 Live from Las Vegas, extracting the signal from the noise. It's theCUBE, covering IBM Insight 2015, brought to you by IBM. Now your host, Dave Vellante and Paul Gillin. We're back in Las Vegas, everybody. This is theCUBE. We're here live at IBM Insight 2015. This is IBM's big analytics show. We heard the keynotes this morning. Bob Picciano kicked it off with a number of innovators from places like Whirlpool. We saw the Weather Channel. We saw Twitter. We saw Aaron Levy from Box. Really interesting ecosystem that Paul and I and George Gilbert have been breaking down all day. We'll be here for two days. Shahid Shah is here. He's the co-founder and CEO of Netspective Communications. He's a thought leader in healthcare and one of the influencers here at IBM. Shahid, welcome to theCUBE. Thanks for coming on. Thanks for having me. So tell us about your operation, your blogger. You follow the healthcare space. You're based in DC. Lots to talk about there. Yeah, so I've been spending a lot of time in healthcare for about 15 years or so. About 25 years total in IT. I'd still think of myself as a geek because I make products and build solutions and things like that for hospitals, pharma companies and other health IT vendors. So what I like about what's going on here is that over the last five years or so as we've been moving away from old style fee for service kinds of payments into more outcomes driven and value driven payments in healthcare, the need and desire for people to generate, accommodate, collect and then hopefully manage and at some point analyze and work with that data is growing ever larger. So I thought it was exciting, all the stuff from this morning, especially IBM Watson and Health. Why the mind share shift in that business? Is it technology? Is it public policy? I mean, combination? I wonder if you could break that down for us. I think you're seeing a mind share shift because our fee for service payment system, sort of like what we've been using for decades now where if you go see a physician, you're going to get paid for whatever services the physician thinks is good for you. It's not because you absolutely necessarily need it or don't need it. It's just the service that he's performing. So that model called fee for service has been around for so long that it's created some sometimes good incentives, sometimes bad incentives in the sense that we're rewarding physicians to do more services that are more expensive rather than fewer services that are less expensive. And that's primarily because the idea of getting paid for the results that one gets out of a healthcare setting is not what's being paid for. So over the last five or six years and you've seen a little bit of movement from the Affordable Care Act itself, plus then just the structural inefficiencies within healthcare are starting to break down a little bit, everybody understands. You can't really go much longer by just keeping and paying for physicians to do things in higher and higher volume and higher and higher value. So what we're trying to see is if we tell physicians that they can manage populations, would that be more efficient, easier to do? And when managing a population of one, let's say me or you or anyone else, if we connect multiple payers and their providers and the pharmacies and the drug companies together to care for the whole of the patient, will that be cheaper? Will it be more effective care? That's the big experiment. So that's why you see a lot of movement. We cannot do fee for service world where we were paying for volume. Doesn't really need a whole lot of data. You come in, you ask for the doctor to see you. He might give you some drugs. He might do run a therapy or treatment on you and then you're gone. I mean, doesn't really need anything to analyze you in any real way. But now imagine that the treatment for me as an individual patient is being spread across pharmacies, pharmaceutical companies as drug manufacturers, healthcare providers, as well as imaging centers. Now imagine if my payment has to come from the holistic care across this when everyone is working efficiently, there's no way to do that without enormous amounts of data that we don't collect today. So that's why you see this. How do you get there? I mean, what you're talking about, I think many people would agree that affordable care was basically insurance reform. It wasn't healthcare reform. Much bigger issues involved in changing how all this ecosystem exists around the patients. Can we do that? I think the answer in general is yes because we're seeing a movement in the entire community now that recognizes that you can't care for the total health of a patient by doing it in episodes in different organizations. There's a general acceptance that that is the case. What we don't know today is what's the fix? We know that it's broken. So volume-driven fee-for-service care is generally broken in the sense that it will never care for the entire patient as a whole. What we are experimenting with, and we have to give it to the Affordable Care Act. One thing is it called these experiments. So when you think about value-driven care and outcomes-driven care, nobody said this is what we're gonna do and this is gonna work and we're done. Good sums of money, 100 million here, 200 million there, a billion there was put in to run these experiments. One of these experiments is the accountable care organizations where you can actually give people an entire population, either a geographic population or a population with specific conditions. Like you could say, I'm really good at managing diabetics, give me a thousand diabetics to Medicare. Medicare would then say, okay, you have an accountable care organization of these diabetics or these care conditions and now go forth and conquer. See how you can work together with all the different stakeholders and players in your geographic area or other mechanism. However, you decide to put them together. You create networks of networks, but it's really hard because we've created the network of networks. We like the idea that you can make more money when you're being more efficient. The problem is we don't have enough data to measure. What does it mean to be more healthy? Well, there's an incentive, there's a stick too. You're not gonna get paid unless you do this, right? The stick isn't quite there yet. So we have the carrot and the carrot is good in the sense that if you do the right thing, you get bonus payments. The stick isn't, it will be there. It's promised to be there, but it's not in a major way today. So it turns out that even the institutions that got together believing that if they worked together, they would be more efficient and therefore make more bonuses. It's turned out that roughly half, maybe three quarters of them did not get their bonus money, which meant that they actually made less money, but they were still pretty much a break-even, not really at a loss. But what's interesting is will we see, so Medicare started the trend, which as it does normally, but then health insurers should be picking this up and taking the ball forward because there's a lot more people that are outside of Medicare than are inside. Now what we're seeing is bigger players, you're United Health, Aetna, Anthem, et cetera. Many of them are starting to say, okay, we want you to go ahead and try this as well outside of Medicare. I think they're going to put the screws on a little bit tighter. So where are you seeing change happening? You mentioned that Affordable Care Act was instituted these experiments. You follow these much more closely than most of us do. What experiments seem to be bearing fruit? The ones that are bearing fruit are the idea, three main places. One is in patient engagement. We realize that providers and payers that are more engaged with their patients will have those patients reacting better to whatever they're asking them to do. It seems like you look at me and say, well, no kidding, Shahid, that would seem like an obvious thing to do, but it's not common, right? Because think about the volume-driven fee-for-service world. If you do not show up at your physician's office, they don't get paid, which means they could have cared for you on the phone. They could have cared for you on your iPhone, for example, or an old, plain old telephone system, but they ask you to come in. They ask you, when you need a renewal for your prescription, why do they ask you to come in? I mean, they know what they prescribe. The problem is, if you don't come in, a physician is not going to get paid, and that's the system. So we can't blame the physicians, like, you know, why you're doing it. Isn't that easy to fix? It's not so easy to fix because of the amount of fraud, waste, and abuse that happens in healthcare. We're talking something as low as $50 billion, or some numbers, as high as $300 to $400 billion. So imagine the amount of fraud that happens when Medicare asks people to show up at offices. Now imagine what would happen if they pay for people to be cared for on the phone, or on telemedicine, et cetera, right? So until we figure out how to account for the fraud, waste, and abuse, it's going to be hard to do a massive shift. But the good news is the experiments that we've learned is patient engagement matters. Care coordination matters, meaning that if we have a, like in the old days, we just called it managed care. So in the modern world, it's called care coordination where you have a coordinator, which will make sure you're seeing your imaging provider, then you go to your specialty provider and to a nutritionist, et cetera. So we're seeing patient engagement matters, we're seeing care coordination matters. The third thing we're seeing is population health, management as a whole works. Treat all your diabetics as a whole, and then individually work with them on engagement and core coordination. That works better than just hoping that people show up in episodic care. So there's a lot working. So you talk about fraud, the other side of that coin is cybersecurity. So can you talk about how that might help solve the problem or is security like many of the industries would fall a do over in healthcare? Yeah, I think right now with cybersecurity is such a big issue because of the number of breaches that we've seen in the news that some people are becoming reticent and sharing their medical data, which is getting to be worse because if we don't share the data amongst each other, then we've got a problem. So, the problem- Medical records, as I understand it, are now more valuable to hackers than credit card records. They absolutely are. Some upwards at 10X the value. So if a regular record went for a dollar, medical records would go for 10. And worse than just the value is, with financial records, you can actually have some restitution, right? If your money got stolen, your money will come back. If your records got stolen, let's say your genetic sequences or the fact that you have, even something as innocuous as an allergy to a drug indicates that you've taken that drug and that's why you know that you have that allergy, right? So, when medical records accidentally get out, there's no putting the genie back in the bottle. So it's a big problem- And they're more valuable, explain why they're more valuable. Sorry, we cut you off on your thought and we want to get back to it, but help me understand why they're more valuable because people can perpetuate fraud by having that. Exactly. So if you bought financial records, there's some level of fraud that you can do, but if you buy medical records, then what you can do is like there are mafia, the mob and others who do perpetuate a lot of these frauds, they often will create records that don't match quite well. So the fraud gets caught really quickly, right? So if you're making up synthetic records, it's easier to get caught, easier to get caught, let's say, because there are just as many nice fraud detection systems as there are fraud creation systems. But if you were able to buy, if I was able to buy your two records, that's not synthetic data anymore. You're real people with real credit card numbers, with real social security numbers, that I can say, I can prop up a fake doctor who is going to see both of you, right? And that will be much harder fraud to catch. It's insurance fraud, basically. It's insurance fraud, that's right. That's what this is useful for. You said that ranges from $50 to $300 billion a year. Yeah. Either way, it's huge. It's huge, it's huge. And the reason why that range is there is, the question is, are you talking just about Medicare or about the total, if you're talking about the total industry with over, meaning overuse, like abuse, if you count abuse and overuse in addition to fraud, it goes to $300, $400 billion. If you just say fraud by itself, it might be about $50, $75 billion because that's what the FBI is working towards, mainly is fraud. But then everybody else wants to reduce the cost of healthcare. So, in the credit card world, if I'm in Las Vegas and I fly to Boston and I go to the airport, I have a couple of clam chowder at legal seafoods, I get in a plane, I go to London and they pay for it all with credit cards. They know where I am and then I land in London and my credit card gets denied because they don't have the capacity in the database to actually, maybe IBM can help with that, to actually process that end-to-end transaction, it's getting better. How about, and we've all seen credit card fraud, it's improving the feedback to customers, even though it's somewhat annoying sometimes, the false positives. And so, how does that compare to healthcare? Are the systems rigorous or no? Not at all. So that's a huge opportunity, big problem right now. Very big problem, very big opportunities, also difficult. I mean, think about, if you take a look at a typical transaction in a credit card, maybe it may have five, 10, 15 pieces of information. A typical medical record, say somebody at 30 or 48 years old or above, they're going to have a ton of data. And this is before, by the way, we're talking about genomics and proteomics and metabolomics and all of the things that really tell us about ourselves inside the body. We're just talking about the outside part of the medical record. Healthcare data, it's not commonly referred to this way, but it's got an enormous amount of breath, an enormous amount of depth per person. So not only is the population huge, we're talking about hundreds of millions of people who have these records, but every person's data, by far medical records would be the most data that anyone will ever know about you because there's so much to know. Particularly when your genome is in there. Exactly. I want to ask you about government responsibility in light of a conversation I had last night with a good friend of mine who's a doctor here in Las Vegas, a very high ranking administrator, in fact, in Las Vegas medical community. And the subject of electronic health records came up and he got, there was steam coming out of his ears. He got spitting man. And he said that his colleagues all feel the same way about these things. They hate EHR. And I said, why? Supposedly it's such a good thing, a unified single health record that could be easily shared. And he said, the systems are so damn difficult to use. All these menus and sub-use because there was no incentive for the software developers to make them easy to use because it was shoved down our throats. The government said we had to have this stuff. And so there was no reason to make it useful to us. In light of a comment like that, what is government's role in mandating the use of technologies like this? That's a great question. And when you look at the problem with the electronic health records as a solution or system or software, we have to remember that the reason the physicians are mad as hell about it is because they're being asked to do more administrative work than they've ever had to do before. And they're being now punished if they don't do it on the computer. So now imagine, for example, you and I are having this nice conversation. What if, while we were having this conversation, you had to take meticulous notes about everything that I said. And oh, by the way, if you didn't properly put the notes in, you could be sued because I would say, hey man, you didn't copy what I told you here. Now multiply this by 35 during the day. He's had a typical doctor, may see a specialist will be about 15 to 20, but a primary care physician may see 30 to 35 patients. There is no piece of software on earth that was going to help a knowledge expert because think about a physician is just like us. They're knowledge experts, consultants ask for their advice about care for us. Now, we never walk into any other, like imagine if a CEO of a company or a business unit head had to do all of the CRM data entry that the sales personnel have to do in retail. In any other industry, we will not accept this, that there is a class of people, such as the physicians, the knowledge experts who have to go through all this work. So they're steaming mad for all the right reasons. The problem is blaming the vendors is also crazy. The vendors were asked to produce a electronic health record which is designed to capture documentation so that we can build a government for services rendered on behalf of a patient. So they created a system that did exactly what it was supposed to do, capture data to bill insurance companies. Now, the government and its wisdom got together and said, oh, I guess they're not using these EHRs because we didn't incentivize them. Under what circumstance you have to incentivize anyone to do anything where they want to do it. They don't. So they didn't want to do it because they don't want to be in administration, right? They don't want to be scribes. That's not what they went to school for. So when you look at all of these, what the main thing that the government did probably, the most difficult thing that they had was getting the electronic health records out there. And I think we've seen that many people are using it. Many of them are mad. The difficult part is how can we make it so that we don't tell the vendors and the physician what feature functions need to be in their software? Because this is the mistake that they made. They said there are 32 features required. The first feature has to do this and the second one has to do this and you lose innovation in that, right? What you should basically do is to say, here's the data that I need from your system. For every patient I need X, I need Y, I need Z, you figured out how to do it. I mean, if you want to do it on your iPhone, you do it there. If you want to dictate it, you dictate it. You want to put it into a system. That's your call. I need the electronic data. They didn't do that. And that one decision about focusing on feature and function instead of data, this is, I mean, we're sitting in a data conference here. This is why data is so important. Get the metadata right, get the structures right, to tell people, incentivize them to collect that data in that structure that you need. And in theory, everything else will work itself out. But I completely agree with any physicians about why they're pissed. Government should not be in the software-designed business. That's exactly right. The tail end of that is to create transparency for the individual. So that he or she can see what's going on with their healthcare. It's impossible now to find out. They can't tell you anything about your own situation. Yeah, exactly. And really, when you look at it, if you and I were to go in and ask our physician for our data, the government has done a good thing. They have mandated that if you or I go and ask, they must produce it within, I think it's like 24 or 48 hours. They've even mandated a nice format. So there were parts of this law that did the right thing, which was to liberate, make available at least the liberation of the data, which is good. Like if they didn't do at a minimum that, we wouldn't even be having this conversation. So we've got a good start there. The question is, was it worth 36, 37 billion dollars? Maybe not. So there might have been a cheaper way to do that. But we are where we are. And what we've got to figure out now is if we're going to do outcomes driven and value driven care and move away from volume and fee for service care, what is the minimal set of data that we need and how do we compute the metrics and the measure that tell us that you've achieved reasonable results or achieved reasonable outcomes? That's our next 10 year, 20 year project, which is what you saw this morning. I think in the Watson presentation, they were talking about Watson being able to use machine learning to figure out the patterns. We don't even know the questions to ask Watson, right? Watson was taught with some kinds of information and enough to be able for us to ask it questions. We don't know what questions to ask it. So we need machine learning to tell us. But supposedly the strength of Watson is that it will tell you what questions to ask. Precisely, that's what I'm really looking for. I mean, some of this obviously is hype in that they need to promote, but I think it's good hype. We need to promote the idea that there are machines that can augment and do, you know, the cognitive era that they're pushing is, I like the general terminology because it has some truth to it. We have plenty of information in other sectors, but we are still lacking that information. In fact, in the electronic health care world, we have another problem where even if you have the data, sometimes it's hard to extract it and combine it because you think about the number of physicians you might visit in a regular year, getting them all to connect up and have a unified record, very, very difficult. And so in the case of health care, even though the data is starting to trickle in, only until you put it all together is called longitudinal record. It doesn't have value until it's longitudinal, but then getting to longitudinal is difficult too. Well, and as far as the size of the pie goes, it's essentially a zero sum or even a negative sum game. The size of the pie is going up every year and that's the reason why there's this problem and basically the promise is through efficiency, you'll make the same amount of money, maybe even more, but that's a big if. So the doctors are saying, wait a minute, today I know if I see 35 patients a day, I can determine how much I get paid, you know. And really though, what you have to think about is that they're trying to cover the money because they have to hire extra personnel in the high amount of administration that you have. They could actually, if you think about it, easily on the low end many physicians pay about 10 to 15% of their revenue to these overhead costs, some of them are up 30 to 40%. So there's a plenty of inefficient that we can iron out and still give physicians a good income and revenue that they deserve. They went to school for a long ass time. All right, we have to leave it there all the time. Thanks very much for coming to theCUBE, it was really a pleasure, Shahid, having you on. Sure, appreciate it. All right, keep right there, everybody. Paul and I will be back with our next guest right after this is theCUBE, we're live from IBM Insight 2015. Check out ibmgo.com, we're right back.