NYSPFP Potentially Preventable Readmissions Reports Training September 24, 2014

Transcription

1 NYSPFP Potentially Preventable Readmissions Reports Training September 24, 2014 Good afternoon and thank you for joining us. We're going to get started in just a moment, but we'll start with a few housekeeping announcements first and foremost, as always. All of your lines have been muted upon entry due to the large number of attendees today, so you will have an opportunity to ask questions of our presenters at the end of their presentation. You can do so in one of two ways. You can type a question in the Q&A box, which you'll see on the right-hand side of your screen, just type your question in, hit send, and then we'll let them queue up throughout the presentation and answer them when we get to the question-and-answer segment. Otherwise, you can also click the raise-hand icon. It's a little hand-shaped button at the bottom of the participant window on the right-hand side of your screen. You click the hand, we'll see that you have your hand raised, and we individually unmute your line so you can ask your question over the phone. Today's program is being recorded, and the slides and the recording are available on the New York State Partnership for Patients website so you can refer back to it or share it with colleagues that were unable to make it today. Finally, closed captioning is also available. You can access the closed captioning service by expanding the media viewer box on the lower-right hand side of your screen and logging into the closed-captioning service. With that, I'm going to turn it over to Zana from the New York State Partnership for Patients to introduce today's presentation. Good afternoon everyone. Okay, thank you. Good afternoon everybody. This is Zeynep Sumer-King with the New York State Partnership for Patients, and thanks, all, for joining on this Tuesday afternoon. We are going to be talking

2 about the three MPPR readmissions reports that you all have been receiving throughout the course of the Partnership for Patients over the last two-and-half, almost three years now. We're coming up on the closing of the final three months of this initiative. As you all know, we're winding ourselves down, but we are not certainly winding down the actual work that needs to be done. As we look at readmissions and our overall statewide performance, we certainly have room for improvement, and I know everybody is very focused on reducing hospitalizations overall across the state with the DSRI program, so we wanted to make sure that weren't appearing as though, or, you know, you knew we were all still here and ready and willing to support your efforts around readmissions and wanted to remind you of some tools you have in your tool box. There are tons of different reports out there and data that you're inundated with in terms of your readmissions rates and other quality data. But what we try to do at the Partnership, and with your projects managers that work with you directly through the Partnership for Patients, is turn that data into usable information, because that's really the only time it's going to be useful to you. Data is great and interesting, but if you can't use it to improve, then it's actually overwhelming. So we wanted to take the time to go through, once again, the PPR reports that you do have available to you on the New York State Partnership for Patients data portal website. Your hospital-specific reports are there from the beginning of the initiative, so you can look at them sort of longitudinally to see your trajectory through this initiative, and your improvements. But we wanted to drill down and go through not only just the anatomy or the architecture of those reports but really talk about examples of how you might use the reports and the data within to impact your readmissions efforts -- readmissions prevention efforts. And this goes beyond the New York State Partnership for Patients. I think, you know, discussing how you could use this data, slice and dice it to your needs to improve, you can actually apply those same tactics to reports you have from other sources, other types of readmissions report, other types of quality improvement reports, and we want to go through that exercise, sort of as a sustainability activity moving forward. But additionally what's different about this report -- and Gloria Kupferman and Nancy Landor will talk a little bit about what sets this report apart a little bit in a moment. But I'll just say, you know, CMS and New York State move into looking across the board on hospital utilization and readmissions not just the three Medicare conditions. You're going to be, and you are being forced, to look at systemic reasons for why patients come back. And I think this report, as holistic as it is, is going to give you an opportunity to really study, drill down into those reasons. So with that, I want to turn it over to Gloria Kupferman who is going to review, as I said, the architecture, or the anatomy of these reports, and then once she goes through that, Nancy Landor is going to look at Page 2

3 more case examples of how you might apply some of the rich data that's in the report to your actual improvement efforts. So, Gloria. Thank you very much Zeynep. It's good to hear your voice. And thank you, everybody, for joining us. So I'm going to take you through some very introductory stuff on readmissions. I'm going to go through it very fast, because I think just about everybody on this call knows what a readmission is and knows why we're here on this call, so we really want to get down to the meaty stuff, which is the functionality of the reports in case you haven't been working with them, and then some of the case studies or examples that Nancy will go through just to give you some ideas as to how you can use these reports, as the last issuance of the reports is coming up in another month or so and before the end of the partnership, and so you have this arsenal readmissions data that you can be using and evaluating going forward. So what is a readmission? It's a return hospitalization. Anybody that comes back to the hospital after they were originally admitted, within a certain span of time. That span of time is up for discussion. There are different policymakers, different quality manager, different clinicians out there that may look at readmissions through, then, certain timeframes, whether it's a smaller timeframe to find those people that are bouncing back immediately or a longer timeframe at -- the timeframe determines whether or not something is considered a readmission. For the New York state Partnership for Patients, we are looking at a 30-day window of time for the PFP potentially preventable reports that we're sending you. 30 days seems to have become an industry standard in terms of the length of time. It seems to dovetail most often with other readmissions measures that are out there. There are a lot of different flavors of readmissions. The readmissions method that we are using for the reports that you get for Partnership for Patients, use the 3M potentially preventable readmissions PPR algorithm, and I'll go into that little bit. We chose that because, first off, it is the method that is being used by New York State's Medicaid program to determine the Medicaid payment penalty. So it's something that everybody in New York is familiar with. Plus, it is an established methodology that we can get our hands on, unlike what you see next, the CMS readmissions methodology, which is also obviously being used for payment penalties on the Medicare side. However, it is not open source, and it is not purchasable, so we cannot exactly recreate. It's a black box, so we can't recreate that methodology. So, in wanting to tap into something that related to payments and something that everybody out there was working with, the PPRs became the best choice for that reason, plus the PPRs do some reading out for, quote, unquote, potentially preventable readmissions, so there is some recognition within this software and algorithm that those traumas or those cancer cases that certainly were unimaginable and nowhere near related to why a person might have been originally admitted, are weeded out. Another -- I Page 3

4 guess it's not a bonus, but another point is that now these PPRs will be used as one of the key metrics for the DSRI, program, the Deliver System Redesign Incentive program under our new Medicaid waiver. The CMS hospital readmissions are all-cause readmissions, so they have some exclusions for really extraneous stuff, but they don't try to weed out as many preventables. They do try to weed out some planned readmissions, but they don't weed out as many, and so they're considered to be all-cause readmissions. And there's a hospital-wide measure out there, which CMS has come out and said in this final in-patient rule they will not ever, ever be using for the Medicare penalty. They're going to, instead, choose the second set of measures, which are condition specific, and they're just going to continue layering more condition-specific readmissions penalties on us as part of the readmissions program. It's a topic for another discussion as to what the impacts and implications are of that. We also have readmissions in the bundle payment program. With bundle payments, it's sort of built in there. When you are receiving an all-in payment for an episode of care it will include any time a patient gets readmitted. The window, then, for the readmission is however long the episode is. And there's also, which are not on this slide now, many of you have already just received data from the New York State Cardiac Data Reporting program on PQIs and CABGs. Those are readmissions rates. That's a whole other flavor of readmissions that's based on patients that you report into the registry and then they're paired up to the Sparks data, and they're PCIs and CABGs and there's another flavor of readmissions. So what differentiates all of these different readmissions programs? There's certain things that make these methodologies different. First of all, how long is the window, how many days are you going to count to look at to see whether or not there was a readmission; how you deal with readmission chains, patients that bounce in and out more than once during that window of time. Some readmissions methodology counts those as one readmission no matter how many times that patient is in and out of the in-patient setting within that 30 or 15, or whatever day window. Let's just keep using 30. So some methodologies count that as one readmission no matter how many times the patient hits the setting. Some methodologies count a readmission, then reset the clock, and the next readmission is yet another readmission. So there's different ways, believe it or not, to count readmissions. So each of these methodologies that you're seeing are going to yield different readmission rates for the same hospital and the same patients. And that's very important, because I know it gets very confusing to a lot of you. The other thing that really makes the difference in all of these methodologies is how you risk adjust, okay. A hospital-wise all-cause type of readmission can still do risk adjustment, or it might not. Okay, the CMS methodology does a lot of risk adjustment, and those risk adjustments are based on things about the patient that happened before those patients showed up in the hospital setting to begin with, looking back in their claims history. The PPR methodology also has risk adjustment that we're going to go in. Page 4

5 The New York State cardiac data is this using data that you reported into the registry to risk adjust, and then all of those exclusions for planned, unplanned, related, unrelated. So you can take the same set of data and run it through all of these different readmissions methodologies, you'll end up, unfortunately, with different readmission rates in each. I can't even say unequivocally that you would see the same trends in readmissions use all of these. We've never really done full data on it to see. My gut would say, yeah, you must see the same trends if you're reducing readmissions you're reducing readmissions. But when you layer in all those risk adjustments, et cetera, not so clear. So I apologize for the confusion out there, but readmissions, everybody has their own version. The 3M PPR methodology, which is what's used to create the reports that you have, are looking for -- 3M are called potentially preventable readmissions but they're really clinically related readmissions, and the algorithm is relatively simple, though not completely transparent. It's essentially a matrix that lists every APR DRG in the columns and every APR DRG in the rows, so you end up with this grid, and then just like, you know, those distance maps where you see -- you know, you go from one place to the other and find the box where they intersect and you see a number. You find the box where one DRG intersects with another, and it's either a yes or a no, are they clinically related or aren't they. And that's how this algorithm works. So, you know, if you've had questions, some of you have called us in the past, some of you will probably still continue to call us, you know, was this included as preventable, was it not. Usually there is a chain we have to go turn around and ask the folks tat 3M, and they'll get back to us to let us know, but they don't actually provide us with that matrix. The 3M PPR methodology does use chaining, okay. So a readmission chain is the first admission, the index admission, and then you go out 30 days from index discharge, 30 days from index discharge, and you look for any readmission since that index discharge, and if it's one, two, three, or ten readmissions during that 30 days, it's one chain during that 30 days. When they 30 days is over you start again. It's a clean slate. I'm looking at Alisa just to make sure. Yeah, and it's just a reset. So it doesn't tell you -- you don't know how many readmissions are in the chain and it's not counted -- again, a readmission isn't counted if it's not clinically related, so if you have up third case an index admission and then ten readmissions, okay, but nine of them were not related, they don't count. But you still have one, you still have a readmission chain, same as if you had all ten of them being related. So there are some global exclusions, and they're what you would expect, you know, end-stage major cancers, multiple traumas and burns, that would be, you know, the result of the story we all hear about being hit by a bus after being discharged, and that's not preventable. The PPRs would pick that up. HIV is an exclusion. Some admissions are just plain out. They're not index admissions, they're not readmissions. Same-day transfers to an acute care hospital, okay, so if there was a transfer to another acute care hospital, the hospital that received the patient first, the transferring hospital, that's not the admission. The receiving acute care hospital that actually discharged the patient to a non-in-patient acute care setting, that's the index admission, okay. So, but transfers themselves are not readmissions. Page 5

6 The observed PPR rate, the observed readmission rate is the number of chains, as I just described them, so if you have one or ten readmissions within that 30 days, it's a one for that patient over the total number of index admissions or at-risk admissions. That's the observed pure unadjusted unadulterated readmission rate for your hospital. The expected PPR rate is at number of expected readmissions for your hospital over that same number of at-risk admissions. And the way we calculate, this is the 3M methodology, the expected readmissions is using the state average. Since this is a 3M product, all of these readmissions are categorized into APR DRGs and the severity of illness levels one through four. So you have statewide, for every APR DRG and every severity of illness, a readmission rate, state average readmission rate. Then you would take your hospital patient counts, okay, and assign them to those same APR DRGs and severity levels and multiply those counts by the state average readmission rate. Those are your expected readmission rates. What would your readmission rate be if your experience was exactly average for New York can State? So we refer to that as normalized for New York State. New York State is the norm here. Okay, so that's an important distinction too. Remember, I said that the risk adjustment methodologies make a difference. Your expected rate here in New York State, comparing to you to an New York State average, means that, you know, if the average was some sort of a normal curve for every single one of these expected rates by DRG, about half the hospitals in New York State would be, you know, below expected, and half would be above expected. Okay, we're setting ourselves up because we're comparing ourselves to the state. If we were comparing ourselves to the U.S., if we were performing really well as a state, it's conceivable we could all look much better, okay. But when we compare ourselves to the state, we set ourselves up with that as our norm. So it's also very important when you're comparing if you're going to look at an expected rate, and I'm not really keen on it personally, but if you're going to compare yourself to an expected rate, you've got to be careful, because if you compare your expected rate here to an expected rate from the CMS program, which is using a national norm and national risk adjustments, the New York experience itself is going to flavor the difference between your expected rates. I hope that makes sense. I'm sure I'll hear it if I don't. So then the observe to expected ratio is taking your observed -- your raw rate over your expected rate. So think about that. If your observed rate is higher than your expected rate, that means you have more readmissions than expected, than according to the norm that we're using, and your observe to expected ratio would be greater than one, and that means, compared to the New York State norm, you've got room for improvement. So it's something to look at if you have a lot of readmission rates that are greater than one, because that's certainly quote, unquote, low-hanging fruit. But just because, as I described before, because we're comparing to a New York State norm, not a national norm, not a best practices norm, just because you're observed to expected ratio is less than one, meaning that you're doing better than the state norm, that doesn't necessarily mean you are a top performer nationwide. It means you're a top performer in New York State. Page 6

7 And then in the reports we also provide your severity adjusted rate so that you can compare yourself to some other peer groups, so that you can compare yourself to the state or to other hospital groups, and that's sort of turning that risk adjustment on its side. Now it's showing you what the state or the comparison group would look like if they had your severity of illness mix. Okay? So I'm going to switch over and take you to a sample report, and I'll go through this kind of quickly with you so that, then, Nancy can take you through a real case study. So this is what the reports look like. They're in Excel. The first tab is just the limitations of liability -- the name and the date. So right here you can tell which version of the report you have, each one of them are dated. You should have a whole pile of them out there that you can download, again, from the Partnership for Patients website somewhere in in-house, so you could do your own comparisons from quarter to quarter. The readmission rates that we do here, they're 30 day PPR rates, so the window of time is 30 days. They're always a 12-month period of time, and the reason for that is because the way the PPRs do the chaining, and you can't just do a quarter at a time, you need to have leads and lags and all that sort of stuff. So it's always a 12-month period of time. We updated these reports every quarter. You have one more report coming, so what happens is we drop the oldest quarter and add the newest quarter. So as you look at your trend, it's going to be smooth, in the sense that you're not going to see big changes from quarter to quarter in terms of these reports because it's really 12 months of data and you're only switching out one three-month period for a different three-month period. The first report -- and all these tabs, this is all Excel -- when you open this up, you may get a message asking if you want to enable macros. Yes, yes, yes, you need to enable macros, otherwise all the cute little pull-down boxes, et cetera, won't work, and you want them to work. The first table is at the highest level of review. They're by service line. And what it shows you is the name of the service line, which the service lines are an aggregation of APR DRGs, and you'll see that later; the count for this period of how many at-risk admissions with all those exclusions taken out, the HIVs, the deaths, et cetera. Out of those at-risk admissions for the year, how many readmission chains were there during the period? And then here are those statistics I just explained, your observed rate, which should be this number divided by this number, your observed rate, your expected rate, which is, as I said, adjusted to the New York State norm, your observed to expected ratio. Then your severity adjusted PPR rates using the national -- your experience and the national -- the state rate. Okay, so here is how you can compare yourself to another system, whichever system we've been in there, other teaching or non-teaching hospitals, other urban or rural hospitals and the state, and these are all normalized so that they're on a comparable level, so that there's some more information for you as to how far, close, or away from some comparison groups you are. Page 7

8 The functionality is is that you can sort this. If you use this little pull-down box, you'll get a whole list. Every single column in here can be sorted on. You can sort by service line name or, if you want, you can sort by let's say readmission chains, and you can sort ascending or descending, so let's sort descending. And when you choose that, it will show you the service lines that you have the most readmission chains in at top, and the least at the bottom. If you were to change this to ascending it would go the other way. The readmission chains that you have the least number of chains for the service line to the most. So you could do all this in Excel too, if you know how to sort and sift in Excel. We created these pulls-downs for you to make it easier. Okay, you also should know that at any time you could right click on one of these tabs -- no, I guess you can't move or copy. You can't move or copy tabs? Yeah. Oh, there we go, you right click on the tab. You could click move or copy, create a copy, and save it in a new book, and you could save that one tab some place else. You'll get all sorts of errors because of the visual paste basic code. That wasn't smart. Okay, we're back. But you could copy it over and you have another version of the table here, and then you can play with this. You can pull the numbers out. You can make charts or graphs with it. It's Excel and it's yours and it's your data and it's your number, so you can use it however you want. Oh, I froze. Okay, here's hoping I'm unfrozen now. I really did myself dirt here. Switch windows, there we go. Okay. So moving along the tabs, this tab is breaking out these service lines into the APR DRGs. So now you just have a lot more data. The product line is the same thing as the service line, okay, so this is listing the service lines, the DRGs, the DRG description, same thing, at-risk admissions, readmission chains, all the same statistics. Again, you can sort by any of the columns. You can sort ascending or descending. So if you wanted to sort by product line ascending, you would be able to see everything grouped by product line, and you would be able to see which APR DRGs are being grouped into product lines. And then if you had a particular product line over here that you were particularly interested in, let's say general surgery because you had a lot of readmissions there, you could go here and you could scroll down to the general surgery section, and you would see all of the DRGs that fall into general surgery, and you could see if there was any one or two DRGs where the bulk of those readmissions were happening. So you're drilling down into it. The admit source table is making it a little bit more refined for you. Now we're looking by DRG, and we are ranking them for you based on the number of readmission chains for that DRG. So DRG 173, for our sample hospital here, has had the most readmission chain chains. That's at DRG level, okay. And this allows you to drill in by the admit source to see whether or not there is anything to be gleaned from where the patients are coming from. Page 8

9 Similarly, now you've got the same DRGs, the same order, now this is the discharge status, where did you send these patients for those DRG 173, and is there anything to be gleaned from that information? Sometimes yes, sometimes no. What we're trying to give you is lots of opportunities to look at the data many different ways. Payer table same thing, same DRGs. Now we're looking to see, for 173, what the readmissions were by payer. This table, the top readmit DRGs table, this is based on the analysis that Dr. Jenks did several years ago, back in the "New England Journal of Medicine," where he and his team looked at readmissions and then looked to see what the five most frequent or popular reasons were for patients being readmitted. And what you really want to be able to see -- and this is again based on the same ranking of those DRGs for you highest number of readmissions -- why are people being readmitted. And what you want to see, I suppose, is, you know, what kind of related conditions. On the medical conditions, for COPD for instance, you would expect to see COPD and other types of related conditions for this medical condition to be there. Okay, other vascular procedures, most frequent coming back from more vascular procedures, does that or does that not make sense? But how about postoperative post-traumatic other device infections, other complications of treatment, malfunction reaction? These don't sound very good. So this gives you some insight as to what the most common reasons are, why patients are being readmitted, and can point out to you if you've got some sort of a quality issue going on that might be causing those readmissions. And then finally, the two most detailed of the tables, which would help you, again, for those DRGs, that same list, the top DRGs, this is all blinded here because this is sample data and, you know, we have certain restrictions. But your data does show you information on those readmissions, and if they were readmitted to your hospital, all the information about the readmit to your hospital. And the physician table shows you, again, those same top DRGs -- those same top DRGs showing you who the physicians were that had the most readmissions. So lots of different ways to drill into the data. I'm not sure whether I went too much in the weeds or not enough into the weeds. Always tough to tell, especially on a webinar like this. But what I'm going to do is is I'm going to stop now and I'm going to make Nancy the presenter, if I can. Yes, it looks like I can. And Nancy is going to take you -- share her desktop with you and take you through some of her case studies and the way she's used the reports to ferret out interesting information. So, Nancy. All righty, let me see if I can move this ahead. Okay, what I'm going to try to go, and we really only have 15 minutes, so you all know in the crowd that if you go to analyze data for readmissions it's a very thoughtful, very frequently, time-consuming process, and so obviously I'm not going to be able to do it that way. But I just want to at least go through this enough to give you some helpful hint. Perhaps there are Page 9

10 some things or some ideas that we'll go over today that you haven't thought of when you've done the review. I want to preface this that we're using blinded reports from a major academic hospital and a small community hospital outside of New York State so that we don't run into any problems at all. With the whole readmission piece, before I get started, there's a real reason that these reports are called diagnostic reports. And that's because they are not necessarily ideal to use. If I wanted to do statistical process control reviewing my readmissions patterns and trends over time, this is not necessarily the report I would go to. But this report feeds a lot of the different practice pattern issues and other new development issues that are very unique to your hospital and can provide very valuable information. So when you're looking at your readmission work and when your CEO, you know, now in our faces, when your CEO comes down and says, "Well I know you've been working hard on readmission but you need to double the efforts" and all that kind of stuff. One of the important things to recognize, I think that's very different, is that this side of the curve here is more the generic type of readmission that you're doing, the factors that would apply to any patient, you're risk assessing them, you're doing post-hospital discharge planning, you're using many of the readmission bundle elements. What we're really talking about today is really focusing on the diagnostic piece and how that diagnostic piece feeds information that is very different in some ways from the traditional readmission work that you're doing. So this is an academic hospital that I'm looking at. And I cut and printed parts of the report. And I'm going to take a look at this hospital, and, first of all, I'm not going to quit, because when I first look at this hospital -- and I tend to use observed over expected, and I tend to sort by that particular element, because that's a good first cut to see what's going on and what is particularly a problem. So when I look at my hospital, literally, every single area except five area, you can't see the whole report - - the observed over expected is higher than it should be, and in some of the areas quite drastically. So other than quitting, where the heck do I start here in trying to really take a look at, you know, meeting my CEO's needs when he says, or she say, you know, you need to drop the readmission rate further. So the first thing I would say is, when you take a look at this you try to find your service lines. Gloria showed you the service line chart. And if you look back at that service line, what I'm seeing in here is that the cardiac service line, particularly around, you know, procedure and general cardiology, Gastroenterology are problematic, and there are some areas that look problematic in general surgery and general medicine. So probably it's going to worth it to me to drill down to the APR DRG level that we mentioned. Page 10

11 But I'm going to go, "hmm," because, you know, there's some issues here that are really, really strange, in that there's an extremely high OB-GYN readmission rate, and some focused areas like hematology and renal failure, and that's really important, because when you take a look at specialty areas, you're not necessarily going to be doing your readmission work on OB-GYN. But if they're sitting there, especially in the Medicaid work in this group, with an observed rate over expected is really high, you know, it's going to behoove your medical staff and your nursing staff to say to that department, "You need to get involved and figure out what's going on here, because that's very unusual." In this particular large medical center, it's obvious that they are very heavy in ortho and neuro. You'll see it with their numbers. And that actually is the area that they're not over. That's the one small area that they're not over. So that's an interesting thing that you would keep in the back of your head. So if you take a look at the cardiac line, you know, this is what I'm seeing. I'm seeing cardiac cath, not an area that usually has or should have high readmission rates. We're talking about cardiac cath, relatively common procedure these days, and you're in this hospital is sitting here with twice as many readmissions than it should have. So, you know, this is not necessarily -- it could be related to the fact that this hospital tends to -- I'm going to make a big assumption that this hospital doesn't do well in readmission management at all, because their numbers are high across the board. So they got endemic problems in terms of everything, in terms of how they assess patients, how they case manage patients, how they transfer patients. But aside from that, they certainly have issues in cardiology. So what would be really appropriate, especially in this focused area, is to reach out to those cardiologists and really find out if they can tell you. You'd be surprised, when doctors and nurses see this information it really becomes eye opening, and they tend to sort of spew out pearls in terms of what things are going on to cause this. But if they don't and you can't get enough information from that, you know, that might be an area where if you don't do post-hospital calls or have a hotline for post-hospital follow up, it might be worth it to do for a month and just see what's going on. Why are these patients coming back at this high rate, especially in procedures that should be, you know, relatively risk free, relatively. And then we'll go into general medicine, and I've already told you, it's time to drill down into the DRG, APR DRG level because it's not telling you anything on the service line. And when I look at this -- and I'm going to make a lot of leaps here, you guys, because, you know, we only have 15 minutes, and so I'm trying to sort of draw conclusions to try to help you see where these things can go. But as I look at this, you see in general medicine that the bulk of their issue is in APR DRG 861, which many of you are very familiar with, because that's that catchall DRG on signs and symptoms and other failures influencing health. But then you look at the next three areas that are problematic for general medicine, and there are other complications of treatment, allergic reactions and poisoning of medication agents. Page 11

12 So for those of you who are very familiar with these DRGs and are clinicians that might be working with your analyst to assess this, I mean what is the first thing that comes to your mind? There's a medication management issue, highly likely, a root cause here someplace, because most of these diagnoses have something to do with electrolyte imbalance, patients coming in frail, weak, either not hydrated enough, hydrated too much, take their medicines wrong. And so in general medicine, my inclination would be to flush that out a little bit more and see if what I'm seeing in general medicine, based on the APR DRG, is really and truly fact, and does medication management play a factor in all of the areas that happen to be high in general medicine. Literally those are the one that is are high. And if that's the case, then, you know, you would do the obvious, and lots of the stuff that you're now in our partnership program. You know, you would look to your pharmacist for maybe some assistance in your medication management program, or your high-risk management program or your medication reconciliation program, because point in fact is that might be a root cause when you take a look at what's problematic in general medicine in this hospital. And then, of course, the use of the observation unit, because if these patients are all situations where they come back in and all they need is a couple of shots of IV Lasik, or they just need a couple of bags of fluid, you know, they could be treated very adequately in an observation unit and have those things stabilized. So there's a general medicine reason, and I'm hoping that this is example of why these diagnostic reports are really important. They are not about tracking my monthly rate and trending it and using statistical process control to see if the readmission improvement work is helping. They're diagnostic tools to really help you drive to what might be root causes or other kinds of factors that are playing a big role in readmissions but it's just hidden. So the next one is gastroenterology, and this area happens to be high too. And, you know, I take a look at it and when you look at it, the Gastroenterology, all of the majority, I'm not going to say all, but the majority of the high-volume areas have a lot to do with high GI distress alcohol liver disease, and maybe some pediatric things with the non-bacterial gastroenteritis. So it may be very appropriate, again, once you delve into that a bit more, to reach out to those divisions in your hospital and say, okay, what is going on, and perhaps maybe part of the issue is, is that for all of these GI patients with these diseases that are related to substance abuse or other things, maybe a more comprehensive integrated general medicine psychiatry substance abuse program is appropriate to prevent readmissions. Or a lot of them could be end-stage liver diseases where palliative care would be important. Or in pediatrics, if you're really readmitting all of those pediatric patients, especially those of you in the world of DSRP with Medicaid, you know, we need to learn how to control those kids coming back in for nausea and vomiting. So this is where it really helps you understand not necessarily whether or not you're Page 12

13 transferring and handing off information but really what's going on in the pattern practices and the protocols and the way your staff handles clinical care delivery at your hospital. Same thing with endo, I think you're getting the theme here. When you look at endo it's electrolyte imbalance and, of course, diabetes. And this hospital happens to have a particularly high diabetes volume, that's high, because most of our hospitals only have people readmit -- people admitted with diabetes in 2 to 3% percent of their diagnoses. Diabetes is one of those hidden diseases that they usually don't come into the hospital for, but when they do it's problematic. Well, you know, the message that you guys all know is it's problematic when they're secondary, because it gets hidden and, frequently, not addressed effectively. And so when you're seeing diabetes as a high readmission in your primary, you've probably got an issue around diabetes. And the issue around diabetes discharge planning, when it becomes clinical, again, has a lot to have do with medication management, the use of nutrition, whether or not you have certified diabetes educators on staff, whether or not you have diabetes self management follow-up programs at discharge. The way you resolve those practice patterns, you know, is really by dealing with the protocols and the services available in your area. So I'm going to take a look at general surgery, and in this major medical center general surgery is just like looking at the hospital, forget it. They are literally -- every single type of general surgery is high. And so I'm going to make the assumption with general surgery that, you know, it is confounded by the fact that this hospital, in general, doesn't have really good readmission rates, and, really, general surgery is just a microcosm of what's going on in the whole hospital. And so there's not much you're going to glean out of here from a particular practice pattern or opportunity at first blush, because they're bad across the board. So in this situation I might start my readmission pilot or my readmission work in a surgical unit, because the opportunities are very clear and you know that you can't really go after focus practice -- excuse me, practice pattern changes, so that might be a really good area to start working on high-risk criteria, posthospital care planning, discharge planning. However, those of you, as you delve into the data a little bit more, also know that there are opportunities specific to surgery, should you find that those are the cases. And a lot of those opportunities really are what you already are dealing with in terms of preadmission readiness and preadmission preparation for elective surgeries, all of the hospital acquired conditions and infections that are generated out of surgical admissions and our SSI infection rates and wound care and those kind of things. But with this one, I think general surgery, in this report, is so bad across the board, just like the hospital, that it's hard to glean any practice opportunities here, because they're just no good any where. So either delve into this a lot more or pick this area to really do some focus readmission bundle work. Now we mentioned OB-GYN, and, you know, this is ridiculous. Page 13

14 I mean if you look at these people's data, you know, they have one-and-a-half -- they're 150% high, above their normal readmission rate for the state. And, you know, that's just unacceptable data. And so there's no way that a quality department, a chief medical officer or chief nursing officer shouldn't go to the OB- GYN department and really request that they take a look at what's going on here, because it really requires a very focused root cause analysis to determine what opportunities are going on. There is something missing in the clinical care delivery system in a hospital that has these kinds of numbers, and there's no way to address that with, oh, we're going to high risk assess the mother. You know, that's incredibly important, but there is something else going on here. And hematology and renal dialysis, the only reason I mention those is that it's not necessarily unusual to have high sickle cell anemia readmissions. That's pretty usual. They're low volume, so they don't always come to the top. But, again, with Medicaid and DSRP, they'll come to the top a lot. And, you know, that just calls for a sickle cell protocol that transcends the hospital and the community. And, you know, there might be an opportunity for a division to look at that. And the same with renal failure, everybody's renal failure rates are -- it's hard to be competitive in that area. But certainly if your rate is two times the normal you might want to look and see what's available for post-hospital care and across the system so that you can improve your rates a little bit in that area. So those are sort of specialty type of things. The next thing, and as Gloria pointed out, and I'm not going to go over it for the sake of time, but there's point of origin and discharge disposition. And I will say that what's very interesting is that when people really get into their reports, that nine times out of ten they're going to find coding opportunities as well. And let's face it, coding opportunities are process improvement. They're not clinical but they certainly will help in the bigger picture of things, especially for our colleagues in finance and billing. And so, you know, take advantage of them. Just, you know, coding also plays a role in those risk adjustment factors. I'm not saying that you should -- I'm not saying that that's a strategy for reducing readmissions. Readmissions should be reduced. However, when it comes to payment penalties, which I know this isn't really about this webinar, but when it comes to payment penalties, those risk adjustments matter. You know, we're getting calls from hospitals that have low raw readmission rate, and then risk adjusted, their readmission rates are above expectations, observe to expected ratios are high, and that might be a coding problem. Sure. So don't discount the fact that you're frustrated because something doesn't make sense. I mean if you can send it to your HIN department and figure out what's going on in coding, there could be an operational improve there. The other thing is that there's discharge dispositions, and what is interesting and I'm trying to reflect what you can find in this, is I mentioned to you already that this medical center has a huge volume of ortho and neuro surgery. I think they do like 5,000 cases a year. But their readmission rate is not high. It's not great, but it's not high. So, of course what do I think when I first look Page 14

15 at that? Oh, my goodness, they must be sending everybody to, you know, rehab or nursing home to have a low readmission rate and nothing going on. I go to the disposition, they hardly send anybody to rehab or long-term care, and they actually have disproportionate amount of people who even have home care, so what the heck. So what that tells me is this is an organization that has a high volume ortho/neuro program that they probably have resourced very well, and they have protocols and practice protocols in place, probably prior to admission and posthospital, and they just have a well functioning clinical delivery system there, because it doesn't have any - - when you look at the admission and discharge disposition, it shows nothing. So it's very interesting. And these are the kinds of things you can glean. Now that's a major medical center. I want to just say that this is a valuable, it's small community hospital, which, in fact, is maybe easier. Because this small community hospital that Gloria gave me from another state to analyze, it's the opposite. This particular hospital has only, I think, nine service lines that have higher than expected rates out of 30-some. And when you look at the nine service lines, they're not particularly high volume and the only thing that really sticks out here as a opportunity is pulmonary. I'm not saying that's not the only thing they want to work on, but, literally, in comparison to the medical center where, you know, you might want to quit when he when you look at their numbers, this hospital, eh, pretty obvious small community hospital. They have an opportunity in pulmonary, and when I drill down to the DRG level, their opportunity in pulmonary is almost exclusively in COPD. So if I am a small community hospital and that's what dragging me down, I'm going to think seriously about my clinical delivery system and respiratory rehabilitation programs or respiratory nurse champions, because that probably is going to be what makes the difference. Also, I want to, before we quickly end, I want to make sure that you guys realize that there's a physician profile here that is phenomenal. And I have it blinded obviously, because we can't show details. But let's pretend we're looking at the cardiovascular surgery line for this physician, and I'm drilling down to physicians, and, you know, I'm finding my problems in the product line in these APR DRGs, and so I'm getting a sense of where the problems are. And then so I go from service line to the APR DRGs, and then I go to find physicians by those APR DRGs, those three different tabs that Gloria went through. And I find out, interestingly enough, the physician that does the highest volume at this hospital doesn't have a readmission problem at all. Isn't that interesting? I wonder what that physician does differently. On the other hand, what I found is many, many, many of the low-volume physicians -- and it's not showing it well here -- have readmission problems. So there's certainly something going on with cardiology, where the physician doesn't do a lot of cases, they have a higher rates. And I apologize if this doesn't show it well here. And that may be a factor of a medical group that switches hospitals all over the place and they're just not familiar with the system. But there's probably an opportunity there. And at best, I'm going to speak to Dr. you know, 141, because that physician is doing excellent and has the highest Page 15

16 volume in that organization, and what is that physician doing correctly, because I'm going to guess I have a lot to learn from this physician. Oh, here this shows it better. I'm sorry. This is DRG 175, and if you look at it, I sorted from the readmission chains. And if you look at the physicians that have the higher readmission rates, it's the lowvolume physicians. That's a concern you'd want to check into, and that's a little bit of a different practice concern, but nevertheless, important. So in wrapping this up, this is actually a pretty easy diagnostic tool to use once you get used to it. Anything looks overwhelming. And one of the big recommendations I would have is not only for you guys to drill into it a little bit more but give it to your CMO, your chief medical officer, give it to your department chairs. When you go to OB or when you go to cardiology, let the chairman of cardiology or OB go through and play with it. And I'll bet dollars to donuts, knowing most physicians, they'll be intrigued and they're going to want to play with it. So get it outside of the quality department and let other people play with it and analyze it and provide feedback. I think it will be really very good. And don't forget to use focus groups, because when you see data like this and you're not sure because you're over in the quality department or in the innovations department or whatever, you know, hold a little focus group with your cardiologist or your OB and show them the data, and see what comes out. It will really be important. So with that being said, I'm going to turn it over, I want to end quickly and turn it over to Loretta for some Q&A. I have one question here from Kathryn Rose [ph]. She's saying that she just sent an to CMS with regards to the codes for readmissions. Some PCI cases might be readmitted for CABG because of a stage procedure. Is there any possibility that that might not be -- that's not being addressed appropriately in the CMS methodology. Kathryn, we'll take that offline because I don't think any of us has the answer here readily. So we'll take that offline, but we'll get that answer to you. Yeah. Although I will say, and this is with tongue in cheek, because we don't know, that PCA to CABG surgery is the type of algorithm that PPR uses to say, "Oh, well that was sort of planned. That was progressive." Yeah, that's PPRs. But the CMS readmission -- Well, CMS, yeah, we're not talking about that today, so we'll talk that offline. Page 16