SXSW Interactive 2015

March 13-17, 2015; Austin, TX; Days #1-3 Highlights - Draft

Executive Highlights

Greetings from Austin, Texas and SXSW Interactive. We’ve been dodging past Miller Lite beer stands, tech startups handing out American Apparel t-shirts, and posters advertising the latest apps (e.g., Hater) to hear about digital health and medical technology. This track is much smaller than larger non-healthcare areas like “startup village”, “design and development”, “global impact & policy,” but pretty much everything in healthcare we’ve attended has been standing-room-only. Most of the sessions have taken place in the large Austin Convention Center or the smaller JW Marriott conference center; jeans, tshirts, and sneakers are standard for this crowd.

Like CES Day #1 in January, there is serious optimism about the future potential of technology to change healthcare delivery. But remarks have also laid down clear challenges for healthcare in general: patients don’t want to pay for anything, behavior change remains highly challenging, providers can’t deal with the data (along with worries over about liability and losing authority), payers don’t have the right incentives, our data is siloed and not interoperable, and the best engineers aren’t coming into healthcare. These remarks have spoken generally to healthcare, but certainly apply to diabetes at a high level. Still, most here are enthusiastic that an important foundation is being laid – ACA is making digital health more viable; excellent design and user experiences will hopefully improve engagement and outcomes; and some of the most say oft-cited companies like Omada, Proteus, and AliveCor are proving that some of the great challenges can be addressed (prediabetes, adherence, democratizing EKG scans). Below, we show our top ten highlights from the first half of the conference, followed by detailed discussion and commentary from sessions and panel presentations.

1. A valuable panel on the consumerization of healthcare featured Livongo Health CEO Glen Tullman, myfitnesspal co-founder Mike Lee, and Spruce Health (telemedicine) CEO Ray Bradford. All spoke to the growing financial burden on the consumer (ACA Bronze plans), the expansion of remote monitoring/telemedicine, the need for better design, engagement strategies to avoid, and more. This was an excellent panel for the honest commentary it generated.

2. Charismatic New York Times bestselling author (“DRIVE”, “To Sell is Human”) Dan Pink shared seven tactics for changing behavior. We were moved by all and found his discussion of “using questions” and “making it easy to act” to be particularly valuable.

3. A panel discussion featuring big thinkers in the data field – Scripps’ Dr. Eric Topol and Scanadu’s Mr. Walter De Brouwer – tackled a pressing and provocative question: Is big data the next wonder drug? Both Dr. Topol thinks and Mr. De Brouwer were highly optimistic. Lofty remarks from the brilliant panelists did not change the big takeaway – which we might summarize as #WeAreSTILLwaiting...

4. Dr. Thomas Morrow (Chief Medical Officer, Next IT) gave a fascinating, compelling overview of virtual health assistants (VHAs), software that combines artificial intelligence and natural language processing to improve patient engagement and outcomes. 

5. A panel discussion on the interaction between pharma and tech startups emphasized that startups need to solve a very specific problem, while pharma needs startup’s innovation to go beyond just another pill.

6. Proteus Digital Health’s Chief Product Officer David O'Reilly boldly challenged the entire field of digital health with one bedrock design principle: make it invisible. Of course, Proteus has done it already! There was no commentary on potential diabetes applications.

7. A standing-room-only conversation with author Mr. Malcolm Gladwell and Mr. Bill Gurley (the venture investor behind Uber) touched briefly on healthcare reform amidst a broader discussion on innovation. Mr. Gurley called out misguided efforts to reform electronic health records, while Mr. Gladwell questioned whether the growth of digital technology is actually undermining the patient-provider relationship.

8. Two panels on healthcare consumers and achieving scale  shared similar themes on where digital health is heading: the best business models leverage the existing system, care will be democratized through the smartphone, the balance between technology and humans will be tough to strike, providers are overloaded with data, and more.

9. A WebMD-sponsored panel on the pitfalls of too much health data shared perspectives from ER Physician Dr. Robert Glatter, Atrial fibrillation advocate Mr. Shannon Dickson, and WebMD’s Chief Medical Editor Dr. Michael Smith.

10. A lively session on patient privacy (Including Medtronic Diabetes’ Amanda Sheldon and diabetes advocate Mr. Manny Hernandez) discussed some of the challenges in the modern digital era – the need to deal with HIPAA, despite how old the law is.

Top 10 Highlights

1. A valuable panel on the consumerization of healthcare featured Livongo Health CEO Glen Tullman, myfitnesspal co-founder Mike Lee, and Spruce Health CEO Ray Bradford (online dermatology consultations). All spoke to the growing need to focus on the consumer, given the current trends: growth in high deductible health plans (millions are now on the ACA’s bronze plans, where the first $5,000 is out of pocket), the ubiquity of smartphones, growing government reimbursement of telehealth and remote monitoring (these happened recently but did not receive much press), and widespread frustration that healthcare is inconvenient, wasteful, and hard to understand. Similar to commentary from CES, all appeared nonetheless excited about the future, particularly remote care enabled by technology. Said Mr. Bradford: “We are in the first innings. Less than 1% of doctor’s visits are occurring online. We see a future where more than 50% of interactions occur online.” This could be fairly optimistic coming from someone who runs an online dermatology telemedicine service, but is nonetheless instructive and in line with what we heard from Kaiser at JPM 2013. The panel also offered lots of perspective on how to design apps that people want to use, as well as the types of feedback and engagement strategies to avoid. We’ve transcribed it in full below! We felt this was the strongest panel in terms of useful, actionable remarks for companies and how to develop products that align with some of the key trends the industry is facing.

2. Charismatic New York Times bestselling author Dan Pink shared seven tactics for changing behavior in one SXSW’s featured sessions. His entertaining, example-filled talk was grounded in both behavioral science studies and his experience filming his new National Geographic show, Crowd Control. The seven tips for changing behavior are outlined below in more detail, along with possible implications for diabetes/obesity: (i) use fear the right way (when narrowed focus is needed); (ii) Questions (often) beat answers for changing behavior (when the facts are on your side); (iii) use social proof – people look to other for cues about how to behave; (iv) use rhyming, alliteration, and repetition; (v) make it easy for people to act; (vi) put a human face on things; and (vii) experiment and try stuff. Though the large conference auditorium was only about half full, we thought this talk was a truly outstanding overview of tactical approaches to changing behavior – we hope to see many more of them tried in diabetes/obesity! We were very happy to hear Mr. Pink answer our question on diabetes/obesity via Twitter. More details below, and see a very cool graphic recorder summary of the talk here – we like this because it really captures the spirit of the talk in a very pictorial format.

3. A panel on the future of big data tackled a provocative question: Is big data the next wonder drug? Panelists opened by emphasizing their high enthusiasm for the potential of big data, though made it clear that we are still in early days – “We’re probably at 0.1% of 100% of what we are going to know about our health.” This emphasis on the untapped possibilities in digital health was a theme throughout the session, which was headlined by both Dr. Eric Topol (Scripps Translational Science Institute, La Jolla, CA) and Mr. Walter De Brouwer (CEO, Scanadu, Moffett Field, CA). While such technologies will power the new age of big data, panelists were quick to point out that collecting data is not the issue – instead, the challenge is running analytics to generate actionable recommendations for patients and providers and, in the words of Dr. Topol, stealing the engineers capable of doing this from “Twitter and Pandora.” Ultimately, the panel reinforced how much generalized confidence there is in the future potential to shift health using data, but we felt the session was more notable for the questions it continued to raise about big data: (i) How do we actually use data to improve our future health? (ii) What happens when data is extended from the individual to the population level? (iii) Who owns the data and where should it be stored? and (iv) When is Big Data actually going to deliver on the promise, particularly in diabetes? It was valuable to hear from such notable thinkers, but we were somewhat deflated by the generalized discussion – there is nothing truly notable to report on yet. For the most quotable quotes from this discussion, please see the detailed discussion below.

  • Mr. Walter De Brouwer is an especially prominent thinker in this field, best known for founding Scanadu – the early stage company is developing Scout, a scanner placed on the forehead that provides information about body temperature, heart rate, blood pressure, and more. It’s very sleek, very cool, and currently under FDA review. The usual questions on regulatory and reimbursement still exist. More broadly, Scanadu has been named a finalist in the $10 million Qualcomm Tricorder XPRIZE to accurately diagnose 16 medical conditions with one device.

4. Dr. Thomas Morrow (Chief Medical Officer, Next IT) gave a fascinating, compelling overview of virtual health assistants (VHAs), software that combines artificial intelligence and natural language processing to improve patient adherence, change health behaviors, educate, and motivate patients – see Next IT’s illuminating two-minute video on Alme Health Coach here. Notably, VHAs are trained to know clinical guidelines, handle adverse events, and even ask patients questions – all at a much lowest cost than people. In one pilot hospital study, 74% of participants actually said they would rather interact with an intelligent virtual assistant than a human being! Dr. Morrow’s talk made a strong case for the vast potential of VHAs to help overcome many of the concerning healthcare trends of our time: non-adherence, difficulty of behavior change, the shortage of providers, and the high cost of effective engagement strategies (e.g., office visits, phone calls). It’s early days for these programs, and there are only a couple on the market: Next IT’s Alme Health Coach for chronic disease (we were not able to download it) and MedRespond’s health assistant for end of life questions (we cannot find anything online about this). Dr. Morrow mentioned diabetes and prediabetes many times, implying both areas are ripe for such approaches. However, anything that ventures into the realm of medication titration (very doable with today’s technology) will require FDA approval. Though the talk had a science fiction feel to it, we left moderately optimistic that VHAs could be useful for some patients – of course, they won’t be for everyone, a point made clear in the Alme Health Coach YouTube video’s comments.

5. A panel discussion on the interaction between pharma and tech startups approached the issue from several different angles, sharing the challenges and keys to success for each side. We heard a startup perspective (Cohero Health), a Big Pharma perspective (Boehringer Ingelheim), and an investor perspective (Mohr Davidow Ventures). Panelists emphasized the need for startups to solve a very specific problem that pharma has, such as a particular medication going off patent or patients not adhering to medications. However, it also became clear how challenging it is for startups to even get Big Pharma’s attention – even large companies like BI have major bandwidth issues (“150 cold emails per week,” said the member of senior management there who was on the panel)! But the ultimate punchline is that these two entities need to work together – startups need Big Pharma’s scale/funding/validation, while Big Pharma needs startups’ innovation and fresh thinking (presumably as drugs come off patents). The latter has been a major theme of many panels through SXSW thus far, with speakers calling for a value-added, pill-plus approaches to pharmaceuticals. Said BI’s Megan Lopresto, “I think in less than 20 years, pharma is not going to exist in the traditional sense any longer. With biosimilars and generics, it may not even be profitable to continue to market and sell the way that we do. There must be a value-add.” Quotable quotes appear below.

6. Proteus Digital Health’s Chief Product Officer David O'Reilly headlined a foursome of 15-minute talks, challenging the entire field of digital health with one bedrock design principle: make it invisible. He argued that most current digital health tools are status symbols, appeal to the worried well, and have a high drop-off rate – consequently, he said, the industry “is failing” to provide devices that fade into the background and are private, intimate, and unobtrusive. Mr. O’Reilly very briefly discussed the work at Proteus, which embeds these design approaches into its poppy seed-sized ingestible sensor that can be embedded into pills. Once swallowed, the sensor communicates with a disposable, discreet, body-worn patch, offering biometric data and medication compliance information on a smartphone app. We also loved talks in this session on technology’s ability to foster resilience, health literacy (some truly daunting stats), and Mylan’s CIO Mr. Michael Smith on the three key barriers to global healthcare access. All are summarized below.

7. A standing-room-only keynote session with New York Times bestselling author Mr. Malcolm Gladwell and the venture investor behind Uber, Mr. Bill Gurley (Benchmark, Menlo Park, CA), opened with a brief debate on the status of healthcare reform efforts. We were excited to hear such prominent thinkers share their thoughts on the field, though were not particularly surprised by their overwhelming message: “Healthcare is mucked.” (We heard that phrase at least three times in the first ten minutes of the conversation alone!) Mr. Gurley became particularly animated in calling out misguided efforts to reform electronic health records systems, citing inefficiencies on this front as characteristic of broader reform efforts – this was fairly depressing if not surprising. He noted that for the past five years a US government-sponsored program (Health Information Technology for Economic and Clinical Health Act, 2009) has monetarily incentivized providers to implement EHRs in their clinic (essentially providing a “$44,000 bonus” for those who have complied). In his view, this incentive program has generated $29 billion in government spending for doctors to implement software that they do not want nor necessarily even use – “It’s the exact opposite of what entrepreneurship is supposed to be.” Of course, there are serious merits to EMRs – and we commend the government for promoting them nationwide – but they are clearly falling short on the interoperability front.

  • Mr. Gladwell took a more unconventional stance, questioning whether the growth of digital technology is actually undermining the patient-provider relationship. As screens and gadgets infiltrate the clinic, he suggested that technology is moving us away from the face-to-face interaction that patients truly seek. To an extent, he was playing devil’s advocate – likely in light of the big data enthusiasm at this meeting – though we do believe he has a point. After all, part of the multitude of problems in diabetes care is dealing with a growing patient population coupled with a shortage of providers, resulting in less time for patient care. When that dwindling appointment time is spent staring at a screen, it is no wonder that patients feel lonely and unsupported. Ultimately, neither panelist offered a compelling resolution to the debate and the sense of frustration – almost fatalism – that seemingly touches all discussion of healthcare reform was best captured by Mr. Gladwell’s closing question: “Well Bill, do you have anything hopeful to share?” He did not.

8. Two panels, “Radical Healthcare: What do consumers want?” and “Your Health Startup Is Cute: Gotta Play at Scale” shared similar themes on the somewhat underwhelming current status and exciting potential future of digital health. We heard lots on the growing trend of consumers paying for more healthcare (though correspondingly, a general low willingness to pay for anything), barriers in healthcare (aligning incentives came up a lot, as did user-centered design), business models and achieving scale (leverage the existing system), the democratization of care (through the smartphone – the Economist says 80% of the world will have smartphones by 2020), the balance between technology/automation and humans (hopefully synergistic!), the provider challenges of dealing with all the data (they cannot!), changing behavior (Spoiler Alert: It’s HARD), and the need for more engineering/programming talent (everyone is going to Silicon Valley tech companies). Below, we have included quotable quotes from these two excellent panels, which included perspectives from Mango Health (medication reminder app, via gamification); Lantern (emotional well-being assessment); Kinsights (advice sharing network for parents); Wellframe (risk management); Jawbone (activity tracker); ElevateHealth (an HMO), PillPack (full service, mail order pharmacy that intuitively packages medications – this is a very exciting approach, particularly for those with high pill burden); and The White House’s Deputy Chief Technology Officer (the very young Ryan Panchadsaram, who helped turn around).

9. A WebMD-sponsored panel on the pitfalls of too much health data shared perspectives from ER Physician Dr. Robert Glatter, Atrial fibrillation advocate Mr. Shannon Dickson, and WebMD’s Chief Medical Editor Dr. Michael Smith. The session began with a video on wireless medicine, one of a five-part WebMD series on the future of health (the others are infertility breakthroughs, 3-D printing, obesity treatments, and advances in vision). Said Dr. Eric Topol on the threat wireless medicine poses to the hierarchical medical establishment, “We knew medicine was getting digital. But we didn’t know if was getting democratized.” This theme continued in the panel discussion, which was headlined by emergency room physicians Dr. Robert Glatter. He emphasized that doctors are not equipped to deal with all this data – some actively resent it – and healthcare should be thinking about data management positions to handle the deluge. In addition, there are potential liability concerns if providers miss data that is concerning, and worried patients compound the program by sending unnecessary data in. Manufacturers can help on this front, a point emphasized by Afib advocate Mr. Shannon Dickson – he highlighted digital health poster child AliveCor’s handheld EKG device/app, which now includes algorithms that automatically analyze the readouts. Notably, Dr. Robert Glatter highlighted WellDoc’s BlueStar as one example of a valuable, FDA approved digital health solution that has helped his patients with type 2 diabetes. The panel discussion is enclosed below.

10. A lively session on patient privacy brought together key stakeholders to discuss the promises and risks of digital health, including Medtronic Diabetes’ Amanda Sheldon and diabetes advocate Mr. Manny Hernandez. Though we often hear about the vast potential of our digital footprint, this debate was grounded far more in the consequences of the digital health explosion with a keen eye toward protecting patients. We heard concerns that the risks associated with confidentiality and integrity of data have grown exponentially of late. Panelists called for more transparent and patient-centered processes, though the question of how to drive toward these goals remains wide open. We continue to wonder about the modern relevance of HIPAA, since the bill was signed into law two decades ago in the Clinton administration. There’s no question that protecting health data is critical, but most patients gloss over user agreements and will happily agree to share data. Meanwhile, concerns over privacy slow down development and regulation. Ultimately, the debate remains quite nuanced and to provide a flavor of the discussion, we have pulled out some of the most notable quotes immediately below:

  • “Data surveillance is a $156 billion industry. That industry is worth twice the money the US Federal government spends on intelligence ($68 billion).” – Ms. Jane Sarasohn-Kahn (Health Populi, Philadelphia, PA)
  • “Cybersecurity comes down to ethics as an industry and putting care first and putting patients first.” – Ms. Amanda Sheldon (Medtronic Diabetes, Los Angeles, CA)
  • “Coming from J&J, I can say that the companies that are going to survive in the long term will be the ones that hold themselves to the highest security standard.” – Mr. Marc Monseau (Mint Collective, Princeton, NJ)
  • “Companies that are more cavalier about data security are putting themselves at risk because patients are going to go with the ones that they trust.” – Mr. Manny Hernandez (Founder, Diabetes Hands Foundation, Berkeley, CA)
  • “There are a lot of data flows outside the scope of HIPAA. HIPAA does not help you out in a lot of instances.” – Ms. Sarasohn-Kahn
  • “The most fundamental element is that data sharing decisions are made with the patient first and foremost in mind.” – Mr. Hernandez

Detailed Discussion and Commentary

Digital Health and Consumerization of Healthcare

Panel Discussion

Glen Tullman (CEO, Livongo Health, Mountain View, CA), Mike Lee (CEO, The Future Market, Brooklyn, NY), Nathan Olivarez-Giles (Technology Reporter, The Wall Street Journal, New York, NY), Ray Bradford (CEO, Spruce Health, San Francisco, CA)

Mr. Olivarez-Giles: We have a fantastic panel on making health more accessible and more friendly to the consumer. Can you guys introduce yourselves?

Mr. Lee: I’m the co-founder of MyFitnessPal. We live in a world where it is easier to be unhealthy than healthy. We want to change that. We have a really popular app to track what you eat and your exercise. And we’ve made it easier than ever before. We have over 80 million members. Most people use it to lose weight. Our community has lost over 200 million pounds combined. That’s more than the weight of the population of Las Vegas. I don’t know how you feel about Las Vegas, but it’s gone. [Laughter]

Mr. Tullman: We are all about empowering people with chronic disease to live better. We combine smart devices with a smart cloud, and ultimately, smart people. We focus on remote patient monitoring and keeping in touch with people. If you have OnStar for your car and Nest for  your house, you understand remote monitoring. We have thousands using our product today. Ultimately, we want to give back the power of healthcare to people. We’re seeing this consumerization. People don’t want to be engaged. No one wants to be engaged. If you are well, you don’t want to be engaged. If you are sick, the last thing you want to do is be engaged. So we’re about empowering people, starting off with diabetes.

Mr. Bradford: Our mission is a better way to see a doctor. Now, we have the best way of seeing a board certified dermatologist. It’s a comprehensive software experience. You take photos of your skin and answer questions. Then, you pay $40, and we take HSA and FSA. That’s where the comprehensive experience kicks in. We have a network of affiliated dermatologists that have software on iPads. Prescriptions are sent digitally to your preferred pharmacy. A care team checks in to make sure you are getting results.

Mr. Olivarez-Giles: Where do you think we’re at today? We have a lot of apps that claim to help you live healthier. A lot of them, frankly, just suck. Many claim to be fitness coaches. It feels like the very early days of something that is going to happen. Lots of stuff you are trying to do now wasn’t possible years ago. Livongo is a cloud-connected glucometer. You’re feeding in real-time information from the glucometer into the cloud. We’ve heard about virtual reality for years, but it’s still not in our living rooms. How far are we?

Mr. Lee: I’m not very good at predicting the future. But I’m pretty sure we’re just getting started. There are a lot of foundational things that are happening in the digitization of healthcare. The proliferation of smartphones – it’s a powerful computing device that people check 24/7. We have wearables on the market, and even devices that can perform diagnostic tests at home. We have data connection between the phone and devices and EMR systems. Those are all things that have to happen before we have the kind of experience that could be really impactful. I’ve heard stats that you go to see the doctor, you wait 40 minutes, and then you spend five minutes with the doctor. That’s a terrible experience. In the future, you will do your own diagnostic tests, which will automatically be transferred to the doctor. That’s a much more seamless and user-friendly experience. But it’s going to take time.

Mr. Tullman: You’ve done a great job of predicting the future. There are a few things that are different now. One, we didn’t have the technology. Now we’re starting to get it. Second, we didn’t have the impetus. Now that’s coming through major trends. We’re out of money. We ‘re also seeing the consumerization of healthcare. People are saying, “Why can’t my healthcare work like every other aspect of life?” Healthcare and education are the two areas where we kept technology away, and that’s changing in a very rapid way. And last, we have Obamacare – whether you love or hate it, the end result is we have millions of people on bronze plans. The first $5,000 they spend is their own money. They’ve just become consumers. They’re going to make better healthcare decisions. They are armed with things we used to call phones. Now, they’re little computers.

Mr. Bradford: We are in the first innings. Less than 1% of doctor’s visits are occurring online. We see a future where more than 50% of interactions occur online. The timing and how quickly we get there is hard to predict. But we underestimate the percentage of care that can be done remotely. There is a 30-day wait time to see a dermatologist in America. You can do that in under 24 hours on the phone. It’s better, faster, and cheaper when it comes to using technology. We have the ubiquity of mobile devices. The camera itself is a sensor. And the incentives are here – there’s been a tripling in the number of high deductible health plans. The average deductible is $1,200. You care because it’s your money. You want a great customer experience at an affordable price.

Mr. Tullman: The government has done two things recently that didn’t get much press. They are now reimbursing telehealth. And second, they are now reimbursing remote patient monitoring as long as it’s for two or more chronic conditions. We have payment changes combined with technology and consumerization. We haven’t seen that before. I think we’re in for some exciting times.

Mr. Olivarez-Giles: That sounds hopeful. However, it’s clear that there are a lot of hurdles we need to clear first. I think there needs to be a conversation amongst ourselves as to what data and information we’re comfortable giving up to guys like you. You need to make that as clear as possible to us. All too often with where we’ve gone with mobile technology, those conversations haven’t happened yet. We give our locations to all our apps. It’s not always clear what’s happening with that stuff. Healthcare is a very sensitive and very personal space. I don’t always want everyone to know how I’m always feeling. How are you guys addressing this? How are you making people feel comfortable and transparent about what you are doing?

Mr. Bradford: It’s about what your incentives are. If you’re seeking medical advice and you don’t have trust in the provider, it’s going to be a problem. We’ve got to be transparent about how we’re using data. Beyond that, it’s about looking around the table and thinking whether as a consumer I would be comfortable with what we are doing. You feel often times in healthcare that you don’t know who’s looking out for you. We want you to feel as though we’re in your corner. You talked earlier about doctors making evidence-based decisions in the context of care. You want your doctor to be looking at a broad set of data to be making decisions, so that’s where we come in.

Mr. Tullman: I came to Livongo from running one of the largest EMR companies in the country. The key is that it’s all about patient choice. We always say at Livongo that it’s all about patient authorization. They decide who can see the data. We have to get back to patients knowingly giving their data. We try to simplify it. There can’t be a complicated document with nine pages attached to it to sign away their data. You’re going to lose that trust, and if you lose it, you’re going to see whole companies impacted by it. I think going forward apps are going to change to reflect this and it will affect spending models. People are going to have to pay for the value they get from an app.

Mr. Olivarez-Giles: What will it look like when the consumer loses faith?  It seems like we have a data breach every day. Target had a data breach, and I think Anthem had something similar recently. Do people really care? We should honestly care. But from my perspective as a reporter, I haven’t seen people going out of business.

Mr. Tullman: If someone hacks Target or Anthem, that’s different from somebody knowingly selling your data on a regular basis. You have to really go overboard to try to secure the data. What I’m talking about it is someone selling the data – you give the product away for free on the front end and are selling the data on the back end.

Mr. Lee: A good example of people caring about privacy is Moves, a great app. Facebook bought it. As soon as that happened, the ratings went to one star, and it has basically disappeared. We need data in order to help our customers succeed. If they don’t give data, we can’t do that. But they have to trust us. It’s not just a moral imperative. It’s business critical. We have to get this right. It’s something we really care about.

Mr. Olivarez-Giles: How do you build that patient-company relationship? This is sort of a new kind of relationship we’re trying to have with our health in general. You guys are trying to build this relationship with consumers.

Mr. Bradford: We discussed earlier about how talking about “engagement” does not always make sense. We don’t want you to use the app just for the app. We want to be there when you want it. Think about it this way: When your doctor doesn’t have time for a you and runs out of the room and throws paper at you and you’re lost, we want to be there. We don’t want engagement for it’s own sake. We want you to be able to get back your life in a very convenient way.

Mr. Tullman: Think about a time you’ve been sitting with a family member and have been confused with a medical question. Imagine if the phone rang at the moment. What would you say? You’d say, “Perfect timing.” That’s what we’re trying to create. Because of population tracking ability, when a patient has a high or low blood sugar, their phone will ring. And it works because that’s when patients are receptive to care. However, when you call them in the middle of the day and say, “how’s it going with managing your chronic disease,” they are going to be annoyed. So it’s about timing. But the system works. I would just point out that there’s a system like this at a UMass clinic. Typically, healthcare models reward healthy patients with better reimbursement and penalize unhealthy patients with worse reimbursement. So who do you think comes into the office? So now, this reformed system identifies the people in better health and says that they shouldn’t come in to the clinic as often. However, they also get on the phone with people in worse health and encourage them to come in. It’s a more efficient system.

Mr. Olivarez-Giles: What lessons have you learned in trying to build your products?

Mr. Lee: We’ve learned a lot of lessons. One thing we think about at the company is that we are dealing with people. The data is incredibly powerful, and there are so many things to do to help users succeed. But there are people on the other end of the digits. In terms of getting people to change behavior – a lot of it is art and less science. We had a feature idea to rate foods. Take what you ate and give it grades. We ran it by bunch of users and they hated the idea. Everybody knows chocolate cake is an F. The issue isn’t that knowledge. It’s that chocolate cake tastes good. How can we do this differently? Your app shouldn’t make people feel bad. We’ve come up with the idea of a soda report. “Here’s how many sodas you drank last week. Here’s how many sodas the average person in the community drinks. Here’s how many sodas a successful person drank that was your age and gender. If you reduce your soda consumption by one this week, you increase chances of success by X%.”

Mr. Tullman: Some of our UI designers thought the meter should show a smiley face or frowny face when people turn on the unit. One day is all it took. They are not with us anymore. [Laughter] People don’t want to turn on your app to know they suck.

Mr. Olivarez-Giles: Yeah, dashboards suck. I understand that there’s only so much time and space that you have to make something fun. But I don’t know. What do you guys try to model yourself after? How do you even think about that?

Mr. Bradford: We build off the hospitality industry. We try to think through every element of a patient’s thought process. In the normal healthcare day, there’s a financial burden but also a mental burden. There are many stresses for patients: Who will they go to? Will they be good? There are a lot of touch points. It makes it fun because the expectations are so low, but we can achieve so much more. Sometimes this means taking accountability and responsibility for things that aren’t in your control. What happens when you send patients to a pharmacy and the drug is not stocked? This is why we’ve invested in middlemen in our system and accepted that this is going to be a cost that will pay off down the road. This way, patients are actually satisfied with the services they ask for.

Mr. Tullman: I also think we’re going to see more intervention in apps and that it’s going to be human-directed. Let me give you one example: People with diabetes use a glucometer to test their blood sugar. We went to patients above the age of 50 and asked what they would change. They said that meters are getting smaller and strips are getting smaller, and they simply wanted us to make them bigger. They also wanted an arrow on the strip, because it can be hard for them to see how to insert it. I went to our manufacturing guys and said, “Can you do this?” They said: “This will be the ugliest thing on the market.” But the consumers love it. Healthcare is very much behavioral. We thought we’d totally be in the cloud, but you have to control the whole experience to make it better. We can’t control the technology alone anymore. It’s more than that.

Mr. Lee: Seeing a dashboard is sort of like seeing a balance in your checkbook – it’s knowing the number. What you really want is not knowing the number; you want a financial planner. How do we become a planner-coach and help you achieve what you want to achieve as opposed to just telling you where you are?

Q: How far are you? A lot of apps are using the word “coach.” I guess depending on your experience with organized sports that can be great or deathly. It sounds like a great idea, but nothing is really quite there.

Mr. Lee: When we say coach, we don’t mean somebody telling you to do another pushup. However, we have been testing human coaching through MyFitnessPal, and the results are incredible. We have tested cohorts where 100% of the people being coached by a person were retained though the training period. It’s remarkable how much more effective it is when a person is on the other end. Coaching is not a nagging phenomenon. It’s like the soda report – it’s not “Don’t drink soda,” it’s “Hey, did you know this?” These are suggestions or insights that make it easier to be healthy.

Q: Interaction with humans makes sense. It’s a comfortable thing to trust people vs. an anonymous startup app I’ve never heard of. But who you connect them with and how they interact is not something you can control. Most interactions with people in the healthcare system suck. Healthcare is pretty screwed up as a whole. How do you influence those decisions? We don’t want to wait in the doctor’s office. All of these things are never fun.

Mr. Bradford: Why I found myself getting into digital heath is that there are opportunities to look at problems in a fresh light and to think about what occurs. It’s not just about improving convenience, but about improving efficiency. One example is how rushed patients feel in the office. We’re trying to find solutions that give doctors more time to spend on patients. If you look at data today, 50% of patients say that they are doing what the doctor told them to do. But doctors say that only 2-5% of patients do what they are told. So there’s a lot of room for improvement.

Mr. Tullman: There are some really interesting studies that look at how long it takes for a doctor to interrupt a patient and say that they understand what the patient is saying. It’s like 20 seconds. So it’s not a good experience for patients. However, if you look at dermatology – anything that’s elective – you get very high ratings, because people are paying for it. As we move forward and people have alternatives for quality and cost, there are going to be major shifts. The other thing is that healthcare is behavioral. No matter what you put in front of patients, if they don’t want to change or do not understand why they should do it, the effort is going to be lost. The model used to be that doctors tell you what to do. That information balance is starting to change. I recently taught a course and posed an interesting question to my student. I showed them two slides and asked which person they would go ask to get information about 25 disease conditions: a 20-year-old kid at Stanford with a laptop or a doctor educated at Harvard holding a pencil and paper. 80% of the class selected the kid with the computer. What this reflects is that when you go in to the clinic, you may be better educated than the doctor. After all, they are focused on 50 issues, and you are only focused on yourself. We need to shift the model to say that healthcare is your responsibility as a patient and you are in charge of it. But we haven’t done that yet. When this gets better, the only time you’re going to go in and see somebody is when you really need to be seen.

Q: Obamacare is about providing options. What is this going to look like? Are you going to have $700 apps? What do you guys think the future of this will look like on the cost side for consumers?

Mr. Lee: I think that this is a rare opportunity to do something better and cheaper. Our app is free. We have the ability to deliver services to consumer at low cost and scale. This can only be helpful. I don’t think it needs to be a $700 app.

Mr. Bradford: We cost $40 per visit, and we take FSA and HSA. The median price for a dermatology visit in the US is $110. Often times people are paying out of pocket, and out of pocket continues to go up. There are more players in the ecosystem looking at the efficiency of telemedicine: “Can I shift encounters to telemedicine?” I heard a stat that one-third to one-half of ER visits are simply because people don’t know where to go. That’s lots of time. If they could turn to urgent care on the phone, and avoid going to the ER, there would be lots of visits done remotely. There are tremendous opportunities for cost savings and ways to make better use of providers’ time. In our business, time spent on admin paperwork by a dermatologist is around 9 hours a week. You just think about workflow and the costs of things.

Mr. Tullman: I agree 100%. A $700 app would save the country a tremendous amount of money if the right app did the right thing. Half the ER visits could be eliminated. Those are $3,000 or $5,000 visits. If only they could talk to someone before they went. There was a study that looked at how much time people that buy a car or sofa spend vs. people getting heart surgery. If you were going to have your chest cut open, you would do an enormous amount of research. But most people don’t even get a second opinion. One quarter of all hospitals that do heart surgery aren’t even certified by the AHA. You want to go somewhere that has done tons of heart surgery. This is not like going to a nice restaurant. We have to really start to think differently about healthcare. If we think like consumers, we’ll actually be okay. We’re moving to a catastrophic model of insurance. “You figure it out, but if it’s really bad, we’ll take care of it.” We know that because all of you are going to be smarter than a lot of big health plans. They are in a tough position.

Mr. Bradford: It speaks to the opportunity beyond the three of us. There is a financial burden that consumers face. And then there is navigating the healthcare system. What is the right place to go? There is no transparency in pricing, and no easy way to identify the difference in quality between providers. There are huge opportunities for innovation and new businesses.

Q: You said that trust is the biggest challenge for you guys. From a business perspective, how do you get people to want to use it and keep it cost effective?

Mr. Lee: The way we market is by focusing on building a great project. We think about how we can make our users succeed. That way, everything we do at the business works better. The great thing about MyFitnessPal is that your users are walking billboards. When people lose 30-40 pounds, people ask about it.

Q: Do you have plans to work with health and pharma brands as part of your monetization strategies?

Mr. Tullman: We are most directly into actual reimbursement and pharma. We’ve got to be very, very careful. We want patients to use the best solution. If we’re taking money from a company, we might direct them to use certain medications and insulins. So we keep an arm’s length distance.

Mr. Lee: Whenever we are approached by health and pharma, we ask, “Does this help users? Would they be comfortable with this? Does this help them succeed?” I think there are ways we can work with health and pharma, and there are ways we can’t.

Mr. Bradford: A similar answer. We provide coupons for medications to make more affordable. We’ve really got to stay Switzerland. People want the highest quality care, and they are trusting that doctors are making neutral decisions. We never want to do anything that compromises that.

Q: When we start looking at stakeholders – patients, providers, and payers – how do you see growth occurring because of these relationships rather than in spite of them?

Mr. Tullman: You have to deal with all three. As with any other industry, the consumer is going to drive this, but right now, there are large payers and health systems that you have to deal with depending on how close to medicine you get.

Mr. Bradford: We’ve gone direct-to-consumer, which has helped. We talked earlier about all the lack of money in the system. A CFO’s biggest concern is how much they are spending on healthcare. A lot of that money is spent on chronic conditions. So they are thinking about how to make better uses of services, and it’s important to get those stakeholders involved.

Q: When talking about chronic disease, you often think about the over 50 population, where smartphone penetration is lower. Do you have other tech more geared towards them?

Mr. Tullman: We took a standard glucometer and put a cellular chip and screen on it to make it more usable. We didn’t want to introduce a new device, but we also didn’t want to make people sync with their phones. So we took a new device, as opposed to being on the phone. The other thing is that phone companies hate blood. They don’t want to be regulated by the FDA. We did a separate device for now. It’s unique, but there will be other devices besides the phone for more and more sophisticated monitoring. We’re looking at any monitoring out there for other things. We’d love to connect with you all.

[Comment]: I’m currently working with HHS. My comment is that we’re very interested in data sharing. The Office of the National Coordinator is paying providers to use EMRs. Also, the NIH is planning to institute data sharing requirements for researchers getting public money. This is coming.

Q: My question comes from my perspective as a grandmother. Is there any interest in engaging children and putting in groundwork for the future? Are any of you working with Pixar and Disney, for example?

Mr. Lee: I have two kids, so this is something I think about all the time. However, our product wasn’t designed for kids. Those under 18 years old are not allowed on MyFitnessPal. We actually don’t want kids so focused on calories or weight. This is a really important problem. Other startups are trying to tackle it, and we’d love to address it too. We’d have to change our product to meet those needs, and we’d need to be thoughtful about that.

Mr. Tullman: I have one example of how we’re engaging with children with type 1 diabetes. Our product provides small insights on managing diabetes while patients are waiting for blood glucose readings. We have a group working on how to make those messages more appropriate for children. However, we’d gladly take a meeting with any of those companies. If any Disney characters want to deliver those messages, we would welcome that.

Q: What about the idea of the employer as stakeholder and partner?

Mr. Tullman: We sell to self-insured employers first, like Iron Mountain, Activision, Blizzard. Larger companies are embracing that. These are companies that care a lot and are also going to have employees for a period of time. Anyone self-insured owns the risk. But they are still just learning about it.

Mr. Bradford: With the employer opportunity, it’s about how we make employees healthy and happy. Something that is already getting uptake with the consumer directly is more likely to get employer traction.

Q: How critical is HIPAA compliance and EHR integration?

Mr. Tullman: HIPAA is incredibly important. Having just relocated to Mountain View, a lot of the great tech folks don’t understand healthcare and HIPAA. When you’re dealing with most patients, they waive those rights pretty readily. HIPAA seems like a big deal, and you have to understand it in terms of sharing information. But most patients will give you what you need to do to perform your business. You have to understand HIPAA and can’t ignore it.

Mr. Bradford: We run on Amazon web services. The infrastructure is a lot more mature now and it’s easier to put together HIPAA compliance in a robust way. We feel comfortable about it.

Fear, Shame, Empathy, & More Ways to Change Behavior

Dan Pink (Author, Drive)

New York Times bestselling author of Drive and To Sell is Human, Mr. Dan Pink, shared seven tactics for changing behavior. His entertaining, example-filled talk was grounded in both behavioral science studies and his experience filming the National Geographic Show, Crowd Control. The seven tips for changing behavior are outlined below in more detail, along with possible implications for diabetes/obesity: (i) use fear the right way (when narrowed focus is needed); (ii) Questions (often) beat answers for changing behavior (when the facts are on your side); (iii) use social proof – people look to other for cues about how to behave; (iv) use rhyming, alliteration, and repetition; (v) make it easy for people to act; (vi) put a human face on things; and (vii) experiment and try stuff. Though the large conference auditorium was only about half full, we thought this talk was a truly outstanding overview of tactical approaches to changing behavior – we hope to see many more of them piloted in diabetes/obesity!

  • Use fear the right way for changing behavior. Mr. Pink shared that fear can be a motivator and a mechanism for behavior change, but only in certain scenarios. Fear is a negative emotion, and consequently narrows humans’ scope and focus. This makes it particularly good when blocking out extraneous things is useful, such as in emergency settings (e.g., leaving a burning building). Fear works poorly as a mechanism for changing behavior when people need to have a wider focus (e.g., coming up with new ideas).
    • Example: Airline safety announcement. Mr. Pink and colleagues went to Jet Blue University’s flight simulator, a training location for flight attendants. The group tested three different ways of reading airline safety instructions, followed by a simulated plane crash. The team used: (i) a standard announcement; (ii) an announcement with a financial incentive - $250 for anyone answering all the safety questions correctly on a quiz; and (iii) revamped safety instructions and seat cards infused with fear tactics, pictures, and statistics (“opting to ignore the instructions increases your chance of dying by over 40%”). The third condition worked by far the best in ensuring passengers followed the directions and exited the plane quickly. 
    • Implications for diabetes/obesity: Fear is often used to motivate people with diabetes/obesity to change their lifestyle, and for the most part, it seems to be fairly ineffective. We wonder how narrowed focus may apply to these conditions: (i) managing diabetes/obesity require an expansive view (e.g., changing many elements of lifestyle); or (ii) people stop paying attention and zone out. In addition, we wonder how credible of a scare tactic long-term complications are – saying “you may go blind” or “you may have a heart attack” may not seem sufficiently scary or likely, since it occurs so far in the future and it’s hard to see the direct link between lifestyle actions and negative health outcomes.
  • Questions (often) beat answers for changing behavior. When questions are used instead of statements, people come up with their own reasons for doing something, which they believe more deeply and adhere to more strongly. Emphasized Mr. Pink, “Questions, by their very nature, elicit an active response – they are more engaging.” Statements, by contrast, may wash over the person receiving them.
    • The corollary to this point is also true – people can also come up with their own autonomous reasons for disagreeing with you. The implication is that when facts are clearly on your side, persuading with questions is enormously effective. But when the facts are not clearly on your side, questions are not an effective strategy.
    • As an example, motivational interviewing uses two “irrational” questions to encourage behavior change: (i) On a scale of 1 to 10, how ready are you to _____ (e.g., clean your room)? (ii) [Following answer] Why didn’t you pick a lower number? The second question encourages someone to come up with their own reasons for doing an unwanted behavior, which are believed more deeply and adhered to more strongly. If the person answers “1” (no willingness to change), the question is modified to, “What could we do to make you a 1?” This question typically surfaces some kind of barrier that is preventing someone from doing something. However, it is rarely needed, since most people do not rate their desire a “1” on a 1-10 scale.
    • Implications for diabetes/obesity: We wonder how questions might be better used in interactions between healthcare providers and patients, since the dialogue is often hierarchical and authoritative. Motivational interviewing is used throughout the field, though perhaps it should become standard practice for all interactions between providers and people with diabetes/obesity.
  • Use social proof – enlist the crowd and send message, “This is how we do things around here.” Given our tendency to look to other people for cues about how to behave, social proof can be very effective for changing behavior. Mr. Pink discussed a hotel towel experiment, where guests were encouraged to reuse their towels as part of an environmental program. The most persuasive tactic used social proof: “In a study conducted in fall 2003, 75% of the guests who stayed in this room participated in our new program.” A similar tactic worked in Washington DC after cellphone talking lanes were painted on the sidewalk – once actors started using their cellphone in the cellphone use lane, random pedestrians began following the rule too. What did not work was directing people to go to one lane or another; in fact, this made people angry. 
    • Implications for diabetes/obesity: We wonder about the potential of peer support and patient credentialing to encourage people with diabetes/obesity to better manage their diseases. A doctor saying, “Do this,” may not be very effective, but a fellow patient saying, “This is what I’m doing,” just might.
  • Make time to rhyme – “as cheesy as it sounds, don’t dismiss this technique.” Studies suggest that rhyming increases “processing fluency,” allowing for easier absorption. Repetition and alliteration also increase processing fluency. “Rhymes are a woefully underused technique,” said Mr. Pink.
    • Implications for diabetes/obesity: Education in both diseases tends to use a lot of jargon and unfriendly vocabulary. We wonder how more attention to processing fluency (rhyming, alliteration, and repetition) might help patients keep key more lifestyle principles in mind.
  • Make it easy for people to act. In one study, individuals were sent either “general” or “specific” letters about a food drive. Sending a very specific letter with a map, instructions, and details on what was needed dramatically increased participation in the campaign vs. the general letter. The study also tracked participation in two peer-evaluated groups – those deemed “likely” to contribute and those deemed “unlikely” to contribute. The group vs. group results revealed another important lesson: when we try to explain or predict behavior, we always overweight personality and underweight the importance of context. For boosting participation, the “specific” letter sent to the “unlikely” group was way more effective than the “general” letter sent to the “likely” group. As humans, our first instinct is to predict/evaluate behavior based on personality (“he is a jerk”, “she is a wimp”), but the context of an action is more important. As a result, behavior change should always ask, “Have I made it easy for the person to do what I want them to do?”
    • Implications for diabetes/obesity: Talk about a principle applicable to diabetes and obesity. Medical advice often asks patients to exercise more, eat healthier, or take medication, but pays little attention to context. Does someone live in an unsafe area or food desert? Can a person not afford a medication?
  • “Put a face on it” – empathy wins. As an experiment on his show, photos of local Austin residents in wheelchairs were put on handicap parking signs under the slogan, “Think of me keep it free.” The signs were intended to discourage able-bodied drivers from parking in disabled parking spaces – and they were incredible effective! After a month of surveillance, not a single able-bodied person parked in a handicapped spot. The key, said Mr. Pink, is to make things concrete; too often they are abstract and don’t motivate changes in behavior. Two dozen communities around America are now using these signs.
    • Implications for diabetes/obesity: We wonder how more empathy and patient faces could be brought into product development and manufacturing. In addition, could patient faces be added to data/lab reports to bring a more human side to analysis?
  • “Try stuff. Some stuff isn’t going to work, other stuff will.” Mr. Pink’s TV show sought to discourage bike theft in New Orleans, so the team placed intimidating, life-size cardboard cutouts of police officers throughout the city. Thieves were definitely distracted from stealing bikes (a success!); however, they stole the cutouts instead. Said Mr. Pink, “You can be very empirical about all of this. We’re in the era of A/B testing.”
    • Implications for diabetes/obesity: We thought this was a fitting note to end on, as we’ve been hoping for many more pilots in diabetes. We imagine there are so many experiments that could be conducted, whether on the clinical delivery side, motivation, reimbursement, software/apps, telemedicine, data management, etc.

Questions and Answers

Q (@diaTribenews via twitter): What are the most effective behavior change approaches for obesity and diabetes? The challenge is short-term sacrifice vs. long term gain?

A: There are lots of confounding variables. The main confounding variable is that these habits are ingrained in our bodies. We evolved in a landscape of scarce calories. We want calories. All of the sudden our Stone Age brains are in a world with abundant calories. It’s a harder problem to solve than a business pitch. First, go for small wins. Set it up so people can make progress. The other thing is changing the environment. So much of what we do comes from default behaviors – we do it because it’s what we always do. Vending machines have crappy food and we eat it.

Q: Have you ever felt your self-confidence was gone? How did you handle that and get it back?

A: All of us engage in self-talk, largely before something important. What should our self-talk sound like? It should be positive. “You can do this.” “You got this.” Here’s what research shows. Positive self-talk is much better than going in neutral, in terms of confidence. But it’s not the best thing. The best thing you can do is interrogative self-talk. Instead of going in and saying, “You can do this,” ask yourself, “Can I do this? And if so, how?” Self-help gurus that think every emotion should be positive hate this. But the evidence is very, very clear. Questions by their very nature elicit an active response. Pretend I’m pitching a new book to a publisher. I go in with positive,  affirmative self-talk, “You can do it.” I like hearing from myself that I’m awesome. I love telling myself I’m awesome. But it’s not the best thing. Instead, I should say, “Dan, can you do this? And if so, how?” I have to start answering the question. “I’ve pitched this before.” “I researched this.” I’m preparing, I’m rehearsing, and I’m practicing. When we think of something like self-confidence, it’s a very loud, muscular thing. Sometimes it is. But there is a quieter muscularity that is often more effective.

How Is Big Pharma Interacting With Tech Startups?

Panel Discussion

Jennifer Lannon (Life Science Program Manager, Springboard Enterprises), Phyllis Whitely (Venture Partner, Mohr Davidow Ventures), Megan Lopresto (Boehringer Ingelheim), Mike Marett (Cohero Health)

This session touched on the interaction between pharma and tech startups from several different angles – the startup perspective (Cohero Health), the pharma perspective (Boehringer Ingelheim), and the investor perspective (Mohr Davidow Ventures). Panelists emphasized the need for startups to solve a problem that pharma has. However, it also became clear how challenging it is for startups to even get pharma’s attention – even large companies have major bandwidth issues! But the ultimate punchline is that these two entities need to work together – startups need pharma’s scale/funding/validation, pharma needs startups’ innovation (particularly as drugs come off patents). Quotable quotes are enclosed below.

  • “I think in less than 20 years, pharma is not going to exist in the traditional sense any longer. With biosimilars and generics, it may not even be profitable to continue to market and sell the way that we do. There must be a value-add. That’s why [Boehringer] has a separate innovation group – how can we build our business to go beyond typical pharma?” – Megan Lopresto (Boehringer Ingelheim)
  •  “I probably receive somewhere around 150 cold emails a week. That’s a lot. That’s in addition to what I’m doing on a daily basis and the partners I already have.  Getting through that clutter proves incredibly challenging. For my group, we work with trusted agency partners – they tell us the things they’re seeing out in the landscape. We don’t have the bandwidth to evaluate every opportunity ourselves. BI is primarily in primary care products. Being in the PCP space is increasingly challenging. [As a potential startup partner], you must bring an understanding of the challenges and tell us exactly how what you have solves it. If we don’t have that solution piece, it’s too tough.” – Megan Lopresto (Boehringer Ingelheim)
  • “What is the big problem you are trying to solve, and how are you the solution? If pitching to pharma, you have to know what their big problem is. You might have a great idea that is really cool, but it may not be their problem. There are great examples out there. What Novartis did with Proteus. That’s one of the great challenges – knowing about compliance and getting feedback on drug levels. Or Genentech and PatientsLikeMe to work on better ways for patient recruitment.” – Phyllis Whitely (Venture Partner, Mohr Davidow Ventures)
  • “We need to appreciate that doctors are business people as well. We’ve got to make it easier, not harder.” – Mike Marett (Cohero Health)
  • “I want to provide a different vision of pharma. I’ve worked as a BD person in pharma, as well as somebody trying to sell into pharma. Imagine a maze. Think about all the different ends that go nowhere. And that one that gets you to where you want to get to. That’s pharma. But that maze gets different every few months. Oh and by the way – it’s different for every pharma. So who is your internal champion? Who believes in you and your product? Who are the decision makers in that organization? What is the process for decisions being made?” – Phyllis Whitely (Venture Partner, Mohr Davidow Ventures)
  • “One of the things that is really different is consumers taking control of their own healthcare. We’ve seen a real change – we’ve become digital. We’re driving payers and physicians to change into the more tech savvy digital area. It’s an incredible time to talk to the consumer as an entrepreneur.” Phyllis Whitely (Venture Partner, Mohr Davidow Ventures)
  • “Most entrepreneurs come out of academia or physicians. But they don’t have the business experience. Where I’ve seen success is when you approach it from a clinical perspective with outcome data to tie to the product. The best way to cut through is to distinguish yourself by creating and owning a category that you can define.” – Mike Marett (Cohero Health)
  • What pharma is trying to do is extend the lifecycle. If [your product] can suddenly change the marketing end or sales end of prolong IP, that’s a distinct possibility.” – Phyllis Whitely (Venture Partner, Mohr Davidow Ventures)
  • “We are seeing decreased influence of physicians in the treatment of patients because of exterior factors. Payers and managed care have greater influence. The empowered patient now comes into play. There is an opportunity to do more as pharma than just provide a pill. We need to show additional value. It’s not just about coming up with a me-too drug – another drug for type 2 diabetes. How do we take a pill-plus approach?” – Megan Lopresto (Boehringer Ingelheim)
  • “What has changed [from a drug discovery perspective] is next-gen sequencing. We can do a lot with genes. And Big Data – we can correlate a lot. We have started to see a real difference in how we approach research. There’s also 3-D bio-printing. We’re only on the cusp of understanding how drugs get into organs...and we can use social media to help recruit and attract patients.” – Phyllis Whitely (Venture Partner, Mohr Davidow Ventures)
  • “Naturally, pharma is risk averse, but they want to be innovative.” – Mike Marett (Cohero Health)

Future 15

Digital Health: Indispensable means Invisible

David O'Reilly (Chief Product Officer, Proteus Digital Health)

Proteus Digital Health’s Chief Product Officer boldly challenged the entire field of digital health with one bedrock design principle: make it invisible. He argued that most current digital health tools are status symbols, appeal to the worried well, and have a high drop-off rate. “How many things do you see on the wrist that are going to be transformative?... As an industry and an emerging category, we are just failing. We are not developing products that people are using. We can do better.” He very briefly discussed the work at Proteus, which operates under the core design principles of invisibility, seamlessness, and fitting digital health into people’s daily lives. The company’s poppy seed-sized ingestible sensor can be embedded into the pills patients take every day. Once swallowed, the sensor communicates with a disposable, discreet, body-worn patch, offering biometric data and medication compliance information on a smartphone app. “You cannot be shackled to technology,” he emphasized – these devices “should fade into the background” and be “private, intimate, and unobtrusive.” We love the lofty goal and continue to fervently hope that the company is working on a diabetes application. For more on Proteus, see our Closer Look from last August on the company’s $52 million in Series G funding.    

Resilience Through Tech: Designing to Help People

Fred Dillon (Director of Product Development, HopeLab, Redwood City, CA)

Fred Dillon, HopeLab’s director of product development, focused on technology’s ability to foster resilience. He broke this important quality down into three key components – purpose, connection, and control – sharing several app examples that boost feelings of each. HopeLab has built an impressive array of apps to improve health and well-being, including the widely publicized kids activity tracker/game, Zamzee – see a full list here.

  • Purpose: OptimizeMe (record, analyze, and optimize your life – uber quantified self tracking), SmartGoals (goal setting and tracking), Golden Scale (goal setting and tracking), On Purpose app (living life according to values you set).
  • Connection: Couple (for significant other to stay in touch) and Text to Connect (building empathy among college students and teens). Notably, Facebook is working with the Yale Center for Emotional Intelligence to optimize the language around disagreement messaging.
  • Control: Mood Meter (building emotional intelligence by selecting how you are feeling and why – beautiful design!). Mr. Dillon also mentioned Re-Mission 2, a collection of online games to help young people fight cancer.

Smart Yet Illiterate: Hacking Health Literacy

Christina Mullen (Director, Strategy & Innovation, Cadient A Cognizant Co)

Christina Mullen’s short talk on health literacy had two truly compelling slides. The first offered several statistics on health literacy from the US Department of Health and Human Services – we were quite shocked to see how prevalent health illiteracy actually is (see below). The second slide showed the seven styles of learning (verbal, aural, visual, physical, logical, social, solitary), noting that “Digital is the only medium to use all seven styles of learning.” That is compelling from a health education perspective. Ms. Mullen offered three tips for improving health literacy: (i) simplify (create an easy story and break down complex language); (ii) visualize (e.g., infographics); and (iii) play (enhance digital tools with interactivity that makes learning easy and fun).

  • Nearly 9 out of 10 adults have difficulty using everyday health information;
  • 99% of Americans can read; only 12% have ‘proficient’ health literacy;
  • 1 in 2 adults can’t read a drug label
  • 50% of patients leave the doctor not knowing what to do or what they were told;
  • 9 out of 10 adults may lack the skills needed to manage their health and prevent disease;
  • 50% of adults cannot understand a vaccine chart.

All Access: How Tech Is Redefining Health Care

Michael Smith (Global CIO & SVP Global Business Services, Mylan)

Mylan’s CIO Mr. Michael Smith discussed global healthcare access, boiling it down to three main obstacles: shortage of knowledge supply (e.g., nurses, doctors, other HCPs), physical/logistical constraints (e.g., rural population); and cultural constraints (e.g., stigma going to an AIDS clinic). He expressed optimism that technology could help address all three areas, but didn’t give any specific examples or share more about what Mylan was doing. In a teaser to end his presentation, he asked attendees to tweet @mylanCIO #ALL_ACCESS. His talk had a number of heartbreaking stats:

  • In the last 80 seconds, 35 children under age five will have died from diseases that were treatable or preventable.
  • 75 countries currently don’t have the bare minimum number of healthcare providers required to deliver the basic, most fundamental health services.
  • This decade, 125 million people will die as a direct shortage of the knowledge supply.
  • Over 50% of the world lives in rural areas, without the most basic infrastructure

Is Big Data The Next Wonder Drug?

Panel Discussion

Eric Topol, MD (Scripps Translational Science Institute, La Jolla, CA), Walter De Brouwer (CEO, Scanadu, Moffett Field, CA), John Nosta (Founder, NostaLab) 

This panel on the future of big data put the promise and challenges of the field into perspective. Panelists opened by emphasizing their high enthusiasm for the potential of big data, though made it clear that we are still in the first innings — “We’re probably at 0.1% of 100% of what we are going to know about our health.” This emphasis on the untapped possibilities in digital health was a theme throughout the session, which was headlined by both Dr. Eric Topol (Scripps Translational Science Institute, La Jolla, CA) and Mr. Walter De Brouwer (CEO, Scanadu, Moffett Field, CA). While such technologies will power the new age of big data, panelists were quick to point out that collecting data is not the issue – instead, the challenge is running analytics to generate actionable recommendations for patients and providers and, in the words of Dr. Topol, stealing the engineers capable of doing this from “Twitter and Pandora.” Ultimately, the panel reinforced how much generalized confidence there is in the future potential to improve health using data, but we felt the session was more notable for the questions it continued to raise about big data: (i) How do we actually use data to improve our future health? (ii) What happens when data is extended from the individual to the population level? (iii) Who owns the data and where should it be stored? and (iv) When is Big Data actually going to deliver on the promise, particularly in diabetes? It was terrific to hear from such notable thinkers, but we were somewhat deflated by the generalized discussion – there is nothing truly notable to report on yet. For the most quotable quotes from this discussion, please see the detail . For more from this discussion, see the most notable highlights below:

  • “Data is a wasteland. I don’t mean this as an adverse thing, but the analytics of this data is going to go exponential. The challenge is all the stored data. It doesn’t get analyzed. The bottleneck is data scientists. We need the distillate for the data to be useful.” – Dr. Topol
  • Dr. Topol, in response to whether we need “information-ologists” to help codify the data: Yes. That’s what holding us back. Most of these folks that have the knowledge to process this data are working at Facebook, Twitter, and Pandora. They don’t think healthcare is even important because it’s the last thing they think about in life. If only we could suck them into this world…”
  • We’re just learning what a healthy proteome is and what a healthy microbiome is. We had such a fixation on disease for so long that it’s taken this long to study health.” – Dr. Topol
  • “Almost every patient I see has blood pressure problems. But now, thanks to apps, they have blood pressure data on their phones, so they change their medication dosing based on that data. Because that is based on “real life.” And it’s only going to get better. Data is going to drive therapeutic change.” – Dr. Topol
  • Mr. De Brouwer, in response to whether data or devices are the “nut to crack”: It’s the signal and noise. Certainly, manufacturing something for hundreds of millions of people is difficult. But we know how to do it. However, the fact is that our data are all unique.
  •  “We saw it with Apple’s ResearchKit this week – having individuals who are willing to have data collected and used and shared. It’s unprecedented at that scale. It’s a global story – 24 hours after it was announced, it enrolled over 11,000 people in cardiovascular, over 6,000 in Parkinson’s, and over 3,500 in Asthma.” – Dr. Topol
  • “That’s why we have a problem on the medical side – the ability to let go and not have full control is very tricky. Forget about reimbursement and knowledge gaps.” – Dr. Topol
  • Today, medical data is owned by doctors and hospitals. That doesn’t make any sense. It’s that paternalism. The right framework is people should own their data. A logical place is one’s own personal cloud or locker. Today, there are no shortage of clouds, but they’re not set up for all that raw scan data. A perfect example: over 10% of medical scans are repeated unnecessarily in this country because the doctor cannot obtain the scan from the person who did it. That’s hundreds of millions of waste. If each person had their own data, that’s one immediate fix to the problem. So we need new laws. Patients should own the data.” – Dr. Topol
  • We don’t want any work involved. That was the failure of Google Health and Microsoft Health.” – Dr. Topol
  • Innovation to me is coming from unusual places. We’re not hearing from the Pfizers coming out with breakthrough innovations. It’s small tech companies, or its Apple and Google. That may be part of the reality.” – Mr. John Nosta
  • “We have learned to make detectors – it’s a yes or no thing. We’re very good at that. We need instruments that don’t know what they don’t know. So depression is probably reflected in your typing speed. Or weather, like in Shakespeare plays: it thunders when Othello is angry. You want to put all this big data in a mixer and have things fall out.” – Mr. De Brouwer
  • “Now that we can gather data, it offers a totally different look for medicine. But we haven’t begun the much larger scale effort. It’s all research, publications, and not practically useful for people.” – Dr. Topol
  • That’s the big, big challenge: behavioral change. People have never had data. It’s easy to see once people get it, it has an effect. But sustaining the effect is the hard part. We can do that through gamification, incentivization, competition, and co-opetition. So they stay with it. That’s the biggest challenge at the consumer, user level.” – Dr. Topol
  • “Behavior is sort of a conspiracy, Bin Laden peer pressure. If we start collecting our data, other people will do it.” – Mr. De Brouwer

Radical Healthcare: What Do Consumers Want?

Panel Discussion

Jason Oberfest (CEO, Mango Health); Alejandro Foung (CEO, Lantern); Jennifer Chung (CEO, Kinsights); Vinnie Ramesh (CTO, Wellframe)

This standing-room only session on what consumers want touched on all aspects of digital health: business models (leverage the existing system), the democratization of care (through the smartphone), barriers in healthcare (incentives came up a lot), engagement and user-centered design (very absent in healthcare, according to the panelists), the need for more engineering/programming talent (everyone is going to Silicon Valley tech companies), the balance between technology/automation and humans (hopefully synergistic!), scaling digital health companies (only a few ways to do it), and the provider challenges of dealing with all the data. This was a session full of gems and we’ve enclosed the most quotable quotes below from Mango Health CEO Jason Oberfest (medication reminder app, via gamification), Lantern CEO Alejandro Foung (emotional well-being assessment), Kinsights CEO Jennifer Chung (advice sharing network for parents), and Wellframe CTO Vinnie Ramesh (risk management).

  • We actually believe that mobile is the way to democratize care. If you saw the Economist last week, 80% of all adults in the world will have smartphones by 2020.” – Jason Oberfest (CEO, Mango Health)
  • As we move to accountable care, digital health is more viable. Providers and hospitals become focused on quality of care. It becomes easier for digital health companies to survive. Then incentives are aligned. The harder problem is how do you survive in the mean time?” – Vinnie Ramesh (CTO, Wellframe)
  • “Opening a box of clothes is fun. When you are sick, it is not fun. We need healthcare solutions that are just as compelling in design by putting ourselves in the mindset of consumer.” – Alejandro Foung (CEO, Lantern)
  • Two things are missing in healthcare. Lack of user-centered design and lack of engineering leadership. We have really smart people working in aerospace, finance, defense – you don’t have the smartest engineers going into healthcare. That combination, in addition to aligning incentives, is really missing from healthcare.” – Vinnie Ramesh (CTO, Wellframe)
  • “Newer brands [in healthcare] are focused on parts of the relationship with patients that are sporadic and transactional. We need real innovation... to make someone’s life better every day. This is especially true for chronic conditions, which are 75% of US healthcare spend.” – Jason Oberfest (CEO, Mango Health)
  • “You need the best engineers working in healthcare. Get them out of Facebook and Apple and Google.” – Vinnie Ramesh (CTO, Wellframe) [Editor’s Note: They may very well be working on healthcare in these companies! But this was a common theme throughout many panel discussions at this conference – the best engineers are not drawn to healthcare, since it is less sexy and slowed by regulation.]
  • “Who pays? Consumers are not used to paying for healthcare. The other challenge is behavior change. Drastic behavior change is really, really challenging.” – Jennifer Chung (CEO, Kinsights)
  • One big challenge – what is technology good at and what are humans good at? Healthcare is very service oriented; the tech people want to replace doctors with AI. But how do you keep that human touch in healthcare, or use tech to amplify it?... Humans are good at communication. Let’s amplify their ability to do so. Computers are great at taking a lot of data and turning it into something a nurse can digest. But you still need the nurse completing that feedback loop.” – Vinnie Ramesh (CTO, Wellframe)
  • “A lot could be automated, but the patient may not want that. I just think patients aren’t quite ready for a robot to do surgery on them. I’m certain we’ll get there at some point.” – Jennifer Chung (CEO, Kinsights)
  • “If you believe that the future of healthcare is largely about building amazing experiences for patients – and we certainly subscribe to that – then you’re the kind of company that obsesses over user experience. I have a list of 800 things we want to do in our app. However, you have to actually focus on other areas too. It’s a hard balance to get right.”  – Jason Oberfest (CEO, Mango Health)
  • Once you do align incentives, everyone is super motivated to do something. You can make some really meaningful interactions happen.” – Vinnie Ramesh (CTO, Wellframe)
  • That’s how medicine has started. Let’s measure outcomes and grow it. What if we started the other way – let’s get tons of people using it, and see what good we can do with it? One company that’s an example is Fitbit. I wouldn’t say that it’s really changing people’s lives. Counting steps – good job. On the other hand, they own 40% of the wearables market. It changes small decisions. Think about what they can do now once they’ve reached the size they have.” – Alejandro Foung (CEO, Lantern)
  • “It’s a logic tree. Are you building something that you think consumers will want to pay for? For Fitbit, it’s a yes. In some parts of healthcare, it’s yes, in other parts it’s no. Then you need to focus on one of two things. Either generating evidence that your stuff works, which is the primary upstream metric. Or you build a business around billable codes. You can get commercial scale going if you can productize against billable codes. Companies like Omada have done a great job of that.” – Jason Oberfest (CEO, Mango Health)
  • “If you’re building something for people with cancer, the person building may not have had cancer. You still have to develop that empathy for the user. In e-commerce, you buy stuff online as the developer. So the friction is a lot lower. It’s really, really hard to initially relate.” – Vinnie Ramesh (CTO, Wellframe)
  • “I’m a physician that does retinal surgery. In talking about healthcare apps that patients are using – patients will pay $4 a month for one app that measures a change in vision. An email gets sent to the doctor. From a provider perspective, lot of doctors don’t like that. The doctor doesn’t get reimbursed for the time. You have to deal with this extra data. But since the doctor is not getting reimbursed, it’s less likely to be adopted. This is real world stuff. If it is a billable CPT code that already exists, the doctor is more likely to use it. For telemedicine, they are developing CPT codes, but they’re slow to develop, and doctors are saying, “How do I bill for that?” To go for the thing that’s already in place is something that is going to make a lot more money.” – Audience Comment

Your Health Startup Is Cute: Gotta Play at Scale

Panel Discussion

Andrew Rosenthal (Group Manager, Jawbone), Corbin Petro (President, ElevateHealth), Elliot Cohen (Co-Founder, PillPack), Ryan Panchadsaram (US Deputy Chief Technology Officer, The White House)

This diverse panel included touched on a wide array of issues in digital health, including achieving scale, working in government, “pilots” vs. “research,” using data to actually improve health and drive behavior change, developing a business model, payer challenges, and more. Quotable quotes are below.

  • Be clear about what you’re doing. Most pilots are research projects. And they don’t think that’s what they’re doing. When you have a pilot, you are solving a business problem for the person piloting, and they’re going to be a customer. Research is about improving outcomes. Research takes a long time – it takes 20 years to pay off. That’s why its funded in government.” – Elliot Cohen (Co-Founder, PillPack)
  • “You start with user needs. The first time touched a user was on October 1 [when it launched]. In this room, you test with users throughout the development process, and you roll out iteratively.” – Ryan Panchadsaram (US Deputy Chief Technology Officer, The White House)
  • “A really great idea may not be a really great business, particularly in healthcare.” – Corbin Petro (President, ElevateHealth)
  • Behavior change is hard. Good design makes it better. Good design shows up in the consumer market. In environments that are heavily regulated, sometimes design is never a part of the conversation.” – Andrew Rosenthal (Group Manager, Jawbone)
  • In my experience, consumers have limited willingness to pay for healthcare. The one exception might be for weight loss – people are pretty willing to pay for weight loss or fitness. Generally, the expectation is that someone else is paying – government, insurance, or an employer.” – Corbin Petro (President, ElevateHealth)
  • Businesses in healthcare are always more scalable if they work within the business context that already exists. Healthcare understands how to pay for services and how to pay for providers.” – Elliot Cohen (Co-Founder, PillPack)
  • “At Jawbone, data belongs to users. If you add data into the system, that’s your data. But we make great use of it. If you haven’t been sleeping well, we might notice that and try a behavior change intervention, “Today I will...” – it’s an approach to get you to bed a little earlier. About 23 minutes earlier, we find, on average. We use data in interesting ways to improve health. But you are welcome to delete your data at any time from your system. – Andrew Rosenthal (Group Manager, Jawbone)
  • “Government will only be as strong as the people that participate in it. If you want to play at scale, come join us.” – Ryan Panchadsaram (US Deputy Chief Technology Officer, The White House) on the need for more technologists and engineers from Silicon Valley to join the government.
  • “The world’s best sleep labs at Harvard Stanford have beds in the room and a one-way mirror. You are hooked up to 36 different electrodes. That’s best in class sleep research. So if you want to understand why you are having trouble sleeping, the best way to understand that is to bring people into a foreign environment and wire them up to 36 electrodes. [Laughter] People track sleep with Jawbone. When we try interventions around sleep, we have 40,000 people. We ask, “Do you want to try this?” And we look at the outcomes. We don’t talk about pilots. Traditional people in healthcare haven’t seen that scale.” – Andrew Rosenthal (Group Manager, Jawbone)
  • “Everything we are trying to do is incremental. We need to change incentives and align them to reduce costs. We’ve found challenges within that. Health insurers benefit from low utilization and low costs. HCPs benefit from volume of services.” – Corbin Petro (President, ElevateHealth)
  • Project Argonaut is a new project to standardize APIs for all EHRs.” – Ryan Panchadsaram (US Deputy Chief Technology Officer, The White House)
  • “In the consumer space, we have a saying: health is what happens in between doctor’s visits.” – Andrew Rosenthal (Group Manager, Jawbone)
  • “The federal government listens and listens very closely to folks in public. Policies go out and there is always an opportunity for comment. We hear a lot about lobbyists, but you in this room have an equal chance.” – Ryan Panchadsaram (US Deputy Chief Technology Officer, The White House)
  • A lot of policies from Washington DC, CMS, and the state level are incredibly challenging to implement in the timeframes and budgets. For instance, the High Tech Act – the amount of money to implement EHRs is completely not able to purchase anything worthy of purchasing. It’s a challenge for government to impose things – a lot of the people writing legislation haven’t been there implementing and operating things.” – Corbin Petro (President, ElevateHealth)

Promise and Pitfalls of Too Much Health Data (Presented by WebMD)

Panel Discussion

Robert Glatter, MD (Lenox Hill Hospital, New York, NY), Michael Smith, MD (Chief Medical Editor, WebMD, New York, NY), Shannon Dickson (Healthcare Consumer & Editor, WebMD, New York, NY)

This panel on the pitfalls of too much health data began with a video on wireless medicine, one of a five-part WebMD series on the future of health (the others are infertility breakthroughs, 3-D printing, obesity treatments, and advances in vision). Said Dr. Eric Topol on the threat this poses to the hierarchical medical establishment, “We knew medicine was getting digital. But we didn’t know if was getting democratized.” This theme continued in the panel discussion, which was headlined by emergency room physician Dr. Robert Glatter. He emphasized that doctors are not equipped to deal with all this data – some actively resent it – and healthcare should be thinking about data management positions to handle the deluge. In addition, there are potential liability concerns if providers miss data that is concerning, though worried patients compound the program by sending unnecessary data in. Manufacturers can help on this front, a point emphasized by Afib advocate Mr. Shannon Dickson – he highlighted digital health poster child AliveCor’s handheld EKG device/app, which now includes algorithms that automatically read the report. Notably, Dr. Robert Glatter highlighted WellDoc’s BlueStar as one example of a valuable, FDA approved digital health solution that has helped his patients with type 2 diabetes. The panel discussion is enclosed below.

Dr. Michael Smith (Chief Medical Editor, WebMD, New York, NY): Dr. Glatter, tell us a little bit about your journey into digital health?

Dr. Robert Glatter (Lenox Hill Hospital, New York, NY): I’m an emergency room physician, and over the years, I’ve seen a lot of inefficiencies. I’ve seen people who just haven’t been able to get to the ER when they have needed to. The question I asked myself is: How do you get to these people? So, I started doing house calls but still felt that there was a need to get to people and see them more quickly. I figured that there had to be ways to use digital technology to help people with conditions see doctors more quickly. There are so many people that need access. Therefore, in some sense, I think that the rebirth of house calls has been an impetus for the digital healthcare revolution. I was then approached by Pager, a company with a telemedicine platform that connects patients with doctors. We’ve had amazing success bridging that gap. It has come to Austin and many other places. I’ve also seen the utility in other apps, such as those that monitor atrial fibrillation (Afib). I’ve had patients come into the ER who I don’t think have had an atrial fibrillation event, but they have proof of it. I also use [Welldoc’s] BlueStar – the first FDA approved mobile prescription therapy for diabetes. My patients with diabetes are way more well-controlled. Their A1cs are way lower. Of course, there are downsides to apps, but the bottom line is that engaging patients is something we all need to do. Having an intermediary is important, perhaps, and dealing with the deluge of data.

Dr. Smith: tell us about your experience, how digital medicine has helped you?

Mr. Shannon Dickson (Healthcare Consumer & Editor, WebMD, New York, NY): I’m a 63 year old, former owner of a mechanical engineering firm in Hawaii. I’ve owned and edited the oldest patient forum and advocacy website dealing with atrial Fibrillation (Afib). My path to Afib started 24 years ago – I was formerly heart healthy and felt like a salmon was flopping around inside my chest. It felt like a strap around head. Your first initial impression is that these are your last moments on earth. There have been huge advances in modern medicine to treat the condition. Afib is rarely fatal directly, but it can lead to direct fatal consequences, like stroke. With a simple device like this [pulls out AliveCor] and an iPhone app, I can hit record and get an FDA approved single lead EKG reading. That helps determine if I’m in Afib or not. The sense of the unknown is one of hardest parts of Afib.

Dr. Smith: Patients can send EKG strips to their doctors throughout the day. What is a doctor supposed to do? It’s not only overload but it’s also a liability issue – there’s no establishment around them to help them deal with that.

Dr. Glatter: I don’t think at this point that we are ready to handle the data. It begs the need for a new breed for data management specialists. Large hospitals are seriously looking into this, not just from a legal standpoint, but in terms of taking good care of patients. We’ve got what we need on the front end, but on the back end, it’s very lacking.

Smith: That’s true. Many physicians are struggling with electronic medical records. This data somehow has to get into EMRs and we have to give doctors the capacity to work with those systems. However, are doctors willing to change?

Dr. Smith: Are most of your colleagues open to data?

Dr. Glatter: Depends on the doctor. Internists and cardiologists with large practices have people to manage it. It is an evolving problem. If we don’t address it upfront and soon, data is going to be exploding and lost and not making it into the hands of people making key decisions. We need to empower an additional person besides physicians. What training or credentials do they need to have. Is this a role for the physician extender, like an NP or PA?

Dr. Smith: From a patient perspective, what have you seen people do?

Mr. Dickson: The type of patient that makes it to our forum is somewhat proactive. They take the bull by the horns. That predisposes most people we deal with to be open to and capable of dealing with new technologies in a better way. The big challenge is in making an interface between the patient and doctor. Both are working in concert together. That is one thing I consistently hear. Cardiologists specialize in electrophysiology – they were all gung-ho about the release of AliveCor. And they still are. But it’s tempered a little bit by the overwhelming amount of data. Sometimes people send 10 EKG strips in a day. After a while, most of that is a bit overkill. There are legal aspects too – what if you send a whole stack of EKG strips, and only two of them have five or ten seconds of Afib? Afib is not diagnosed until there are 30 seconds of continuous Afib. But if Afib develops and the doctor doesn’t treat, there could be a liability issue. But you cannot diagnose with just a little blip of Afib. On the other hand, there is a tremendous amount of support and piece of mind this brings. You’re taking out the unknown.

Dr. Smith: That’s true. As we saw in the video, to be able to so readily diagnose an ear infection is really empowering. Have you seen any downside from the patient perspective?

Dr. Glatter: Yes, it can be overwhelming. If people look at reports that have incidental findings, it can lead to panic and concerns. I have had experiences where people see little things in reports, such as a kidney cyst, and it provokes fear. You can’t blindly give information to patients unless they are very educated.

Dr. Smith: I would say that doctors support the use of these devices as a diagnosis tool. However, very few support giving this information to patients. They think it might challenge their medical authority. Have you guys seen any evidence of that in your practices?

Dr. Glatter: I do. The older generation of doctors, 20-30 years in practice, are not comfortable with these devices. The younger generation, say 10 years in practice, is pretty comfortable. As time goes forward, we’ll see a little bit of a change.

Dr. Smith: Do most of the doctors you know recommend AliveCor? Are they really gung ho about it?

Mr. Dickson: Most of them do. They are happy to see when patients use it. They just want to make sure people are using it correctly. It you squeeze it hard, you get more noise. In the past year, AliveCor developed four separate algorithms that go along with the device. It takes away the guesswork for the doctor and patient. If the signal is Afib, it says “Afib.” Same with “normal.” It’s a real life-saver for doctors and patients. Sometimes if there is a little bit of noise, it will say, “Interference.” And it will say “undetermined” if there is a signal for a different type of arrhythmia. Those helpers from the device manufacturer really help. But they won’t alleviate he need for a data collector.

Dr. Smith: Certainly, this opens the door more for self-diagnosis and management. What Shannon was saying is that not only does the device detect and monitor heart rhythms, but it tells the patient whether the heartbeat is abnormal. It advises patients and helps them work with their doctors. Dr. Glatter, let’s think pie in the sky for a second. Where are we going to be five years down the road.

Dr. Glatter: I think where we are moving is toward app refinement and communication with EHRs. There’s been a roadblock with EHRs not wanting to share code with developers. It’s a roadblock because there is no communication between these systems. Companies have not been wiling to share this information and create open APIs. If you could have this information, it would be hugely helpful. After all, really taking care of patients requires real-time information and if we could cooperate, we could take care of it.

Dr. Smith: There are FDA approved devices like AliveCor. But most devices and apps are not FDA approved. How do we think about that?

Dr. Glatter: There was recently an app looking at blood pressure. It was inaccurate in terms of monitoring, and some patients were getting super high levels. We need validity, accuracy, and reliability – garbage in, garbage out. As doctors, we have to know that apps are validated and have been studied. Organizations and medical societies need to do evidence-based validation. That is why there is resistance in medical community to adopt these apps.

Dr. Smith: This seems to have the potential to transform how hospitals are run. Is that fair to say?

Dr. Glatter: Yes, I think one of the major hospital areas this can affect is congestive heart failure. That’s a big group of patients that keeps coming in and getting readmitted. It is a big cost and Medicare is not going to pick up the cost. However, if we use technology, we can monitor these patients though at home. Another other area that can be transformed is glucose monitoring. Glooko has been very effective. So has [WellDoc’s] BlueStar. They have received very good reviews. Glucose management is a big issue in the hospital, and these platforms make patients feel empowered.

Dr. Smith: All this data is coming into the hospital. Doctors have access to it. Is it truly improving patient care? Or are we trying to work our way towards it?

Dr. Glatter: We haven’t seen it yet – it’s a work in progress. We have a national initiative to have an EHR. There has to be a national initiative to manage chronic disease using medical apps, and those need to be integrated with EHRs. They need to be integrated with remote monitoring centers. The apps out there don’t address chronic disease management.

Dr. Smith: I would say another goal of technology is decreasing costs while improving care. We’re striving for both.

Mr. Dickson: What we see in people who use AliveCor is that it speeds up the recognition process and speeds up the process of taking action. One fellow in New York had diagnosed atrial fibrillation for some time and it was weighing on his mind. His initial response to it was to brush it under the rug and one day he had a stroke at his office. It did wake him up and he got an ablation. He bought into AliveCor after that. Over the last two years, he’s had 17,000 EKGs and they have confirmed what he has felt, but it has given him additional reassurance.

Dr. Smith: I want to mention payment. Is it a sustainable thing for doctors – we are asking more of them, since it’s more data. How do we incorporate it into practice?

Dr. Glatter: Overall, there is not a high penetration rate. BlueStar is FDA approved, and CMS and private party insurance will reimburse it. Apps have to be covered by major insurers for the movement to propel forward.

Dr. Smith: Insurers are covering the cost of the device, but there is not necessarily payment for monitoring and review of the data.

Dr. Glatter: It’s an evolution. CMS has agreed to cover a good part of the charges. Third party payers are coming on.

Q: So many people these days have fitness trackers and more available wellness data. What’s the state of using that copious amount of data when you see your doctor?

Dr. Smith: That is a ton of data. One thing we talk about is whether we could use it in a clinical trial. There is a lot of conversation about this data. However, what to do with wellness data is still unclear.

Dr. Glatter: Apple’s HealthKit is a step toward productive data utilization. Major academic centers are looking at chronic conditions and thinking about ways to enter data in a way that can be tracked. The future is that this will be studied in a bigger way.

Q: What is the role of the community pharmacy in screening and the use of monitoring and apps? AliveCor has been used in Australia to screen patients in the community pharmacy setting.

Dr. Smith: There are big uses here for medication compliance –that’s key for pharmacies.

Dr. Glatter: The important part is triggering some kind of alert to the pharmacist. There is a new app on the market, Proteus’ Helius. It’s an ingestible sensor and a patch you wear that connects to the cloud. You know if the patient is taking the medicine. Then, you can confront the patient, “You have not been taking your medicine, your blood pressure is not getting better.” Proteus Digital Health’s Helius. It’s going to be big.

Mr. Dickson: The pharmacy is a really good opportunity to catch people who have no idea they are in Afib. About 30% of the population is asymptomatic. That’s a very big risk of having a stroke. Afib strokes tend to be very serious – about 70% are fatal. Even coming into to get a flu shot, you could have an AliveCor check. I’ve caught five people myself who had no idea they were in Afib - they certainly were. Democratization of the EKG has its benefits.

Q: Can you talk about the certification that a device goes through? Can we trust the data that comes from patients at home? Or is this a second class of data that will have to be confirmed be measurements in hospitals?

Dr. Smith: This really speaks to the processes devices go through to get approved.

Dr. Glatter: We have to be able to trust data from apps, and I would note that the NHS has a very high bar by which to certify an app. So, in England, there is a high threshold. However, here in the US, there are 25,000 apps and fewer than 50 have really been validated or certified. So that speaks to the problem.

Mr. Dickson: I think we have to look at these at-home devices as a good front-line indictor, but one that to be confirmed by follow up. At-home data can give you a red flag, but I wouldn’t use it as standalone data and forget the doctor.

Q: Can you speak about medical schools training medical students on data, and preparing for more technology enabled care delivery? And Dr. Glatter, what about working with patients who want data?

Dr. Glatter: Medical schools are making a foray into technology and medicine. Quinnipiac  University is really forward thinking on this. They are looking at apps and teaching students how to evaluate data from apps. It’s very cutting edge. Other schools are following suit, like Harvard and Stanford. So we’re seeing more education related to digital healthcare. Secondarily, in terms of sharing data with patients, I like to tell them what’s going on. That is tempered with fear. We let someone go from the hospital after an ER visit for chest pain, and their mind is going to start wandering. The problem we’re facing is creating some fear. We have to try to temper that. Follow up is so critical after an ER visit – having physicians discuss the findings. People go online and play Dr. Google and become terrified.

Dr. Smith: A lot of you have access to lab reports in doctor’s office. You might have freaked out when you saw something had an H, for “high.” It might not mean anything.

Q: I run a digital health startup focused on reproductive health. What we found is that these consumers are driving digital health for their families. They are buying digital health technologies and are going into offices and asking doctors about the data they are seeing. Doctors are confused because they don’t know how to deal with these products and the data being generated because they haven’t seen it. So my question is whether there is a way to standardize data formats?

Dr. Glatter: I think there is a real need for a national consensus on this front. We need standards for communication. I think this is a work in progress. It’s not there yet, but this has to be a conversation.

Mr. Dickson: I agree. This has to happen on a formal basis. However, until it gets formalized, there are a few sites with quite a bit of knowledge that can help patients get to this information. That can be an interim step for patients. When this happens, patients will come into the clinic with more information and can create a more collaborative relationship, too. Most fields have at least one if not two or three really outstanding support websites. You have to be careful though. You need to find one that’s evidence-based.

Q: Is management of health data vs. management of human beings something physicians want to do? Are they excited about that? Will there be a shift in the way malpractice is set up, to fend off irregularities and inconsistences in data.

Dr. Glatter: Physicians are challenged by this – health data on the one hand and taking care of patients on the other. At this point, we’re just not there yet. The deluge of data is just overwhelming for many doctors; they don’t know how to deal with it. Having a person to manage it is needed in many cases.

Dr. Smith: What about malpractice?

Dr. Glatter: A lot of docs are concerned about the risks of data coming in. When we train, we have a morning report. It’s about having that data available, but presented in formal fashion.

Dr. Smith: Docs love data. But we’re dealing with too much of a good thing right now.

Q: What do you think of the role of the ACA in driving some of these changes and importance on evidence-based care?

Smith: Conceptually what these apps are trying to deal is entirely in line with the ACA: reduced costs and improved outcomes.

Dr. Glatter: Yes, I agree. Apps are all about bundled care. It’s all about less fee-for-service. I think that’s what is driving this. It’s about more streamlined care and delivering care at a fraction of the cost.

Q: I have a medical device company and a consumer device company. SXSW is exciting, but it’s hard finding software geeks with the mentality and patience to go through the regulatory process and the ISO process. It’s so much slower. The whole mentality is very different. Having you talk here is a very progressive, useful step. We need to be around people who are going to be making the next slew of software, but with some warnings that it’s not quite as fast paced or as exciting. Thank you so much. I have a  lot of sympathy and respect for AliveCor. It’s so important to get devices that we all stand behind. I see so many products that use fluffy words: “It’s quite accurate sometimes if you wear it right.”

Dr. Smith: We certainly need regulation here.

Mr. Dickson: We need creative designers who have the long view.

Q: I understand the topic of the presentation, but I almost feel like we’ve lost perspective of where the data should be focused. It should be focused on preventative healthcare. Are we losing sight of what’s really important?

Dr. Glatter: Disease management is where we should focus. Education is key here.

Patient Power: A Data and Mobile Health Revolution

Digital Power -> Patient Power

Jay Boisseau, PhD (CEO, Vizias, Austin, TX)

Dr. Jay Boisseau gave an optimistic talk on digital health, noting that technology is not a constraint, but is making it possible to solve important healthcare problems. The real challenge, he said, is the healthcare system (e.g., policies, incentives, etc.) – can consumers, entrepreneurs, and researchers outflank the current structure? The title of his concluding slide summed up his view of where things need to go: “Computers + People = Insights.” An illuminating quote (misattributed to Albert Einstein) made the synergistic potential wonderfully clear: “Computers are incredibly fast, accurate, and stupid. Humans are incredibly slow, inaccurate, and brilliant. Together they are powerful beyond imagination.” Dr. Boisseau envisioned a “Super WebMD” in the future consisting of the iPhone + wearable devices + Apple’s HealthKit + IBM’s Watson supercomputer. The latter can especially help with “the doctor’s data dilemma,” where providers don’t have enough time to analyze all the data generated.

  • Context #1: Biological numbers are big – but FINITE. For example, the human genome has three billion base pairs = 1.5 GB logical data (less than a free conference USB flash drive), or 200 GB of raw data from a sequencer. For seven billion people on the planet, that translates to 1.4 zettabytes (10^21) of raw data. Human genomes vary by <0.1%, so storage would be one exabyte [admittedly, we have not heard of these terms!]. But total information per person is ~100 trillion cells, or the equivalent of ~150 zettabytes. So human biological data is truly BIG data, but “it’s finite, which gives us hope.”
  • Context #2: Digital technology and the power of exponentials. Dr. Boisseau mentioned Moore’s law, where processing speed doubles every 18 months. Other digital technologies also have exponential growth. For instance, human genome sequencing first cost billions of dollars. Now, it’s thousands of dollars. “Biological problems are huge,” he said, “but they are finite, and digital capabilities increase exponentially.” This means the world will be very different in a few years.
  • Smartphones are capable, ubiquitous, and connected to vast computing and storage. Dr. Boisseau mentioned Apple’s HealthKit and ResearchKit as two promising examples of connecting devices to cloud software and storage. He also mentioned 23andMe, who has partnered with Genentech to share data for medical research.

Data Quality, Privacy, and Transport

Matthew Cowperthwaite (Director of Research, NeuroTexas Institute Research Foundation)

Matthew Cowperthwaite made three points in his short remarks on healthcare data: there are “enormous data quality problems”; a shift in the notion of privacy is needed (HIPAA “has to go”); and patients should have access to their own data and should be able to move it easily.

  • There are “enormous data quality problems.” For example, the simple Yes/No question, “Do you have diabetes?”, might see 38 different data entries in an EHR (e.g., “Getting better”).
  • A shift in the notion of privacy is needed – HIPAA was signed into law by President Clinton roughly 20 years ago! Mr. Cowperthwaite believes that the law “has to go,” as the current system puts too much risk on the IT department and leaves no room to experiment. In his view, we need incredibly strict fines/penalties for violators of data privacy.
  • Patients should have access to their own data and should be able to move it easily. He envisioned an Apple-Pay-like kiosk allowing patients to download their medical record data to their phone. The challenge is technology legacy barriers and perceptions that need to be broken down.

Healthcare and Big Data

Edmund Jackson, PhD (Hospital Corporation of America, Nashville, TN)

Speaking from his experience with the Hospital Corporation of America (the “largest healthcare provider in the US), Dr. Edmund Jackson put the highly constrained healthcare environment in stark perspective. He acknowledged that the goals of big data movement are laudable (namely, to improve clinical and financial outcomes) but suggested that the magnitude of hospitals’ responsibilities (e.g., saving lives) makes forward progress difficult. Indeed, he suggested that deviating from a system that is familiar entails an underappreciated psychological barrier for both providers and the broader healthcare system. We were intrigued to hear his confidence that the system is trending in the right direction, though he stressed that the adoption of data is still moving in baby steps. Why? Because of one underappreciated fact in his view: Data has to be paid for.

  • Dr. Jackson called the introduction of EMRs into hospitals a “healthcare revolution” in and of itself. Though he received some pushback on the idea – other panelists wondered whether a “revolution” has actually happened yet – he reminded attendees that physicians are being asked to move away from paper systems that have “worked” to online systems that are not provider-friendly. He suggested that the biggest hurdle to EMR adoption continues to be at the front end: “We need to find a way to get the data into the system in a way that doesn’t break the system.”
    • While the front end of EMRs is lacking, Dr. Jackson acknowledged that healthcare systems already have huge databases of information to draw from that provide an accurate representation of the inpatient experience. The challenge moving forward is embracing this data and capitalizing on its predictive potential to improve outcomes and reducing costs.
  • Dr. Jackson concluded his talk by presenting the elephant in the room when it comes to data: “Data wants to be free. But data also wants to be paid for. That’s the message that isn’t talked about.” We would note that significant value has come from liberating data and allowing patients to solve real-world problems as seen in the success of NightScout. Horizontal systems allow for clinically relevant innovation and scaling that is hard to achieve in closed proprietary systems. And once the data is set free, it can benefit industry through an ecosystem of apps, as we heard from Dr. Joseph Cafazzo and Howard Look at the November 13 FDA workshop on interoperability.

Patient Power?

Mini Kahlon, PhD (Dell Medical School, Austin, TX)

Dr. Mini Kahlon shared her view that mobile devices and big data are our keys to “revolutionizing” the healthcare system. She believes that these shifts in health data ownership – from providers to individuals – can empower patients to understand their personal health and engage in self-management. The obstacle, as she sees it, is the complexity of our current system in which industry cannot adopt technology without creating conflicts of interest with business models. For example, she raised the irony that many of those who call for expenditure reduction (e.g., academic medical centers) are those who need the revenue to exist – in this light, she stressed it is no surprise that digital technology has experienced such slow uptake. The key question, in her view, is whether industry can reconstruct the economic incentives to motivate such change. Unfortunately, she did not provide specific advice on this front, though the underlying message was quite clear – “We can’t just enable technology in a corner. We’ve got to change the business model.”

  • Dr. Kahlon opened her talk with a simple but telling statement – “Healthcare is a domain where I often feel disempowered.” She stressed that the field has not yet moved away from the hierarchical model of medicine in which expertise-driven providers are expected to advise “powerless” patients. However, in her view, this is where digital health has the power to disrupt the system – platforms that give patients choice AND the responsibility to make educated and self-informed decisions.

Panel Discussion

Q: Should patients have access to their own data?

Dr. Edmund Jackson (Hospital Corporation of America, Nashville, TN): The obvious answer is, “Of course they should.” But there’s quite a bit of subtlety here. This is not Facebook or Twitter. We are being led to change by government and that’s part of the problem. We’ve got the cart before the horse in a very serious way. The ability to get at data is hindered by governments and the structure of this enterprise.

Dr. Jay Boisseau (CEO, Vizias, Austin, TX): What do you mean though? Do you think data should be free?

Dr. Jackson: I’m just playing devils advocate. Of course, patients should have access to data. But I’m saying that in this age where we’re asking if autism is caused by vaccinations, should patients really have access to data to make poor decisions?

Dr. Mini Kahlon (Dell Medical School, Austin, TX): They should have access. The question is over what patients are going to do with the data. I can guarantee you that my care would be better if my physician had access to my data from my previous operations and appointments. I don’t want to have to worry about systems not talking to each other. I want to be able to give my doctor my data and not have to worry about if the system is broken.

Dr. Jackson: Clearly you should get your data and be able to move it around. I’m just playing devil’s advocate.

Q: I get my data, but I’m a medical professional. You get lost. Those data are hard for patients to understand. What are your thoughts on not just getting the data, but making it understandable.

Mr. Matthew Cowperthwaite (Director of Research, NeuroTexas Institute Research Foundation): The hospital may collect 1,000 data points per day on you. Were your linens changed? Was the trash taken out? There needs to be a middle ground that provides a snapshot of diagnoses made and next steps. There is potentially a danger in having access to a lot of data that patients don’t necessarily understand.

Dr. Jackson: We are actively working on portability. But we have a technology problem. EMRs are from the 80s and 90s. They are not the open platform to which we are used to in technology.

Dr. Boisseau: What we have is a technology legacy problem.

Dr. Kahlon: But on getting your down data, guess who led, by far? The VA. The VA has on so many levels one of the best systems. They’ve had some negative news, but that takes away from the lessons we need to learn. The VA is the first institution to give you your EHR on demand. We should look for examples wherever they may be. 

Comment: On the notion of granting access to data, I would highlight that the most prominent prosecution cases in this field have been when providers have not given data to patients.

Q: Mr. Jackson, you spoke about the Healthcare Corporation of America doing research on the data front. To what extent are your physicians accepting of your results, especially when they are not necessarily published and are not verified through clinical trials?

Dr. Jackson. The thing is that data sells itself. The data can be sold on its efficacy. If it reduces workflow and increases efficacy, the doctors see that. They are data driven people, after all.

Mr. Cowperthwaite: I agree. Physicians tend to be into data. Perhaps they don’t always understand the stats behind it all. However, they are moved by the findings that show that changes are needed; they are moved by the findings that suggest they should change the way they do things.

Dr. Kahlon: I would highlight that the approach to publishing is very different nowadays. That is some of the benefit for a researcher in being partnered with an academic environment. Hospitals, on the other hand, also want academics to come in and do research, but once they get their data, hospitals want a say in what gets published or not.

Q: I come out of the financial services industry. Often, innovation is stymied by complying with rules and regulations, as well as the use of legacy systems. However, I’ve found that if there is improvement to the bottom line, either increased revenues or decreased costs – improved outcomes and decreased costs in healthcare – typically someone senior can push through and make something new happen. Who is that person to move things forward in your organization?

Dr. Kahlon: We worked to bring new ideas from the bottom, choose the best ideas with rigor and evidence, and select some. But the only way was from the CEO down. This is the challenge with system-wide change. If you are going for the top-down approach, you need blessing from the top. You can also use data to disrupt from the outside. There are ways to mine this stuff. Someone should do something like Some are stunning. We pay a lot more for premature babies in the NICU than normal pregnancies. Think about where those incentives lie.

Dr. Boisseau: The private sector is trying to do things around it. 23andMe was big at SXSW last year. They tried to do it in a different way. They did not play completely nicely with the FDA. Now they are back but can only provide ancestral data. 800,000 people paid $100 to have their genomes sequenced. Genentech signed a $60 million deal with them – there is so much good data there, even with partial sequences, that there is a real ability to design new drugs. I hope there is disruption from the private sector as well as academic disruption.

Q: Could you comment on the sharing of genetic information? It doesn’t sound like anyone is disagreeing that patients should have access to their data. Even the government agrees. However, it sounds like the problem is more about the business model. I know that this problem has not been solved. What are some business models that seem promising? That seems to be the crux of the problem.

Dr. Kahlon: We are not going to wait on the government. We have different views on the pros and cons of sharing genetic information. But I don’t think we’re waiting on the government.

Dr. Jackson: No, we’re not waiting. It’s that we’re scared on them. Part of the problem is that they have enforcement power. They use the power of HIPAA and say that if you make a mistake, we’re going to fine you tremendously. So we have to act with great caution. The government is sending us mixed signals.

Dr. Kahlon: I would push you on that, because I would say that government institutions have implemented standards and guidelines quite quickly, such as the VA. However, private and academic institutions are moving slowly. They are not incentivized along the right axes. No one is penalizing them.

Mr. Cowperthwaite: Let’s think about the VA. It’s easier to implement the system. It’s a closed system that very different from a hospital. After all, a patient might go to the ER for a broken leg and get 20 bills. This is because radiologist and orthopedist and ER doctor all have their own sets of data. It’s a very complicated problem because we have so many providers along the point of care.

Dr. Jackson: Getting back to sharing genetic information: If we provide information that is incorrect, what is the consequences of that?

Dr. Kahlon: I had a false positive screen when my baby was born, and I had a right to see it and it was horrible because I thought my child was going to die within a few months to four years of being born. This is a very real ethical issue. I don’t know where I would exactly fall on it, but I want to emphasize that there’s a reasonable question to be asked here. Especially when we consider that these data has have probabilities and statistics and volatility associated with them. I think we have to balance harm and good. I just want to point out that there’s a discussion to be had here.

Q: You said the VA system is the most successful – is that the business model?

Dr. Kahlon: No, I don’t want to make generalizations. The VA fully implemented the ability to get one’s digital electronic health record.

Q: What is your opinion of government organizations and their efficacy at the highest levels?

Dr. Kahlon: I don’t know about their effectiveness. I think there needs to be some coordination, and I’m humbled in front of the task they have. However, I’d suggest that we need local experiments that accelerate coordination beyond where we are.

Q: The conversation is lacking on the quality of data. What’s HCA doing about improving clinical documentation, and the ability to share data within the community?

Dr. Jackson: On data quality, it’s a stepwise process. First, we had to collect the data. Three years ago, there weren’t systems to give an enterprise view. Now, we have the data, and we need to get quality around it. We’re improving clinical documentation with more structured fields. We face issues like prescriptions. We had 300,000 ways of describing the way of taking a pill. We cannot send that to a patient – it’s not meaningful. It’s a serious concern – we’re on it.

Mr. Cowperthwaite: The less and less fields you enter, the better. There are more dropdown menus and finite structure fields.

Q: The challenge with that is that the two of you walk in with flu systems. On paper, you look exactly the same. As a patient, I treat the two of you exactly the same, but you have very different medical histories.

Dr. Jackson: I don’t think we would be treated the same.

Dr. Kahlon: There is great excitement about Big Data – we could say, “screw quality.”

Dr. Jackson: Goodness!

Dr. Kahlon: We have to keep improving quality. It is a painful process. We should do it, however, with new predictive approaches, we can deal with some noise.

Dr. Jackson: Different types of noise; let’s be clear.

Dr. Kahlon: There are much cleverer ways to do this – Bayesian approaches that are more probabilistic. I think both sides need to happen. We need to get much better about getting systematic about entering data. And we need to get better at leveraging dirty data across a variety of data across domains.

-- by Varun Iyengar, Adam Brown, and Kelly Close