 Okay, we're back. This is Dave Vellante. We're live here at IBM's IOD, live from Las Vegas. This is theCUBE, Silicon Angle TV's continuous coverage of IBM's IOD. This is Big Data Week. We're here at IOD, covering information management, covering big data. We're on our way to Strata. Half the team is there, ready. We'll be broadcasting live from Strata on Wednesday and Thursday. This is our last segment of the day. We're here with Craig Reinhardt of IBM, who is the director of ECM strategy and market development for the IBM software division. And I'm also joined by my friend and Wikibon contributor, Gary McFadden. A gentleman, welcome. Thank you. Welcome, thank you. Good to see you guys. Craig, let's start with you. So, I mean, IBM has really done a good job of focusing on specific industry verticals. It really made a big emphasis on that. IBM has always had a strong emphasis on that, but really brought its analytics and its big data and mojo to different vertical industries, really peeling back the onion and supporting clients from a services and a technology standpoint. Doing a lot in healthcare, right? So, why don't you talk about what you're doing there and just kick it off and then we'll get into it. Well, healthcare has been a business. You think about the amount of paper and images and those cases. It's been an industry where enterprise content management has been important for a very long period of time. But, you know, this is now an industry that is grappling with change. You know, but business models are moving from fee for service to pay for performance. You've got new incentives and disincentives to a way that care is delivered. And quite frankly, the biggest challenge that we're trying to help with is, how do we raise the quality of care at the same time lower the cost of care without, you know, obviously delivering, you know, bad care, right? So better quality, lower cost. Now, what we announced here at IOD was a new solution, leveraging a number of capabilities from the enterprise content management business called IBM Patient Care and Insights. And what that is is really designed to help with those healthcare organizations who needed to do exactly that, deliver better patient care at lower cost, whether they're provider, payer, government, pharma, you name it. Research, you spend a little bit of time looking at healthcare. Healthcare has not always been the most dynamic in terms of adopting technology, but it's kind of resistant to technology. Is that changing? Well, I'm not sure you could say that healthcare is a resistor of technology since they spend about $750 billion a year on different types of technologies. It's really the type of technology. The technology has been primarily at the bedside with devices and MRIs and so forth, but where they've been lagging is in the IT space, and in particular, most recently, this whole drive towards EHRs or electronic healthcare records, which has been somewhat driven by Obamacare and the Affordable Care Act initiatives and so forth, and also incentives for hospitals to migrate from older systems to new EHR systems so that they can enable some of these quality programs and initiatives to lower costs and also to improve patient outcomes. So actually in 2010, the statistic was from the industry that they spent about $88 billion on EHR solutions that's representative only the U.S., which is a population of about 5,000 hospitals. So the thing that's happening is really, there's the paradigm shift that we're right in the middle of, and they're the early adopters that have already made the shift from fee-for-service to quality from paper and non-technology solutions in the back office to these EHR solutions that have cost some of the hospitals, some of the larger firms, into the billions of dollars, literally, one, two, three billion dollars for a very large hospital group to do the migration. So to say that they're not spending a lot of money is not true. To say that they may be spending a lot of money but not necessarily spending it all right is maybe also true. And I think that's one of the opportunities that IBM has and other vendors in the healthcare space is to help to direct where the spending is going to be most useful to improve costs and to lower the, to improve outcomes and to lower costs. David, think about it, right? When your whole business changes and you go from being a fee-for-service business volume to one where you have to deliver value based on quality, this begins to put a real importance on having information that you can put to work in your business. If it's a very different paradigm as Gary's talking about. So it seems that the value proposition in healthcare would be twofold. One is, you know, cut my costs. It's a very budget constrained environment. The other is save lives, improve the quality of care. Talk about that a little bit and how IBM participates in those two vectors. Well, if you look at how care is delivered today, there are guidelines which don't cover all scenarios. Your typical care guideline is based off of, you know, a patient with one disease and all the different things that you do to treat someone with COPD or congestive heart failure. But things begin to break down when the patients have more complex scenarios or there are multiple diseases in state. There aren't guidelines in many of these situations. So we're left up to what doctors know and the personal experiences of these two delivery things. You end up with, at times, inconsistent care. It's very much dependent on knowledge to deliver the quality of care. So there's a consistency issue there, right? The best doctors deliver great care but not everyone's the best doctor. And where it really gets interesting is which doctors are the best for your situation? Or you have unique things that are about you that not all physicians are going to be able to optimally treat and care for. So part of where we're trying to help is by complimenting those guidelines, where they exist, those treatment guidelines, with data-driven insights from actual patient population data. Think about all those times you went to the doctor and you had an interaction and they recorded that into notes that went into your electronic health records that are now being deployed. The goal now is to leverage all of that information that's beginning to build up, whether it's structured data or whether it's the unstructured content. Therein lies the value of enterprise content. How can we analyze that data, find insights that are now currently trapped in that data to make more information better at the point of care, giving the opportunity to jump in, intervene earlier, deliver better care? So what consistent care? So what's the model there at the doctor level? A doctor scribbling notes and half the time you can't read what he or she's writing. Is that changing to one where it's a voice activated? You got a dragon translation system that goes into some kind of data repository. Is that actually beginning to happen? There are a number of things going on there. There have been advancements in handwriting recognition technology. Our own data cap technology does that. You have a number of technology speech to text has come a long way. These encounters or interactions that you have with your doctor, sometimes they're recorded during the session. Other times the doctor sits down quickly and into a voice recorder records the consult if you will. Those things can be transcribed manually. There's a giant industry around transcribing notes that are just put into records. Then there are technologies that do speech to text. So there's a number of ways that this information gets generated and then captured to become part of the electronic record. What the industry hasn't done effectively, and this is where we're seeing a lot of change, is begin to analyze that data and surface those nuggets of information that are really in the past been trapped in there. So talk more about that. I mean what's the analytics angle that IBM's bringing into the industry? Well we feel there are three things that are really important from an analysis perspective. The first is content analysis or text analytics. You may have heard the term natural language processing in the past. IBM did the thing with Watson and Jeopardy last year and if you didn't know this, the Watson machine that won on Jeopardy was based on natural language processing. Which is a way to interpret text or human, if you think about text, it's all what comes out our brains into our fingertips is text, right? We write emails in text. It's how we express ourselves. It's how those recorded conversations get transcribed. So using content analysis as a way to read that text and extract from it medications, diseases, symptoms, all these things that have been building up over years but not being analyzed becomes the critical thing. Because what they really want to analyze is usually not what you filled out on the form in the waiting room. That's useful and it generates nice structured data but it really comes down to what happened in the interaction between you and your caregiver, what the caregiver thought, what the caregiver thinks that needs to be done, how all that was recorded. And the outcomes that became associated with that. So the first thing is content analytics. Second thing is predictive analytics. Because once we understand what has happened in the past and we've analyzed the structured data and the unstructured data, now we want to look forward and figure out, okay, if all these things happened in the past, what's most probably going to happen next? So let's predict the possible outcomes and the probabilities associated with things, different outcomes that might happen, because we can do that predict forward view if we do the retrospective view. Example, you've been discharged from the hospital or something, the New England Journal of Medicine says one in five readmissions are preventable and yet they don't happen. So why don't we use analytics as a way to figure out who's the most at risk of being one of those one in five and put an intervention program around that to make sure that the patient actually follows that three page discharge thing that you leave with. So for example, yeah, they might not be taking their medication, they might not be prone to do that. You could predict who's not going to take their medication with reasonable high degree of certainty. Oh, there are a lot of socioeconomic factors here, right? You have people who are in assisted homes who forget, take medications, you have people who can't read, so they don't follow the discharge instruction, you have people who are elderly and don't have transportation, so they don't get prescription sold. So you can identify those, make an investment in providing some services to assist them and then actually cut the cost. If you analyze the social data, then you have a broad, rich set of information to be able to do that kind of analysis. Okay, so okay, we got content analysis, predictive analytics. Yeah, there's one more and this is a cool one. It's called similarity analytics, patient similarity analytics. What is that? So this is brand new and it was invented in IBM Research over the last three years and we just commercialized the technology. Now what this does is it profiles patients across tens of thousands of different dimensions. So all of the history data about you or me, we can look at in a given outcome, like am I at risk of the onset of disease and using these similarity analytics, figure out what are my unique risk factors that are most likely going to determine whether something happens to me or not. Once you know what that is, we can then go out and find within the population, I think thousands or millions of patients who are in this population, who are the most similar patients to me based on my unique risk factors. So if we can identify through analytics, which we now can, who are the most similar patients to me, now we not only know what, this is what's called a cohort in medical terms, now we not only know what happened to those most similar patients to me before the point I'm at, but we also know what happens next to the patients who are the most similar to me. So now for my treatment, we know what are the possible paths forward for me, what are the probability based on my unique attributes, whether I'm going to go this path or this path, is am I going to go on a green path for a good outcome or a red path for a bad outcome? So similarity analytics really bring a whole new level of making healthcare treatment personal and about you, the patient, putting the patient in the center. So Gary, we were talking before about how healthcare's been slow to adopt information technology. Are the lines blurring between information technology and technology? I mean, hospitals and healthcare, they'll spend on medical devices, right? Something that'll save lives. In your opinion and based on what you see with your clients, have we reached a tipping point where those lines are blurred and people will begin to adopt these types of technology because of the business value? Well, I think that they're the leaders in the medical industry, like in any industry, the early adopters and so forth that understand the technology and are partnering with IBM to actually drive these solutions forward. So that's already happening. That already exists. Like who, can you name names or give a good name? Well, the Mayo Clinic, Cleveland Clinic, Kaiser Permanente in California, Mass General. I mean, there are a number of them, right? So they're all really on the forefront of this whole process. They've also made the shift from fee for service to quality. They don't compensate their care providers based on transactions. There's a quality aspect to it and data is something that allows them to do that and to do quality checks and so forth. But there's all transformation going on that is being at the high level that's really not being followed by the aspirational users. The 99%, if you will, of companies out there that don't have the means or don't have the insight or the direction in terms to be able to take advantage of their data. And one of the areas that the hospital industry is really behind other industries like the financial industry, there's a lot of this data that would provide information downstream for analytics whether it's content analytics or similarity analytics or so forth is locked into these different formats like TIF files or unsearchable PDFs or paper or microfiche. And unfortunately that data is not accessible to these solutions. So there's a whole population of aspirational users that are not the early adopters that unfortunately is the majority of hospitals out there that still haven't gotten the fact that they need to invest in these technologies that allow them to make the connection between where the data is, getting the data in shape so that can move it downstream so that they can take advantage of all these analytics capabilities whether it's fraud detection, whether it's similarity analytics, whether it's predictive analytics, all these other things, you can't do any of that if you don't have access to the data. And that's a very big problem in the healthcare industry that the financial industry had to solve 10 years ago and now it's happening to the healthcare industry. Where is healthcare in terms of the whole data life cycle management? Are they in the first inning or are they more mature than that? I would say in different areas, they're in different places. What Gary's saying is absolutely true. What has happened in recent years though is there's much more interest and investment around this stuff. And I'll give you a perfect example. In fact, a perfect example where this readmissions thing that I mentioned earlier. So readmissions is a problem that costs everybody. It's bad for the patient, nobody likes to be in the hospital, it's bad for the provider because valuable resources that could go to other patients are being consumed by a patient who shouldn't be there in the first place. The CMS doesn't want to pay for these things. So we did a project at a provider where when we looked at their readmissions challenges. And there was this one example that stuck with me, patient X. Patient X was readmitted to the hospital six times over an eight month period for the same problem. And the data was there the entire time it didn't change a whole month, a whole bunch. And what they did at the sixth encounter, eight months later, is they started treating the patient differently. So the obvious question is, why didn't they treat that patient differently at the first encounter instead of the sixth encounter? Absolutely. The reason is this is sort of built on what Gary was saying, right? The data was trapped. They couldn't surface those insights at the point of care. It was only at the sixth encounter. They found out something ultimately got this patient in an intervention program. But here's the thing. And here's why you're seeing change. Two things. Number one, 83% of the costs came after the first encounter to treat that patient and were avoidable. Had they treated the patient differently. Ton of waste in that 83% of the costs. The second thing is, CMS and the people who currently pay for these services are wising up to this. They're actually new Medicare payment penalties that went into effect this month for those providers with high readmission rates. And if you're above the line, they start deducting from your readmission. I was going to ask you about that. Is there a stick as well as a carrot here? So there is a carrot and a stick. And you want to cut the waste out because there are always charity care things that you do. One organization we work with delivers over $400 million in charity care that they don't give paid for. So they want to cut all the waste out of that stuff that they possibly can. So, Craig, what's the number? What's the, isn't it, the number is something like 60 or 75% of illness is preventable. Do you know what that number is? I think it's a large number. I don't have it exactly, but it's got to be more than 60% of illnesses are preventable. Well, I don't know what that statistic is, but I'll tell you, the one that blew my mind was, and it's a two part statistic is 70% of us are going to die from a chronic disease. Diabetes, congestive heart failure, COPD. That's what the CDC says. And as those diseases progress, guess what, quality of life goes down and costs skyrocket. 80%, I'm sorry, 20% of the patients with chronic diseases account for 80% of healthcare expenditures. 80-20 rule is in effect here. So, if 20% of the patients are driving 80% of all the healthcare expenditures, guess what? Maybe you should focus on those 20%. No, let's focus on the ones that out here who haven't progressed to this point and keep them well longer so that they don't become the 20%, if you will, pre-remediate them, right? Before they begin to deteriorate or before the onset of a disease. So, this is where these analytics are invaluable. If we can find the intervention opportunities in the people who are today well, but eventually at risk of coming down with those chronic diseases, then it brings a whole new light to wellness because I don't know about you, but if I knew that my destiny had a certain probability that it was going to go this path, and it was bad, I'd do everything possible, not to be on that path. Would you have to give up beer? I might have to, I might have to. No, I gotta keep the beer. Hey, these are the choices we make in life. But so, you think we'll see the day where we can actually get a handle on rising healthcare costs? Can IBM and your cohorts in the industry make a dent through analytics? Yes, I do think that. I think we've got clients today who are on this journey. We've got clients like Wellpoint. We're already doing some of the things I'm talking about. I mean, look, what's going on today is not tenable. The U.S. is the number one spender of healthcare globally, and we spend 50% more per capita than number two on healthcare, and guess where we rank in terms of quality of care delivered. Yeah, not at the top, that's for sure. You want to guess? Like 10th or 12th or something like that. Nice try, Gary. We rank 37th. Yeah. 37th. We've narrowly eked out Slovenia and Cuba 37th, and not much of an ROI there. Who's at the top of the list? Do you know off-hand? No, I don't know. I can get over the fact we were 37th. And it's estimated that 30 cents on every dollar is wasted, duplicate tests, misprescribed drugs, unnecessary procedures, 30 cents on a dollar is wasted. So another paradigm shift here is the revenue model shift, right? Which is interesting because data data is not only an opportunity for healthcare, big data is an opportunity. If you look at that, this morning less, one of the presenters talked about data is the oil of the 21st century, and I forget it's written. Data is the new oil. Right, it's better than oil actually, because downstream, when you use oil, it's gone. Data gets better the more you use it, and you can sell it over and over and over again. So there's a revenue opportunity that a lot of these hospitals and a lot of these organizations are missing because the revenue model will shift at one point where they may get in the future more of their revenue from data and reselling it or reusing it and repurposing it than they are from actually fee-for-service. One example is that the paradigm shift that's already happening in the pharmaceutical industry is that the pharmaceuticals are buying data. They've realized, Dave, that using longitudinal or historical data going back 25 years is a more accurate prediction of whether or not a drug that's on trial will succeed and will work than doing clinical trials. They're changing their whole emphasis on how they do clinical trials. Clinical trials, the way they've been practiced forever, are an artificial environment. And now with data, they actually have real time or more of a real environment for data that as Craig was saying earlier, we're having this discussion, that not only focuses like a clinical trial on one aspect of somebody's illness but also looks at the holistically at the patient and looks at relationships between illnesses that might accompany one another. The data, the richness of the data is so much more useful to the pharmaceutical industry than a clinical trial could ever be. It helps to accelerate the process of bringing drugs to trial and to completion and approval. And it also helps them to identify opportunities to fill their pipeline with drugs that are being developed by startup biotech firms and evaluating them. And it also drives a venture capital and private equity business as well because they have data where they can put their money behind products that are more likely to be winners and to be approved and to be effective. And this is all data driven. You can't do this without this data. So it's changing, it's already changing the whole paradigm shift for how clinical trials are being done. And it's only going to follow in healthcare in other ways. Many of the providers are just, this sort of light is just beginning to come on to them. They're back to the oil metaphor. Who's sitting on the oil? All the providers who have all this information that's been building up over these years. And while in the past, they may not have been managing that information as an asset. Maybe it's still on paper, locked away in a file room somewhere. But now it's a business opportunity. So this is part of this dynamic that's going to change. And it will change business models in the way that income gets generated and so forth. And the pharma companies as Gary's correctly pointing out are among those who are interested in getting access to those insights. Us patients are interested in getting access to those insights too. Well, and that's a great opportunity for individuals to be more proactive in their care as well. If you feed that back, one of the meaningful use portions of Obamacare is the requirement to give more information to the patient and build patient portals. So I don't know about the rest of you, but I'd like to be able to be more proactive in my own care. I'd like to know if I have a certain ailment, what the treatment modalities are, what are the ones that are more effective. I want to contact other people that have the same affliction. I want to do some research. I want to do some research. And the payers. The payers want the data too. Many of the payers are trying to change their relationship with their member providers. And one of the ways that they want to do this, they want to offer incentives to have providers. Most of the major players have announced these kinds of incentives, well-point announced last year, over a billion dollars in incentives for following treatment protocols and these kinds of things. And in order to do that, you have to have the data to figure out which are the treatments that generate the best outcomes as well as at the lowest cost. So this is a whole new game. And it's all based on the theme of the event, big data, think big, of how do we leverage this data, whether it's content-oriented and unstructured in nature or structured data, so that all these things are, they're all happening in healthcare today. They're happening in other industries too, but there's a real urgency in healthcare today. Well, because like you said, the situation's untenable and everybody's pointing fingers. They're pointing fingers at the providers, at the patients, at the payers. Administrative way, and they're all saying, well, wait a minute, we have our own unique problem, so we've really got to attack this. I want to talk a little bit about Watson. Yeah. That's something that's pretty cool and obviously has a use case in healthcare. Yeah, very much so. So what are you seeing there? Can you give us the update? Yeah, well, Watson, we didn't build it to go on game shows as cool as that was. We're not going on the game show circuit. You know, early on we announced that healthcare was a focus for Watson, followed by financial services. And we made subsequent announcements around partnerships in that area with Memorial Sloan Kettering, Well Point, and others. And Watson's still in medical school, but he's working on oncology. And Well Point and IBM are now working with a number of select providers around the oncology solution and around use cases like utilization management. So Watson is being deployed and rolled out in live productions settings as we speak. And what's great about that is, is Watson's role in all of this is to be a source of those knowledge, right? There's so much medical, I saw a stat, something like, you know, physicians can only spend like five hours a month reading updates in journals. And there's so much more than that that they should be reading, but time prevents them from reading it. So, you know, when you have something like Watson that can read all this stuff, understand it, assess it, and have it all available for someone who needs to make a decision based on the latest, you know, approved information, huge value there. So it becomes a source, an information source of knowledge, if you will. Fantastic. All right, Gary, give you the last word, maybe some closing thoughts. Well, I think there's a real opportunity for IBM to make a major contribution to the healthcare, the whole healthcare situation in the US. IBM has a, is in a position that is unlike probably any company, technology company in the world that they have a breadth and depth of capabilities that can really help to really move this whole paradigm shift. One of the things that Craig and I have been talking about is that there's a lot of misinformation out there about technology and its value within the healthcare industry, and there are a lot of bad decisions being made about how to implement solutions. And I would just say, if there are any end users out there contemplating migrating their EHR solutions or getting into this big data game, they really should look at it from a holistic perspective and not go after point solutions necessarily and really have to understand the whole gamut of issues that are applied to playing in the big data game in healthcare. And certainly IBM doesn't have all the answers, but sorry about that Craig, but IBM certainly has a lot of them and certainly has a lot of smart people bringing a lot of brain power and customer insights to bear on this very big problem for us, big data problem. And I'm excited to see what IBM's going to come up with Watson, with the ICPA, which is IBM Content and Predictive Analytics, which is kind of like Watson-like or a Watson-ready type application that is already being implemented at the seat and healthcare facility in Austin, Texas, and some of the other great innovations that you guys have come up with. So I would say it's been a great show for me, good learning experience, a lot of good assets here and people should talk to IBM about healthcare. This hope that our healthcare costs won't continue to rise 15 to 20 percent. We definitely, IBM wish you good luck on that one Craig. So please come through, make it happen. We're going to play our role. Gary and Craig, thanks very much for coming on theCUBE, it was great having you guys that appreciate the healthcare drill down. Keep it right there, we'll be back. I'll bring on Jeff Kelly, we'll wrap and then we'll pack and we'll see you in just a minute. This is theCUBE, live from IBM IOD, keep it right there. Okay, thanks Dave.