Freed: This is Steve Freed with Diabetes in Control. And we’re here with a special guest at the American Diabetes Association, 78th Scientific Sessions, Lisa Latts, who’s presenting I believe. And maybe you can start off with tell us a little bit about yourself.
Latts: Terrific. So, good morning. I am Dr. Lisa Latts. I’m Deputy Health Officer with IBM Watson Health. Internist by training, also have some specialty training in high risk pregnancy.
Freed: And what is the title of your presentation?
Latts: So, I spoke yesterday and what I presented was some analytic work that we had done in our partnership with the American Diabetes Association. IBM Watson Health has a partnership that we announced two years ago actually at this meeting with the American Diabetes Association. And so, among other work that we’re doing, I presented a paper that we had submitted yesterday looking at compliance in patients with diabetes who are on diabetes medication. So, we looked at individuals who were diagnosed with diabetes, who as far as we could tell within the year prior had not been diagnosed with diabetes or treated with diabetes meds. They were started on the med and then we looked at how long they took their medicine. And what we found is that among those individuals who were newly diagnosed with diabetes, 58%, six in ten, were not taking their medicine one year later.
Freed: Is Watson really an artificial intelligence? I mean, who actually does the thinking?
Latts: Yes. So, Watson is really an artificial intelligence. And this is not the robot that you see in Terminator. This is basically, Watson will think — will be able to draw conclusions. So, the thing that Watson can do is it can understand vast amounts of data, so it can read both structured data and unstructured data. So, structured data are things like we used for that claims analysis that I mentioned where it’s reading in claims and we can do things with it because we can do analytics on it. But Watson can also understand unstructured data, so that’s data that’s trapped in the electronic health record or in medical literature, or in places where — if you’re human, you can sit down and read it, but it takes so much time. Watson can read large amounts in seconds and be able to ingest it, and then draw conclusions. And so, the thing about artificial intelligence is that while it can read it and ingest it, it’s like a baby. It doesn’t know what it means, so it’ll draw conclusions but those conclusions may be wrong. So, humans have to then teach it over time just like you would teach a child, “That’s incorrect,” or, “That’s correct.” And then, as it learns more and more, it gets smarter and starts to draw conclusions that are more and more correct over time. And so, we used artificial intelligence in our partnership with Medtronic to develop Sugar.IQ which is in exclusive use with their Guardian Connect Continuous Glucose Monitor.
Freed: So, I have an interesting question and I don’t know if Watson can answer.
Freed: But if you go back to 1950, we had one medication for type 2 diabetes, sulfonylureas. And then it took us 50 years to get a second medication called metformin. So, it took us all that time to get two. As of today we have 6 million possible combinations. So, for the family practitioner to sit down with a patient and come up with the right possible combination is basically impossible.
Freed: That’s where something I would imagine like Watson can actually calculate, put their insurance in, and actually calculate what’s the most effective treatment at a reasonable cost. How far are you guys away from that?
Latts: So, we don’t have that kind of tool for diabetes yet but we do have something very similar for oncology where we have a tool that we developed in partnership with Memorial Sloan Kettering in New York, and what the oncology told us is it takes in all that patient information, as you mentioned, although we don’t have insurance in there currently. And it reads through the individual’s characteristics, looks at the individual cancer, and then comes up with recommendations for both chemotherapy, endocrine therapy, and radiation therapy based on the expert training that it received from MSK. So, we have that sort of thing for oncology. We do not have it for diabetes yet.
Freed: What’s the future?
Latts: So, I think the future is more personalized medicine. So, I 100% agree with you that right now — I’m a physician as well as — it’s kind of a shotgun approach and so you go on what you think what’s in the literature. But the literature is all based on these large studies with large populations. And so, as we get smarter and smarter and know more about individual characteristics and how social determinants of health fit in, and all the other things that make us who we are as people, we’re going to start to make better choices, I agree, and Watson can be very helpful there. We’re sponsoring a panel tomorrow night in partnership with the ADA and leaders from various segments of the diabetes community to talk about health equity and disparities in diabetes care, because one of the other things we see as I’m sure you’re aware is that diabetes care is not equal across all racial and ethnic groups. And so, there’s a huge opportunity to get smarter about how we use diabetes resources for different populations.
Freed: Is there something new for Watson this year besides what you just mentioned for claims?
Latts: So, Medtronic announced that their Guardian Connect and Sugar.IQ has gone live. And they announced a study that we just released, looking at individuals with diabetes who are on the Guardian Connect and Sugar.IQ app for 90 days and found that those individuals were far more likely to be in range in terms of their sugar. So, they had 36 more minutes in range on average per day, so 33 minutes less hyperglycemic and then less time hypoglycemic. So, again, using this artificial intelligence system is really the ability to help patients make more informed decisions about their own care in diabetes because diabetes is fundamentally a decision where patients are by themselves day after day having to decide, “What do I eat? What insulin do I take? When do I take it? What kind of exercise do I get?” And that’s really hard to make all these decisions day after day after day. And so, the goal of Sugar.IQ is to help individuals make better and more informed decisions because they have the data coming in that will help them understand how previous behaviors have led to their sugar changing.
Freed: Is that program available right now?
Latts: It is now. Yeah, absolutely.
Freed: So, how does it actually work? The patient comes into the office….
Latts: So, it takes in sugar data from the Guardian Connect monitor. It goes through an app. You enter your food that you’re taking and then it gives you insights based on previous patterns and analysis, looking at — so, for example, we were talking to one woman yesterday who had been using it and she had found that she had trouble with coffees. So, when she had a Starbucks drink or a coffee drink that her blood sugars were going all over the place. And so, she was using Sugar.IQ to help give her insight into, “Why is it that when I have this latte, which should be predictable in terms of its effect on my blood sugar, it’s variable?” And so, Watson helped give her insight that, “Wow, when I have it in on the weekend and I’m relaxed, I’m not having problems. When I have it during the week and I’m stressed, that’s when my blood sugar is spiking.”
Freed: So, how is it available? Is it an app?
Latts: It’s an app. It’s an app for the IOS system available to individuals who are using the Guardian Connect Continuous Glucose Monitor.
Freed: And what data supports the Sugar.IQ?
Latts: So, what we announced on Friday was the launch of this and the Sugar.IQ and this 90-day study that show that individuals have more time in range as a result of using Sugar.IQ with Guardian Connect.
Freed: And when was it made available?
Latts: So, it was FDA approved in February and I believe was officially launched and available commercially in June.
Freed: And currently how is it used?
Latts: So, it’s available for anyone, type 1 diabetes for the most part, who are getting multiple daily injections and want to use this as a continuous glucose monitor.
Freed: How long has Watson been around?
Latts: So, Watson the artificial intelligence computer was launched with Jeopardy — do you remember the Jeopardy? Yeah, so it was launched with the Jeopardy game. Watson Health as a separate entity was April, three years ago. So, Watson Health, the organization, is about three and a couple months old.
Freed: Are you presenting any other lesson or information here at ADA?
Latts: Yeah. Thanks for asking. So, we’ve got the Guardian Connect, Sugar.IQ, and we’ve got an analytic study that I presented yesterday looking at adherence for individuals with diabetes. We’ve also got a poster that uses machine learning to look at how different classes of diabetes drugs affect cardiovascular outcomes. So, looking at acute myocardial infarction, looking at congestive heart failure, looking at stroke, so looking at those outcomes by different classes of diabetes drugs. And then we’ve also got a proof of concept that we’re doing with the American Diabetes Association as part of this partnership where we are looking at, “Can Watson help make available to researchers and physicians all of this amazing information that’s being delivered in a video format but that is incredibly difficult to find?” So, if you’re a scientist and you’re like, “When I was at the ADA last summer, I know that someone was talking about the effect of heavy metals in diabetes but I can’t remember what it was.” So, the idea behind this proof of concept is that you would ask the question in Watson and it would pop up not just the correct video but actually the clip within the video where they answer that question. So, we’ve tested a proof of concept based on ingesting 90 hours of video. And now we’ve done some preliminary testing. We’ve got very favorable opinions and we’re continuing to improve and continuing to build. And so, now we’ll be looking at developing a beta test and then looking at developing some sponsorship for this app to make it available for scientists and other researchers.
Freed: So, how does the cardiovascular aspect of it work?
Latts: So, for the cardiovascular study, we looked at the different classes of drugs and we looked at in the year after someone was first prescribed these drugs, what happened to them from a cardiovascular standpoint; were they hospitalized for a heart attack, were they hospitalized for stroke, did they develop a diagnosis of congestive heart failure. And then, we group the classes together and so we acknowledge that there might be some drugs within a class better than others. And we were just looking at the whole classes and what we found is that both the DPP4 class and the SGLT2 inhibitor class had lower or fewer cardiovascular adverse events than the other classes of diabetes drugs. So, both of these from our study at least seem to be cardioprotective as a class effect than the other classes of diabetes drugs.
Freed: And how does a physician or family practitioner use it?
Latts: So, I think there’s a couple of ways. One is that if you have someone who either has cardiovascular disease or is at strong risk for cardiovascular disease maybe besides their diabetes, they’ve got a strong family history, they’ve got high cholesterol that you’re having trouble getting down. And as you mentioned earlier there are so many individual decisions that you have to make in choosing a class of diabetes agents for your patients, having this information may help tip you over to — my patients at risk for cardiovascular disease, I may choose something that perhaps will confer some cardioprotectivity and go for one of these classes of drugs to help me inform my choice.
Freed: It sounds like it’s kind of general information helping you to make a decision as to what drug to use. As I mentioned earlier, is it ever going to get down to the point where it can actually recommend or give you a couple of recommendations? And I know for some physicians that could be tough because you’re taking their knowledge and their ability to make the decision away from them.
Latts: Yeah, yeah.
Freed: And that could be a little scary.
Latts: Absolutely. And Watson is never going to make the decision. Watson is only going to make recommendations based on the best available knowledge. There are so much information being created today. It’s been estimated that 90% of the digital information out there was created in the last two years. And so, when you think about that there’s no way that a human can keep track of all that data and all that information. So, the idea behind Watson is that you make it easier for individuals to access the information that they need at the time that they need it. And so, that’s what it’s going to do, is surface the right information at the right time to help clinicians make the right or better informed decision for their patients.
Freed: So, you’re a physician. Do you still see patients?
Latts: I do. Yeah.
Freed: And you decided to go into a little bit different direction.
Freed: Why did you do that?
Latts: So, my background and my training was in public health. And so, I love being a practicing physician. And I see pregnant patients with medical problems, which is a very unique niche but my interest has always been, how do we make the health care system better, because it’s incredibly hard to be a practicing physician day in and day out. And so, how can we develop tools and strategies that make it easier to practice and be practicing physician. We’re seeing a huge problem in the physician community these days with physician burnout and lots of physicians are leaving the practice of medicine. And so, we’re going to face an enormous problem with a shortage of physicians if we don’t figure out ways to make it easier to practice medicine. And so, that was one of my goals, is how do we make the health care system easier, how do we make it work better so that we can deliver more care to individuals and more appropriate care without health care cost rising out of control. Because in the US, we as a society cannot continue to see the kind of rising cost that we’ve seen in the past, and we need to be better about how we’re using our health care dollars, so that’s what motivated me to go into more the administrative side of medicine as opposed to seeing patients full-time.
Freed: The science, Sugar.IQ, because that use is limited, where can family practitioners and nurse practitioners find out what’s available that they can use today, how would they go about finding out what’s available and how they can access it?
Latts: So, in Watson we’ve got lots of areas of diabetes that we’re working on currently. For a family physician coming to things like the ADA may be possible but I think getting information from aggregate sources that put it together in a way and show them the highlights and the important things that they need to know for their practice, right now is where we’re at. But in the future, I think there will be more and more tools that help surface these insights that will help individuals make better decisions based on their individual patient characteristics.
Freed: So, what’s available now that they could possibly use in their office practice?
Latts: So, there are definitely tools out there that help you find sort of summary information. For example, we’ve got a tool that we use within IBM Watson Health called Micromedex that helps surface the best information from a pharmaceutical perspective, looking at different drugs and how they might be used and helping you figure out how to use those drugs in the most appropriate way. And so, there’s tools like that that you can access to and then help you at the point of care get the best information you can today.
Freed: So, for my final question, if this is viewed by pharmacists, family practitioners, nurse practitioners, dietitians, what’s the takeaway message?
Latts: So, I think the takeaway message is that what’s coming is the ability to have at the point of care better insights and more information to help you help your patients make diabetes care easier. So, it’s coming. We’re right on the cusp of that. We’re learning so much about diabetes. You just look at the size of this meeting at the ADA that we’re at, which is enormous. I heard yesterday that 60% of the presenters here are non-US based, so it’s not just the innovations that’s coming out of the United States, it’s worldwide innovation in diabetes. And so, there’s so much changing and so much more than we know now than we knew a year ago, much less 10 or 20 years ago So, it’s incredible changes are coming and you have many more options today than you have yesterday. So, being able to understand these things and get the best information you can for your patients will help your patients have better outcomes.