitals are open not how many hospitals are closed so it's even worse there but you can google and you can find it all over the place the this I just took from Milwaukee Sentinel here's a woman who died because she worked at the medical college of Wisconsin was having a stroke and got filtered through three hospitals before she got back to the medical college of Wisconsin and these stories are all over the internet they're not just crazy stories are really happening and a lot of states have Amos diversion test forces and they have websites and they have all kinds of things going on just trying to get people access to critical services just adults and this is despite admissions being down this is despite patient days being down if you look across all age groups across the country the numbers are down and they're down so much that a lot of hospitals are closing we see things like this all the time obstacle services closing whole hospitals going down we've seen that here in Massachusetts I'll talk about that in a second but it's been you know kind of a confusing picture because the children's hospitals are all growing you can google this and if you google children's hospital expansion you get like the zillion hits and here's the first few things that I put up there so most of the major children's hospitals are expanding big time as did we which is why some of our crowding problems went away temporarily but there's still some big problems out there and it's showing up in the EMS system other places here's a very recent paper from around Baltimore where they were looking at what happens to kids in emergency in the EMS runs and a lot of the runs they're just passing the local hospital to go into town because they don't think they can get care in the local hospital and this is something that people haven't really thought too much about despite the children's hospitals growing everybody's not okay I mean there's pressure on children's hospital too and also just because you're big doesn't mean that you're solvent and so that can be a problem too so we've got all this stuff going on it's a fair question to say what's going on with the hospital system in general and do we really even have a system well if you step back you know there's about 5,000 hospitals that are registered with CMS and they seem to be distributed in a way that's sensible you've got them up in the last gun everywhere else and it seems like they can follow the population but then if you look more closely you see that they're all kind of different they have different missions they have different ownership they have different things that they do some of them are in the same place and are frankly competitive and some of some of them are in the same place and don't have complementary services they're still big gaps in what they have healthcare historian Rosemary Stevens would argue that we really don't have any coherent health policy in the United States and that's why we've gotten to this point what we have is a patchwork of laws that kind of get us to where we are now and then we kind of have to deal with their unintended consequences and deal with their consequences so in the 1800s very late 1800s early 1900s people started with the workers compensation system we had to find a way for employers to take care of injured workers and that started around you know 19 10 15 and then it really didn't filter the state's programs so it didn't really fill the country until the 40s on the last state Arkansas Arkansas was last state to start workers come in the 20s they started the VA system the reason they started the VA system was because they didn't have a place for chronic care for events they needed places for big orthopedic injuries and disabilities in places for tuberculosis and PTSD and the regular hostels of the time which I'll talk about in a second didn't take care of that so they created the VA system separate in the whole 50s were all aware that there were big tax subsidies for employer insurance and that's why everybody who works has insurance even still and then we've had a bunch of one-off laws an example is a Tallah where it kind of creates the safety net by this mandating that you can't turn people away and then we're all familiar with the Affordable Care Act which is under you know a lot of pressure right now and as we kind of talk a little bit more about the history you can see this is just not new this is the same thing that's been going on all along so how do we get here well in the 1800s it didn't really make much sense to go to a hospital there were some some hospitals some decent hospitals now that we've heard of that that were built and there were centers of research and education and care but hospitals were mostly kind of operating as charities or as they were mainly for indigent patients they weren't really a place that you would go because people knew that if you went there you were more likely to get sicker than you were if you just stayed home and there really wasn't a good reason to go to the hospital until Morton demonstrated anesthesia and surgery started to expand and then infectious disease got figured out and hospitals got clean and so you could do surgery without killing somebody and having them diaposed operative infections so if you look around the early 1900s 1904 this is when Paul Stead was was founding Johns Hopkins and found surgical resumacies there were about 1400 to 1500 private hospitals in the country there are another thousand four-profit hospitals in the country and this is the era I don't know if any of you have ever seen the Nick virus in that the show that is a great show and that's based on Hall Stead so that's kind of cool you can check that out but the period is the same period as soon as those surgical problems got solved as soon as there was anesthesia and fewer post operative infections and then guys like Hall Stead to expand surgery hospitals started to grow like crazy so these are the 20s or actually 06 to 23 I'll show you the next slide more what happens but they grew like crazy the number of hospitals they're small mostly small hospitals but they're going to have all over the place a lot of their model was based on experience in the first world war so the American base hospitals during the first world war were very very very successful so medical communities here would get people together and they'd send teams over to staff hospitals there close to the lines and they would take care of casualties and they got good at it you know a lot of people Mayo a lot of people have said you know medicine is the only winner in war and that was the case here that people got very good at doing a lot of acute care trauma care and working on teams and being clean and doing all the things it took to really improve care a lot from where it was at the turn of the century and it was that model that a lot of these newer hospitals were based on as things boomed in the 20s okay so we had tons of hospitals grow all through the 20s we're at nearly 7000 by 1929 which was an important year during this time the growth in hospitals and beds far up stripped what was going on in the population which kind of makes sense they were behind but then it got to be growing much much much faster than the population because we're in the roaring 20s despite this half the counties well maybe not half but close to half the counties in America didn't have a hospital so even though there were all these hospitals and they were growing faster than the population you didn't have hospitals in the reach of everyone sorry no oh yeah so 1929 we're all familiar with subsequent to the Great Depression there was a huge consolidation of this industry and a lot of hospitals went under and this was despite huge improvements in medical care during this time radiology IV fluids finally down here's sulfur and people really were getting good care getting much better care than they ever did before but hospitals were going out of business because of the economics and then the series events that maybe a lot of us here are familiar with maybe maybe not but right after the First World War our Second World War rather Truman realizes that so many of the recruits were for F that we needed to do something about the healthcare system in America and at that time about 4% of GDP was being spent on on healthcare and that was less than was being spent in Europe and so he said you know we need to spend more on health and he was pushing as a lot of presidents have since pushed for a national health insurance but there was tons of pushback from mainly the AMA but also the American Hospital Association they wanted to continue a voluntary approach to living they wanted to continue every man for himself type stuff because they didn't want the government to get control of healthcare and this theme has gone on ever since and it's going on today but they vehemently opposed any attempts at a national health insurance policy and had lobbyists and had big campaigns they had campaigns and they made it sound like it was socialism and everybody was afraid of socialism at that time and so you know that wasn't going to happen but what we did pass was a Hillburden Act which allowed construction of a lot more hospitals so they were even though there was this big consolidation they were starting to build hospitals in urban in rural areas especially Hillburden Act was kind of a match so local places excuse me local places had to come up with the money to get started and then they would match it two for one so a lot of hospitals got built with this then the green light continues for growth of the industry with corporate tax breaks for health insurance as I mentioned already and then finally Medicare and Medicaid big AMA push against Medicare and Medicaid for the same reasons they were concerned healthcare would come under control of government they were right but it got passed which is also right I think but as soon as it got passed people realized that it was usually inflationary that this is going to be a big big problem and so the yellow light starts to come on to slow it down here's during nixon administration they start talking about HMOs so here's the government telling people out of practice medicine and how to organize practice medicine just as was predicted then they can't really get this to catch on so they start going at them in the courts with antitrust you know you're monopolizing care and that kind of suffer your colluding and your setting prices and things like that so that went on for a while with this dead Hickford gets started later becomes CMS to get control over how we're actually spending our money this is dead too let's be a look but I've got a mouse so some more legislation kind of creates some help in a way in the sense of funding for disproportionate share hospitals but also at the same time says oh by the way the states don't have to pay the same for Medicare is Medicaid or Medicaid is Medicare and so that cuts prices a lot then you've got a professor DRG's in the prospective payment system and then you've got all the junk for managed care 1.0 that we know from the 90s which was the first consolidation that we had before or not the first but was part of it and then now we've got that stuff kind of happening again in spades that I don't need to tell you about so from 75 to 2016 we had a second massive consolidation of the industry I'll put that in perspective in a second here in Massachusetts in the 90s we lost about a quarter of our acute care hospitals after that in this slide you see in 1999 we get what 75 or 80 hospitals now there's 65 or so acute care hospitals in Massachusetts so the consolidation continues you saw Quincy Hospital Clothes or pretty much clothes and other places clothes were over at Walt them Walt them hospital clothes so there's that kind of stuff going on and if you look back over history this is what's been going on with hospitals and you can see that the number of hospitals we have now is about the same as we had in the 1900s early 1900s and you might wonder about the distribution of those hospitals and you might wonder about the care delivered in those hospitals and you might say well the size you know these hospitals are much bigger now than they were before so you've got lots of beds but the distribution is what matters you know so you've got you know hospitals of all different sizes in different places and maybe not so many big beautiful hospitals like this in reach of everyone looking from a distance again you see like a lot of good dots but then if you take it apart these are the critical access hospitals the red ones these are hospitals that have 25 or fewer beds okay so the care is really patchy the hospitals that we have are all over the place and a lot of them are in danger in a system like this and underneath all of it is us so here are the children's hospitals first in CMS in their database but you know a lot of some of these are just cancer hospitals some of these are just orthopedic hospitals fewer just burn hospitals it's hard to imagine even though we think we can do it here from Massachusetts it's really hard to imagine how this is going to serve all the kids in America so surprise when you look in the literature that you see big increases in volume of even routine cases in in in academic centers this is from a little while ago in Washington state they see a lot of the cases moving in I'll show you some data that we have now this is a lot worse now but is this is this is this really is this a good thing is a bad thing because we like having kids get their care here you know we want to be able to give care to everybody here and so you know we think we have better quality of care if you can if we can take care of you here but it does cost more and we are probably more efficient in caring for kids than a lot of other places are but maybe we're getting crowded and maybe we're having diminishing returns as far as edificiency we think we have improved safety and outcome but really is that true for all conditions and then is that okay to balance against the loss capacity in the community because these people these kids are coming to first hospitals and they need care then and they get stabilized and they get sent on well what if they can't get stabilized and that's why the all the ambulances are bypassing them things like that and then we think it's an improved patient experience if you come here the kids get better care and we know how to take care of them and we're nice to them we're good with families and I think we give excellent care but everybody can't get in here I mean I've got neighbors that can't get seen for appointments in neurology and everything else and forget just that getting into the hospital is now hard and I can tell you from years and years working in the community there's a lot of unmet need that we just don't know you don't see it a lot of other weird oh sorry was this was were you guys hearing me before it was okay all right is this better yeah all right I'll settle here so there's a lot of other weird stuff going on you've got oh I forgot to mention uh asked Rick around lost capacity and decreased access because some of that's artificial we've got plans that have actually decreased your access it's your insurance rather than the fact that there's a physical hospital around so a lot of people are out of network and so they can't come here or if they come here it's a big problem this thing from the GAO they said that 70% of all air transports were out of network so I mean these families are based with bankruptcy because their community hospital couldn't take care of stuff now you'll say oh well you you airlift a you know a trauma patient and something like that and that's okay and we need to do that trauma care short we're airlifting everything and and I'll show you what's going on with seizures and you'll be kind of surprised and a lot of people realize that the networks as they're structured are not going to be adequate to serve the needs of children and this isn't really a new thing this paper was very recent in health affairs but people were talking about this in uh an act or in other places long before it just makes sense if you've got you know that's where they are you can't have good access for every kid in America so the AP has you know a statement and everybody should have access to hospitals with appropriate pediatric expertise but um how do you know how do you know if the hospital has appropriate expertise and how do you know they really have access to it and who's looking out for this well be nice if we had a measure and if we if we had a measure then you'd want it to be able to capture the full range of hospital services right we wanted to be able to capture capture surgical things and medical things individual diseases because you can't really have just like an asthma hospital and you can't really have one that only takes care of a few things they need to kind of be full service you need to be able to stratify it by variables of interest so you can see if everybody's getting access you know race, payer, age you know you can have something where it looks like they're getting access but it depends on how old you are or it depends on what insurance you have or it depends on your ethnicity so you need to be able to stratify the variable it needs to be insensitive to volume because when you're looking at a hospital it doesn't matter if they did you know if they're able to do a few or many maybe that'll matter with quality and you need to look at quality but whether they can do it or not you have to have a measure that doesn't matter on the volume you just want to know whether they can do it it should be a direct measurement should be something that you're measuring directly and not through three different I guess this is happening type thing it'd be nice if it had intuitive meaning you have to be able to get it for publicly available data and it should be something you can't gain you can't fake you know we all know that we'll put you know all these extra diagnoses that are going on everything in order to hit the risk adjustment stuff you know you can't it has to be something that's hard so we thought a long time about this and we were influenced a little bit by Sir Brian Jarman who I met one time and asked him how he came up with this and he said mortality was the only thing I knew they couldn't fake and so his way of looking at whether hospital is doing good job or not was whether or not you are more likely to survive than if you didn't go there or something that's about that simple and intuitive that you if you lower your mortality risk because you went to the hospital that's a good thing it's not like the 1800s where your mortality risk went up so we were looking for something like this we thought how do we know if a hospital can take care of somebody just look and see if they did and we were looking at hospitals this way every hospital has a bunch of emergency visits and then they have patients that need admission for inpatient care and then if they can't do that they transfer them or if they don't know what to do they transfer them but there were patients that have emergency visits that are maybe you could even handle a lot of those in an outpatient freestanding outpatient clinic but then you've got patients that need something else that you're not going to be able to do there and that's really the essence of what a hospital is and that's what you want to have access to and that's what we were trying to measure so we were looking at the ratio of the number of people that were admitted to the number of people who needed to be admitted that makes sense? So we had probability of admission and I love this measure because it's we can talk another time it's very robust and there's a lot of ways you can use it and I'll show you some of that now we define capability and this isn't maybe a great word because the hospitals get all angry when you say they don't have capability or whatever but we define capability as this this fraction summed up for all conditions so if you are a highly capable hospital you admitted into a care of every condition and if you weren't very capable you transferred those things and then we just divided by the number of conditions to get an index for the hospital it's kind of an average and this is what we use for most of the stuff I'll show you pretty soon in the hospital capability index. Looking at this way if you admit everybody who I'm sorry if you transfer everybody who needs admission because you don't do it there if you there are some hospitals and that's just pretty much like that the hospitals don't have pediatric inpatient care will have you know a HCI of zero for kids because you can stratify it by age so the HCI for kids will be zero and that's a lot of hospitals I'll show you but if you're here and you're admitting everything that needs it then you're close to one and when you actually do the numbers and you can't distinguish a transfer that's a retro from others sometimes it's a little bit less than one so you get numbers that are like a point nine and really most children's hospitals are numbers that are kind of higher than like point seven point eight they don't transfer very much. You can look at an entire system by summing up and averaging the hospitals and you can measure regionalization for a condition by just one minus how much of this is going on right so to make that kind of make some sense in a very regionalized system most of the hospitals in the system don't admit they send somewhere and this number goes to one because this other number goes to zero which I have my pointer but I'm going to get it and if all the patients are admitted the first hospital they visited realization is zero because this approaches one so without numbers it looks like this here's a unregionalized system where everybody comes in they get admitted and they don't get transferred here's a very regionalized system where most of the people come in and get transferred and this guy did get admitted so it's not one it's point nine nine or whatever does that make sense? So we wanted to see if this worked. The first thing we did was looked at hospitals we knew so we looked at Massachusetts and the Department of Public Health has a we all have we all know what the hospitals are like here because we work with them all the time and so we have a good intuitive feeling for what they are but Massachusetts also classifies hospitals according to academic teaching, disproportionate share and regular community hospitals. The academic designation is a little kind of fuzzy but basically it means you're doing a lot of research and you're big and teaching means you've got either medical students or maybe some residents so Cambridge hospital would be a teaching hospital and then we are a mess generally others are all academic hospitals and what you find is what you think with our new metric, the academic hospitals are where you think they would be very very high, the teaching hospitals are kind of in the middle and then the community hospitals are basically the same and it didn't really matter whether or disproportionate share or not. Just probably good. And then we looked to see if the regionalization metric worked and looked at different diseases and the diseases that we think are regionalized, they don't get cared for everywhere and they do go to specialized centers, cystic fibrosis, yeah and things that are everywhere diarrhea, asthma, yep. So we looked at this against the different conditions that we're trying to regionalize here in Massachusetts so trauma, cardiac care, obstetrical care, things like that and pediatric care even but tell you about that later but yeah I fit all the things that have point of entry policies through EMS to get people to a center are more regionalized or highly regionalized compared to the other things. Another thing we realize is that when you start looking hospital by hospital kind of an interesting patent emerge that is also captured by the metric. So these are some of these I really kind of explain but the capability numbers that I'm telling you about are here you know zero to one okay that's the probability of admission for 226 CCS conditions so HCUP lets you aggregate ICT 9 Codes into conditions so these are 200 conditions and what you have is a fingerprint of what a hospital can do. So here's a highly capable hospital where everybody's being admitted and is kept there maybe a few gaps maybe they don't do burns here or maybe something like that maybe they never showed up but pretty much everything is taken care of and it's a high capability number and here's hospitals in the middle of nowhere that doesn't really have many admissions and doesn't keep many things and transfers a lot and there's big gaps and then here are different hospitals three different hospitals that are the same size and the same volume but they have different capabilities and you can capture that and the numbers come out different and that's key because what policymakers will do they say oh well there's a hospital in your community but what does it do can take care of my problem you're insurer will say oh there's a hospital in my network well can it take care of what I have this is what we're trying to get at so we vetted this was this was too too we say we vet with the they're not nerds but we vet with people that know how services research to see if we can get you know if this is passes enough with them to do this is the only thing I've reposted in this journal but this is why so they would see if you know it makes sense and they said did we went to their meetings as I said it did so they were okay with that so we felt comfortable applying it to kids and seeing what happens and so if you take these are these are box clasps of all the hospitals in Massachusetts over time adults and kids grouped by eight or not grouped by but but their HCI's so these are the HCI's for all the hospitals in Massachusetts for kids and for adults and you quickly see two things two separate systems most hospitals take care of most adult things most hospitals don't take care of most pediatric things and it's getting worse going down it's been going down for a decade we were interested in where is it going down and look specifically at different types of hospitals what we found is what you think is that it's not going down in the academic centers but it's going down everywhere else so what does this actually mean well what things are going down the things that are being more regionalized are the things that are routine they're not it's not that the sick are more complicated crazy diagnosis patients are now you know dominating the pediatrics landscape and they're all coming here these are fractures these are abdominal pains asma this is this is you know fever those are the things that are now being sent to tertiary centers that used to be managed I can tell you for sure used to be managed in community hospitals look for a long time to try to figure out what kind of visualization I'll show you a really a better one later on but what kind of visualization would kind of illustrate the differences between the adult and pediatric system and the first one we came up with is this one where we're saying okay here are the number of hospitals in Massachusetts that will take care of 50% of the patients who present with any of these diseases and these are adults and these are kids so an adult is looking adult with fluid and like disorders is looking at a hospital system where there's 60 hospitals or more that can take care of them and the kids are looking at half that and it gets worse as you go down to all the other conditions so the chances that you're going to get care for you know some of these things like epilepsy I'll show you in specifically there's like 10 hospitals in the state they're going to take care of them. Okay Massachusetts had health insurance before everybody else has awesome academic medical centers is rich so it's probably different from other states that's what we did this they're not we looked at California, Florida and New York now we've looked at a bunch of other states and the story is the same all states and when you look at the decline the decline is pretty parallel in all the states they're all declining about the same rate same way these are these are macro effects are national these are not what's going on just in Massachusetts. You wonder well is regionalization the same are the same conditions regionalized in the same way in every place and yeah they are you've got these paraplocks where you look at each state against state that we looked at and these are the conditions and you're just plotting the regionalization of all the conditions and they're all straight lines very little scatter it's pretty tight when you do this the stats they're pretty much the same so it's not like in one state they're you know not regionalizing asthma and another state they are everybody's doing pretty much the same thing. We picked New York since all the states looked the same we picked New York as an example and this is these are the conditions that are becoming the most increasingly regionalized over time abdominal pain, tonsillitis, tonsillitis, asthma. So what does this mean to us on the ground? Well we looked at abdominal pain and appendicitis first because that was one of the things that was most widely treated before still is but was changing the most and it's also important because with you guys no better than I all that's going on with the Children's Surgery Verification Program which is great but part of that is it's envisioning a system where there are level one two and three centers and the one centers need to be you know kind of helping the two and three centers so that care and and health in the community is as good as it can be and so the centers are not expected to be doing the same things but they're supposed to be kind of looking after each other a little bit I think and we wondered kind of well where are we for just plain apes and we looked at four states the same four states for this one now looked for more had a lot of encounters a lot of transfers and what you're seeing is that everybody's seeing abdominal pain and appendicitis and they're all transferring them to high ACI centers they're all transferring them to places like this and so the delta that you're getting is to these huge complex centers from like any hospital to complex centers and the soundbite sickle along with this is that have a bowel, abdominal pain was going to 20% of hospitals 80% of transfers went to 9% of hospitals things like that you can see by age how it varies people worry with the verification program that little kids aren't getting done in academic centers and the places we looked at they pretty much were but you know there are exceptions for sure 60% of all of them were done in places with an HCI greater than point seven and point we're like point nine but the more is was like point seven five or something so you know this is these are not children's hospitals but they're good hospitals and then interestingly among all the transfers and there's tons of transfers among all the transfers have required surgery so the community hospital docs were you know they're right they needed to get something and 20% required some imaging so that probably wasn't available and they wanted to kind of make the diagnosis through that but a third 30% were just sent home they just like came up to the king got transferred in just got sent home nothing got done for them in the emergency department and so I think that's probably an opportunity for telemedicine for us to support some of these other centers not centers from these other community hospitals and that promised I would say a little bit about seizures looked at eight states for seizures and what we find is that they are seen in all kinds of hospitals but they're really only admitted to to the high capability hospitals so what does that mean like how said a problem well I don't know if any of you are familiar with the Dartmouth Atlas but the guys in Dartmouth did a long time ago as they set up hospital service areas and hospital referral regions based on where people get care so in the hospital service areas those are areas who are 80 percent of care is delivered in that area to hospitals in that area and they realized that when they did that that didn't capture everything because these were hospitals that often had to refer also for certain other things so they had to refer to mainly for cardiovascular care and neurosurgery care so they redefined hospital referral regions as the bigger regions that these smaller hospitals refer to for big things like neurosurgery care and cardiovascular care and this is Medicaid so this is Medicare rather this is all adult but you see here's how many hospital service areas there are and here's how many referral regions there aren't many referral regions okay so when we looked at seizures we found was that here the hospitals the acute care hospitals in the state here are the number that saw seizures almost all of them so pediatric seizure rate in these states was about four per thousand per year the period rate for a year four per thousand is the rate the number that admitted that actually admitted to seizure patient were a lot lower the number that transferred was like about as many as saw them because even if some of them admitted they still would transfer a lot of them and the number of receiving hospitals was about as low as a number of Dartmouth hospital referral regions or lower I had to put less than three because HKAP doesn't let you show something less than two just so you know and this means that seizure care is for kids is as highly regionalized as neurosurgical care and cardiovascular care in adults that's how regionalized it is right now and when we're looking at what happened to these kids they're not crazy seizures these are two-day length of stays in all the datasets every all the states across all the years we looked at there are 27 deaths okay these weren't like crazy high-end things these weren't even intubated a lot of and and half of them came in the only diagnosis they had the seizure so if a regular kid has a convulsion for whatever reason febrile seizure whatever maybe not febrile seizure a lot of these a lot of the the community houses are important because kids were coming there and they were getting care with seizures and they were getting sent home but if they needed anything else there was any uncertainty they had to be packed away in some times like helicopters right because that's what would happen that's how big these referral regions are so pediatric hospital care is rapidly consolidating the result is an entirely separate system for kids then for adults the pattern seems to be similar across the country even older kids with common conditions are now transferred for care and the system that is critically dependent on a handful of centers so I worry that we have decreased access especially to critical hospital services we're getting crowded because we're taking care of things that we really should be able to take care of somewhere else there can't be any surge capacity I get something else happened where would they go I wonder about our regional disaster preparedness because when I look at this we've only got a couple hospitals that really take care of kids in really really wide regions and we need to have some way if anything happened there if they got full what would you do in the community there's a real cut in public health by not having pediatric and community hospitals when we were down in South Shore we were the ones that taught the paramedics how to integrate kids you know we were the ones running palace programs blah blah blah there's a million things like that so if pediatric completely dries up in community hospitals it's a big problem and then I think there's probably a lot of potentially unnecessary transfers that we can probably manage some other way so we need to have better communications that are policymakers so they understand that this is an issue we need to have a better understanding a real time understanding about these networks are working and where the pages are going where they're coming from and when I think we need to know a little bit more about what's going on with specific conditions and then we need to develop once on that understanding we need to develop better coordination cross hospitals so we're working on that so here's what we're starting let's see now 7.44 so I have six minutes so but we put this together to try to show people what we're talking about people that aren't familiar with health care and what's going with kids and so this is kind of like the other one that I showed you what it is is that each box here this is pediatric this is a problem conditions are arranged like this by increasing availability of care okay so the more available something is in the community higher it is and the hospitals are arranged by increasing capability so the most capable hospitals are over here and then they're colored by your HCI so how capable are they in taking care of that condition and what you see is that most hospitals take care of most adult things and that most hospitals do not take care of most pediatric things in fact most pediatric things only go here only go to a few places I'm going to try something and see the works but Sandy has been using this showing it to people to try to explain what's going on with pediatric versus adult care and why you know we need sports children's hospitals etc but what we're doing with it now is we turned it into a tool for well policy researchers so this is the website and you can go probably should I'm going to go to Arkansas but you shouldn't go to Arkansas because it's in one hospital it lets you go and look and see by condition who's doing what so so a state rep can go and look and see oh you know nobody's taking care what's the most common condition here your any tract infection is being taken care of in most places in Arkansas and then you know what's not cardiac disreasoning is what that makes sense but you can look through here and you can see what's going on with your system and each state is a little bit different now let's see if I get back to where I was sorry I knew that was a risk you're gonna like those two but okay here we go okay the next thing we're looking we're creating are these procedural heat maps where we're looking at all the procedures all the surgical procedures that are done in any state and then arranging them in heat maps like this where these are the procedures and these are the hospitals and this is Massachusetts and I wonder who this is and this is frequency of cases so you very quickly see where anything is being done how many are being done and what we're going to do with this is look and see how many places could potentially be level one centers in different states because you need a thousand cases for that we're also doing geospatial analysis we just kind of started doing this and this is fun we take a state like Massachusetts and these are ED visits for a condition those are the admissions those are the transfers out these are the transfers in and you can look at them all like this you can say oh well it looks like seizures are not being admitted hardly anywhere but are being transferred to just a few places and you can see that right away but they're being seen all over the place here's asthma I think this is it I'm pretty sure this asthma here you say well guess what everybody's seeing it a lot of places are admitting it some places are transferring it you know into like here this is here they don't admit it but they're not going to transfer it to the hospital next door that does admit it they're going to send in here and there's this kind of stuff all over the place this is this is superficial laceration so these are the families that would like their lips sewn by a plastic surgeon and you see them all over the place and you get these transfers but then they get transferred to just a few places you can look very quick these we have this so these can be generated on the fly very quickly so we can look at any condition anytime and all the stuff here's Maryland just for fun I've usually did some states and they're pretty pictures Maryland Florida New York and you can see the regionalization of all kinds of conditions and what we're doing with that is we're trying to understand the geography like how how far do you have to travel what barriers there might be and what are all these centers then really doing the same things here's Kentucky that's some rural states and then the last thing we're doing 748 is formal network analysis so it doesn't project this kind of projects we can take rather than just looking at the geography we can see who's transferring to who how often and how those connections run and it's too bad those others didn't show because it's very interesting you can see in Massachusetts that for asthma there are certain centers that get it for seizures it's a lot more complicated so if you're trying to fix this problem and arrange what to do it's it's hard to do a one-size-fits-all law because because different conditions have different behaviors and there's a good reason for that is it has to do with what the different hospitals can do but you have some very high capability hospitals that aren't getting very many transfers it turns out and you've got other that are getting probably too many so we could fix that a little bit but we've done California and Florida this is seizures in Florida and New York and then the last thing I'll tell you about is some stuff we do with unsupervised machine learning we're looking to classify hospitals see if we can find an automated way to classify hospitals and then let EMS providers and such know where to take certain conditions and the early stuff does seem to work this is Massachusetts hierarchy of clustering M.C. kind of comes in so the green ones are the ones you'd think would be kind of grouped together so that's us and MGH and Tuffs and BMC and base state and UMass they kind of come out as one but even the subtle differences of BMC are kind of split up and then it's got these red ones here which are ones that we think of as the pediatric hospitals Winchester and South Shore and some of the others and when you've got these others some of them don't admit kids very much so I'm just to a few things and what we're using this is try to see what characteristics are reliably available in what hospitals without knowing the hospitals so it's easy for me I could do this myself by hand you know after living here for 30 years for Massachusetts but I can't do Arkansas and I can't do Kentucky but now when we automated we should be able to do everybody and you can cut it by region and all this work was done with my research partner or Bono Franca who is awesome he's a physicist programmer network specialist and all around good guy and we started on this stuff a couple of years ago and it's kind of taken off a little bit and I will leave you with this do we have a hospital system these are two definitions off the internet so you can decide whether we have a hospital system and if we are going to build a system we have to be careful all right and it is 751 thank you well my thanks for bringing us a view of healthcare quite different from what we normally think of you know I think unfortunately the bottom line on this ends up coming down to dollars and cents and seems to me that what you would want would be a system where you know urgent things can get seen at lots of centers close to where they occur and then you have organized you know transfers to the kids that need higher levels of care than are unavailable in the smaller hospitals that can take care of everything and yet you know looking at your sort of next to last last diagram of where people get sent that's that's clearly as you expect built around you can private hospital systems when you look at the the California one particularly so it wasn't necessarily where was the best place to sit in the kid was where was the place in our system that you can send it yet that may not be necessarily wrong but in this time when there's a lot of contraction of dollars and cents going into the healthcare system how how do you get it a system designed where kids can easily get seen for urgent things closer to their community and yet there's a system whereby they get sent to the appropriate hospitals and not the ones just linked in the system yeah I mean that is the question right that's what we're working on I think fundamentally it's going to involve involve cooperation between and among hospitals and it's a kind of level of cooperation that you know we're starting to see in some areas but even that's not great you know I think we're not thinking in terms of delivering care across multiple institutions we're thinking about you know each place separately and that's been going on for a long time and I think that you know for some of these conditions a phone call to somebody who is on your team who knows more about it than you do can be the difference between a transfer and even in a mission we've seen that in a cake program for sure but but I think it's going to take looking at specific conditions first to see what exactly is going on where they come from how are they flowing through the system why are they doing this and then get solutions for those and I think many of these will group that's what we're trying to do with machine learning many of the conditions will group and so you know one solution will be okay for some other conditions too but I don't know which ones yet. Sure there are other questions for Dr. McManus Dr. Jackson. Mike that was great and very thought provoking in your slide where you showed where the actual pediatric expert centers are it kind of brings up another sense which is common sense if you take the look at that slide there's only one center that is a true pediatric center that's not in Boston which is based state yet all of the others are in Boston so you have Boston Medical Center, Mass General, Tufts that really doesn't make sense and if you look at economies of scale why do we have that and why do we allow that? We have it because historically all the hospitals grew up independently and they grew up where people were in the beginning and why do we allow that? It's this slide. I think try fixing it it's not that easy you know the levers of government are only you know can only do so much you have to be careful so I don't know how to I this is what we're working on I would like to fix it. You know Mike I'm troubled as I think about how we've often taken the specialists from outside I mean South Shore is a good example where there was certainly a cadre of a lot of expertise and then that's that's gone away even I even think of Boston Children's Rating Tufts for cardiology and GI or whatnot I mean there's so that the distribution of that expertise that would be necessary to to address these problems that a lot of different sites seems to be going the wrong way. Yeah I mean fortunately with telemedicine and stuff there are things we can do you know there are ways to go at this if we wanted to go at it if we you know we do have we've we scooped up neurology and things like that all over the place but we have our clinics down at South Shore and stuff but our clinics don't support the hospital it's a separate venture there's no incentive for a neurologist to go see a patient and he merdens you around the hospital here. Like that was terrific and your measures that you created I think are so common sense the number of times I can imagine I can remember the last decade and a half arguing in the contracting side with payers about you know why is Boston Children's better why do you guys charge more and I often said with anecdote without dollars or without numbers well if you would account the number of patients that get transferred in to Boston Children's from those other great pediatric centers and compared to the number of patients that we transfer to them but the only reason we ever transfer patients out of this hospital is because we don't have a bed or they have a serious burn or a spinal cord and they just ignore that right they ignore but now you have data to show this the the better anecdotes that how come they themselves and I've had this with you know a major payers contractingly showing up in my office in Lexington with an a Bill of Hurnia when the child had been seen by a pediatric surgeon at another institution not very far from here their scheduled surgery came for a second opinion and it was like seven months old and I said well I think your child needs an operation and she said oh I get it I said get what she said now I understand your charge for this office that is more than they charge me but we at the payer don't count to zero for the unnecessary operation and more importantly my kid who I'd be fearful of putting on a anesthesia and operation so the points that the gentlemen from me were were sort of diametry opposed right Tom said you know why do we have hospitals you know why is it that we have hospitals so close to each other like in Boston and not spread out that are sort of competing for the same thing and then Craig said well we're rating those hospitals we don't have a system right I think what you're pointing at is we don't have a system and and Bob's point that it's all about the money right so in the last two weeks I've sat with public policy leaders in Connecticut who run Connecticut Medicaid and then the leaders of a private for-profit company who run half of them Medicaid children for New Hampshire right and they're all about we want you to take care of the people you have to take care but we don't want to pay you because the region goes further than state lines so my question is as you've gone at the beginning from what you were studying earlier in your career which you described as tactical which by the way I think we're incredibly important in getting this to be more efficient right you have to accomplish something by looking at things like you know downtime and OR something like that those are really important contributions now you're measuring at system this but the question is can we like this this picture is everything do you think there is a solution I mean in the last 36 hours we had you know a federal government or these the leader of federal government saying I've got the solution but I'm not going to tell you to have the next election right until then he went and met with Senator George Lee who said no no you can't see that we lose the election if you have false promises if you were in there right and you were the head of CMS or Senator George Lee were rising to present do you think there is a solution or is our system is our free market system which has gone all the way back to the 40s of preventing centralization of these decisions not able to do this the really pessimistic way this is children's hospital so there is a solution we will find a solution and it's going to happen I don't know what it is and I think you can you can go off have cocked and you can like say here's my solution to and then have big problems without data without understanding what's happening but sure you know there's there there's there are going to be smart ways to make this work better we're we don't have a healthcare policy and you don't get a health care policy by passing a half dozen laws over 100 years but you know and maybe even we can make the free market work in a way that will help make this happen but I don't think I don't think it starts until we look and see what the heck is actually going on and stop talking about stuff we don't even know the numbers and once we know what's going on then you can say people on the face will say well I do believe maybe not in Washington but I think people here in the state are representatives or good people and they want to get so good solutions to care and stuff and so I think you start by doing pilots and solving problems and building on that so I'm not and I'm not going to give up but you know it's it's is I don't think it's going to get fixed in the capital seven questions thanks Dr. McManus for your talk I just had a quick question I wonder how much of the transfers do you think are due to concerns about lawsuits and things like that and would have you ever looked at any states that have done tort reform and see what the transfers look like there yeah that would be you know everybody asked that you know because especially in the Southern States they all asked that and I'm not really sure I will say one of the things that we looked at I was I had come across folks that shall remain nameless who didn't want to in the community didn't want to do didn't want to do an appie in a girl because they didn't want to end up having to deal with complaints about fertility issues and stuff later on if anything you know didn't work out and so they you know wouldn't do that when we looked overall at sex as far as a as a driver for transfer there wasn't there you know so so we may have you know there may be individual cases one off people doing that but I think I think the systems are kind of burned in the other thing you're seeing is that there's not a lot of difference in at least in this state between Medicaid and non-medicate because we build these systems and and we don't really even know whether somebody's Medicaid or not Medicaid we just have they're just in the system the problem that's when you touch the system problem is getting there in the first place I mean our hospitals that have pediatric care even in the neighborhoods or even in the vicinity and one of the things I learned down at South Shore was that a lot of stuff comes out of the woodwork when you put services on things that you think people had 100% access I mean I was shocked at the stuff that was coming out of the woodwork for kids and so I think that the medical legal stuff is important but that the systems that are in place are such that they kind of trumps that here I don't know about you know we're looking at the other states well Michael our time is up but thank you for bringing us to us this morning