SXSW 2016

March 11-15, 2016; Austin, TX, Full Report – Draft

Executive Highlights

Howdy from Austin, where our team has been on the ground for the past five days amidst summer weather and healthcare highlights (and some music, too, of course!) from this annual gathering of the Technorati, music lovers, and film buffs. Though we have yet to get final attendance numbers, this year’s meeting felt less busy relative to last year – shorter lines, less hustle and bustle – though that didn’t translate across the board onto the content side, where we did have some valuable learnings. Our report includes plenty on diabetes technology and broader learning on digital health. See our top highlights below followed by ten themes and our extensive detailed discussion and commentary.

Top 12 Highlights

1. A panel discussion with Lilly’s Dr. Justin Wright, Livongo’s Mr. Manny Hernandez, and mySugr’s Mr. Scott Johnson, moderated by our own Ms. Kelly Close offered a critical assessment of the inadequate “one-size-fits-all” diabetes care model and speculated on opportunities to transform the system.

2. Ms. Dana Lewis discussed her experience spearheading the OpenAPS artificial pancreas project, sharing new news that 40 people (as of this weekend) are now using the open source closed-loop system (up from two dozen as of a NYT article last month).

3. Jeff Dachis shared that the One Drop iPhone app’s Bluetooth-paired meter and unlimited test strips subscription service will ship this October, pending FDA clearance and approval. The monthly bundle is expected in the range of the cost of a co-pay for those paying out of pocket (presumably ~$30-$50) and insurer discussions are ongoing to, the company hopes, offer a free version. In new news, One Drop also plans to launch 24/7/365 expert diabetes coaching – we love the move we’re seeing toward more nurses/community health workers helping people with diabetes, not always doctors.

4. Omada Health’s Mike Payne shared several gems on the company’s digital diabetes prevention program: (i) it has treated “about 50,000 patients” so far (up from “more than 40,000” in February); (ii) it expects five more publications in peer-reviewed journals this year (building on four published to date – wow!); and (iii) “it’s a possibility that Medicare will start reimbursing soon” for Prevent. That last would be fantastic to see since 83% of people over 65 in the US have diabetes or pre-diabetes.

5. This morning, UCSF announced a partnership with Yes Health, a new mobile app for diabetes prevention. We learned more at Yes Health’s sizeable exhibit hall booth this week – its differentiators include a consumer-facing business model ($39 per month for four months), personal coach, all-mobile offering, and four-week free trial. We hope many technology-enabled diabetes prevention programs can be successful – 86 million Americans will require it!

6. UCSF’s outraged Dr. Urmimala Sarkar shared grim results from a small study (n=10) testing four diabetes apps’ usability in low-income patients. The apps were not named, but this underserved population had trouble even entering a blood glucose value and Dr. Sarkar called for better usability testing among the underserved populations.

7. J&J Senior Director of Federal Affairs and type 1 patient Ms. Jane Adams gave a wide-ranging talk on the current healthcare climate that touched on two majors topics relevant to diabetes: (i) the ridiculous bureaucratic red tape muddling the process of Medicare coverage of CGM; and (ii) the huge resistance among US payers and government to pay for any kind of chronic care management.

8. Digital health key opinion leader Dr. Daniel Kraft (Singularity University) peered into the future of healthcare with a rapid-fire look at technologies that are going to redefine medicine. His multiple examples included Omada’s Prevent, Livongo for Diabetes, Dexcom’s G5, WellDoc’s BlueStar, and Abbott’s FreeStyle Libre.

9. Adaptive Path’s Mr. Chris Risdon delivered a fascinating lecture on how to shape behavior through design, creating apps and services that both fit into people’s lives AND prompt them to act. Through a number of stories, he covered examples of how leveraging psychology and empathy – “embracing the squishiness of life” – can influence behavior in significant ways. A highly insightful talk with lots to think about for diabetes.

10. IBM’s Cognitive Lab was THE most popular place to be at SXSW, drawing a long line to experience a spectrum of impressive computing applications: better health insights through the Watson/Under Armour partnership; automatic personality profiling and finding the perfect gift; and a lot more. We continue to believe Watson has high potential in diabetes to make sense of overwhelming data and limited provider time.

11. An excellent session on wearables argued that the biggest barrier to uptake is quality evidence, not cost, product design, or regulatory concerns. Pharma and health systems are showing greater interest in wearables (with some exciting early data in heart failure), but evidence will be lacking until wearables are used as part of packaged medical interventions.

12. FDA Associate Director of Digital Health Mr. Bakul Patel opened his presentation with a telling gesture – “FDA should NOT be thought of as a barrier. We are your partner!” He was greeted with ample applause from the digital health audience.

Honorable Mentions

1. NY-based startup Hero Health impressed us in the exhibit hall with its new smart appliance that stores up to 10 different pills at one time, intelligently dispenses them on schedule, and sends reminders and tracks use with a mobile app. Yeah, it’s expensive – so is people not taking their medicine.

2. Admetsys’ Mr. Jeff Valk shared updated plans for its in-hospital closed-loop system: (i) CE Marking by mid-year 2016; (ii) a Scandinavian launch in late 2016; (iii) an FDA PMA submission in early 2o17; and (iv) a US launch in late 2017. Lots of unknowns here.

3. NeuroMetrix announced that its next-gen Quell Wearable Pain Relief Technology began shipping at the beginning of March and that an upgrade program for first-gen users will launch “next month.” It was terrific to see the IDEO-designed (at least partly!) device here at SXSW.

4. President Barack Obama delivered a passionate and inspiring keynote before a PACKED auditorium (and thousands of remote viewers), advocating for the technology industry to reexamine its relationship with the government and participate in solving the nation’s biggest problems – “I expect you to step up and get involved because the country needs you”. His impassioned speech was a major hit.

Themes

We’re leaving SXSW with ten themes, including increasingly blurred lines between digital health and medical devices; a strong desire for more research in motivation and behavior; a continued dearth of clinical data supporting digital health; greater pharma interest in using digital health in clinical trials; growing tension as consumer technology comes to healthcare; a continued resistance to pay for chronic care management; near universal hatred of EMRs; more subscription and pay-for-outcomes business models; and growing progress in conversation-based AI and virtual reality.

Table of Contents 

Top 12 Highlights

1. A panel discussion – featuring Lilly’s Dr. Justin Wright, Livongo’s Mr. Manny Hernandez, mySugr’s Mr. Scott Johnson, and moderator Ms. Kelly Close – offered a critical assessment of the inadequate and all-too-common “one-size-fits-all” diabetes care model and speculated on opportunities to transform the system. Themes emerging from the hour-long discussion included the importance of putting patients first (“the thing we lack is the personal story occurring in a patient’s home at 11 pm at night after a crappy week”), the need to generate actionable insights from existing data (“We need to free our human caretakers to be more human because as much automation and algorithmic magic we’re going to do, I still need a lot of human contact in my diabetes care”), the utility of retrospective vs. real-time data (skewed toward the latter), the value of CGM in itself (“revolutionary”), the lack of enthusiasm for endocrinology in the medical community (“Who is going to make diabetes care an engaging pursuit?”), and the need for interoperability between devices and platforms (the status quo is “BURDENSOME”). Please see our detailed coverage below for what we thought were the most notable quotes and some of the most prominent themes of the panel discussion in which we also heard about Lilly innovation being developed at its new Innovative Center in Cambridge, MA.

2. Type 1 patient Ms. Dana Lewis discussed her experience co-creating the OpenAPS artificial pancreas project, sharing new news that 40 people (vs. “two dozen” as of last month) are now using the open source closed-loop system. It was an informative session as Ms. Lewis detailed the desperate need that inspired this grassroots action: the ceaseless fear of hypoglycemia, the overwhelming cognitive burden of diabetes decision-making, and the bureaucracy that has slowed what should be relatively straightforward innovation. She was very persuasive, recounting the journey that has brought the OpenAPS project to where it is – from the insight that initially brought her to hack her CGM (“I wanted louder CGM alarms and industry was unwilling to accommodate my n=1 problem”) to the early iterations of the DIYPS algorithm (that enabled remote monitoring/predictive alerts) to the final form that has closed the loop for a growing number of patients around the world. As a reminder, the OpenAPS hardware is not for the faint of heart, consisting of a Medtronic insulin pump, Dexcom CGM receiver, a raspberry pi device for running Linux OS, a CareLink USB stick to enable communication to the pump, and a battery pack. That device burden is a testament to badly patients want automated insulin delivery. Ms. Lewis did not share new updates per se though she was certainly a fantastic advocate for those with diabetes (on a public stage, no less) and for patient-led innovation. Like Nightscout, we hope OpenAPS is pushing the FDA and industry to move faster, and the recent slew of companies announcing pivotal trials (Medtronic is complete; Tandem, Animas, and Insulet are moving) suggests that may be happening. See our detailed coverage below for more from Ms. Lewis and this movement.

3. Jeff Dachis shared that the One Drop iPhone app’s Bluetooth-paired meter and unlimited test strips subscription service will ship this October (pending FDA clearance and approval). The monthly bundle is expected in the range of the cost of a co-pay for those paying out of pocket (presumably ~$30-$50) and insurer discussions are ongoing to hopefully offer a free version. In new news, the subscription will also include 24/7/365 diabetes expert coaching (CDE or other qualified professional) via in-app text chat, plus a 12-week in-app structured educational program for diabetes and prediabetes rooted in behavioral science. The meter timing is behind that shared last April, which expected a launch in “early 2016” – that was clearly an ambitious timeline, and in seeing the meter up close for the first time, One Drop has made several tradeoffs to keep the costs in line (e.g., thumb-drive sized, no color touchscreen as the original artist rendering indicated). An “experienced” manufacturer is making the strips and technology inside the meter; we’re not sure whom. The business model is highly ambitious – a startup offering unlimited strips and a Bluetooth meter with expert coaching for a single low monthly fee – though we hope it is possible, since many users could benefit from greater access to SMBG. Since launching last April, the free diabetes logging app on iPhone has gathered 15 million user data points from over 500,000 user sessions in 650 cities and 140 countries. That translates to an average 1.5 minutes each and up to two minutes each in the last month. It’s impossible to back those metrics out into number of active monthly users (we assume in the ~5,000-15,000 range), though Mr. Dachis highlighted several new features that are driving recent uptake (comprehensive food and restaurant database, auto-medication logging and reminders). One Drop is currently a featured app in the Apple app store (among just 20 apps), a notable vote of confidence from Apple. The app has a middle-of-the-pack 3.5 stars and 73 reviews overall vs. mySugr’s dominant 4.5 stars and 2,185 all-time reviews (to be fair, mySugr has been in market since 2012, while One Drop has only been in market for ~11 months). The unlimited strips subscription service could certainly accelerate uptake, though obviously a lot has to go right on the business and regulatory fronts for that to happen. More details below on the business model, app, meter, SWOT analysis, and our questions.

  • We believe many apps and connected meters can and should succeed in diabetes. We see innovation accelerating in both areas and hope patients and providers benefit from connectivity and more comprehensive and passively acquired data (glucose, food, exercise, sleep). The hope, of course, is better insight to drive therapeutic change.
  • LabStyle Innovations launched a similar self-pay BGM subscription model today for its Dario system in the US, though it is more expensive than One Drop’s hope for unlimited strips in the range of a co-pay (more details below). We’re seeing more diabetes tech companies going this route or planning for it (Bigfoot, LabStyle, Livongo, One Drop) and assume the trend will continue as healthcare payment shifts to value.

4. Omada Health’s Mike Payne shared several gems on the company’s digital diabetes prevention program: (i) it has treated “about 50,000 patients” so far (up from “more than 40,000” in February); (ii) it expects five more publications in peer-reviewed journals this year (building on four published to date); and (iii) “it’s a possibility that Medicare will start reimbursing soon” for Prevent. The latter was amazing to hear, though “possibility” and “soon” of course do not provide specifics on when it could happen. As a reminder, CMS does pay physicians to watch and manage patients remotely, but it’s a measly $42 a month. Omada has always prioritized publishing outcomes, and we especially loved this comment from Mr. Payne: “In 5-10 years, we believe that the currency of selling stuff in this space is holding up peer-reviewed publications when you walk in the door.” Omada is investing “millions of dollars” in clinical research over the next one to two years and will continue to price its Prevent program based on outcomes (per pound lost or maintained). Mr. Payne said Omada is also doing continuous A/B testing to tweak reminders and subject lines (a huge advantage of digital), while new machine learning work is taking human bias out of hypothesis formulation – that sounds fascinating and we wonder what novel questions it might come up with. Omada has about 300 employees now, up from 185 employees as of September.

  • More broadly, this panel on outcomes in digital health included a lot of relevant commentary for diabetes companies, particularly this quip from Rock Health’s former managing partner Malay Gandhi: “Bundled payments is where it’s going to be. Selling your widgets at high margins – that business is dead. It’s the service model now.” We hope both that device and drug companies are listening – and that patients will step up. More quotable quotes are below.

5. This morning, UCSF announced a partnership with Yes Health, a new mobile app for diabetes prevention. We learned more at Yes Health’s sizeable exhibit hall booth this week – its differentiators include a consumer-facing business model ($39 per month for four months), a personal coach, an all-mobile offering, and a four-week free trial. The Yes Health mobile app enables tracking of meals and exercise, and serves as a platform for communication with personal health coaches who offer tailored recommendations on nutrition, fitness, and mental wellbeing. Yes Health is available for free on the Apple app store; after a four-week trial period, users can opt to enroll in the 16-week core program that includes personalized coaching and chatting 14 hours per day. The out-of-pocket cost is a fairly reasonable $39 per month ($156 for 16 weeks). After completing the core program, users can continue to use the app for $15 per month. The consumer-facing business model and all-mobile offering (no web) with its instant 1:1 coaching feedback is slightly different from Omada Health’s Prevent program, which is normally rolled out by insurers or employers and heavily leverages a web interface (though it also has a mobile app). As part of its new partnership with UCSF, Yes Health will create an “opt-in” function for users to share their de-identified behavioral data (diet, exercise, sleep patterns) with UCSF Diabetes Center researchers. The goal is to combine this behavioral data with biometric data (including genetic information) to ultimately develop individualized prevention and treatment programs for type 2 diabetes. In return, Yes Health will gain access to UCSF diabetes content and experts for improving its app, and will also benefit from advertising through UCSF media. This partnership represents an early vote of confidence for Yes Health, who soft-launched in October 2015 – the feedback so far has been positive from its private beta trial last year and from its early app store reviews since its general market roll out (a small sample of six reviews on the app store, all 5/5 stars). Management told us average self-reported weight loss and increased activity thus far is close to the Diabetes Prevention Program target.

  • We are excited to see another player in diabetes prevention, particularly given the bio of founder and CEO Alex Petrov (who has built consumer brands at PayPal, Safeway, and Nestle). The big questions we have are: (i) can the app sustain engagement? (ii) can it prove outcomes? (iii) how large is the market for consumer self-pay diabetes prevention?; (iv) is Yes Health’s coaching scalable and is the business profitable at this price; (v) how much harder is generating awareness and scaling in a direct-to-consumer model, especially given the volume of apps? Overall, we hope many technology-enabled diabetes prevention programs can be successful – 86 million Americans will require it! More details below.

6. UCSF’s outraged Dr. Urmimala Sarkar shared grim results from a small study (n=10) testing four diabetes apps’ usability in underserved low-income patients. The app names were blinded on the slides, though Harvard colleagues selected them as the four “best” among the top 200 Android and Apple offerings. Usability in the study was measured on two very basic dimensions: can people with diabetes (i) enter their own blood sugar into an app and (ii) retrieve their average blood sugar from an app? Patients’ were given the apps on a tablet and performance was classified on a spectrum (successful – “straightforward” or “prolonged”; partial; unsuccessful). For BG entry, only one of the four apps excelled with ~85% entering a blood glucose successfully (straightforward + prolonged), while the worst app saw just two out of ten patients successfully enter a blood glucose. In two of the four apps, a sizeable 30% of patients were completely unsuccessful in entering any blood sugar number at all. Results were far more disturbing on the second dimension (finding average BG): zero patients were successful with two of the four apps, while the “best” two apps saw just ~10%-20% successfully locate the average BG. Dr. Sarkar ardently called for more participatory app design (include patients and providers!) and expanded usability testing in low-income patients. This study was a bit hard to unpack, as it was tiny (n=10, though some apps only n=5), we don’t know which apps were used (e.g., free apps?), we’re not sure how accepted the usability testing was, and we’re not sure what kind of patients these were beyond “low-income” (e.g., did they speak English?). Still, these are concerning data and we agree that much more should be done to design technology and apps for low-income patients.

  • We can only speculate on why apps are not designed for underserved populations now: (i) many apps are “free” and don’t have a business model or funds to do extensive usability testing; (ii) these patients can be hard to recruit, particularly for small startups – they are not that easy to find (Dr. Sarkar grimaced “they’re on Facebook!” in one panel Q&A – yes, they are, but not open to everyone and not advertising a diabetes status); (iii) established companies design products for those likely to buy them, and low-income patients are generally not a target market for those looking to generate early revenue; (iv) product design can be completely different for low-income patients (e.g., Omada’s Prevent for underserved populations took over a year to develop and an ~$1 million grant); and (v) low-income patients can be stereotyped as “not tech savvy.”
  • US smartphone ownership is quite high even in low-income populations, according to Pew: 50% of those with an income <$30,000 and 71% of those with an income $30,000-$49,000. This suggests apps are a promising avenue to reach low-income patients with diabetes. We expect usability to improve with time, particularly as more established companies with R&D budgets can devote funds to proper development and usability testing in low-income populations. Business models for apps are always a challenge, though we hope as more established companies invest more in apps, R&D work can focus on the most vulnerable populations.

7. J&J Senior Director of Federal Affairs and type 1 patient Ms. Jane Adams gave a wide-ranging talk on the current healthcare climate that touched on two major diabetes topics: (i) the ridiculous bureaucratic red tape muddling the process of Medicare coverage of CGM; and (ii) the huge resistance among US payers and government to pay for any kind of chronic care management. This was among the most valuable and candid talks we heard at SXSW on healthcare financial incentives, particularly from the perspective of someone living with type 1. Ms. Adams clearly appreciated that CGM can provide critical information for making therapy decisions (continuous data, alarms, trend information) and alluded to her personal experience with the “life-saving” technology. More poignantly, she even shared that her mother (of Medicare age) also has type 1 and has had to foot the bill herself. She did not dig into the weeds of Medicare’s primary objection to covering CGM (i.e., adjunctive labeling) but did stress that there is huge federal resistance and tradeoffs to paying for delayed savings of any kind in healthcare and diabetes:

  • “When I point out that Medicare doesn’t pay for CGM, it seems ridiculous. However, when you price it out and realize that it would cost the government $1 billion as a whole, then you begin to understand why they ask whether to put that money there or in Alzheimer’s.” [We had not ever heard that estimate and wonder if it is accurate – we do know the entire system is spending nearly $200 billion on diabetes alone each year, so even $1 billion, if it could keep patients safer and on target, would be fine by us.]
  • “Congress doesn’t invest in the long term. It invests in the right now … They are more than happy to pay for a kidney transplant but won’t pay to avoid it in the first place.” This makes one’s blood boil, of course.
  • We’ve long seen that the challenge with government and payers alike is convincing them of cost long-term savings of managing diabetes now– the incentives simply are not aligned and Ms. Adams stressed that overcoming this inertia is going to prove very challenging. Ultimately, we salute Ms. Adams for further driving awareness about this issue at SXSW and in Washington – she was one of the most charismatic speakers we have heard on healthcare incentives; we include our favorite quotes and her insights into chronic disease management below.

8. Digital health key opinion leader Dr. Daniel Kraft (Singularity University) peered into the future of healthcare with a rapid-fire look at technologies that are going to redefine medicine – his hundreds of examples included Omada’s Prevent, Livongo for Diabetes, Dexcom’s G5, WellDoc’s BlueStar, and Abbott’s FreeStyle Libre. He did not share insights on any of the devices – this was truly rapid fire, less than ten seconds per slide – though we found it striking nonetheless to hear him call out diabetes as a “high-impact” area. The reference to Libre particularly caught us off-guard (Dr. Kraft did not mention the device by name but there was no mistaking the slide he showed) and it struck us as a testament to how far and wide knowledge of Libre has spread, and how knowledgeable Dr. Kraft is about what’s happening. As a reminder, FreeStyle Libre Pro (retrospective, blinded) is still slated for a ~mid-2016 US launch, while Abbott has said that “optimistically”, the real-time FreeStyle Libre (consumer version) could come by the end of 2016. Dr. Kraft also delivered an incredibly optimistic view of the coming decade in healthcare technology, urging attendees to think about how far research has moved in the past five years (iPhone 1 vs. iPhone 5) and how we can continue to expect change in the future (iPhone 5 vs. iPhone 10). He suggested that the exponential growth of technology (Moore’s Law) is going to influence the future of medicine delivery in ways that it is tough to even IMAGINE right now: Will it be malpractice NOT to use artificial intelligence in medicine? Will we move from organ donors to data donors? How soon will we each have our own personal “check engine light” indicator? His remarks got us thinking about a future that moves beyond siloed and reactive care to a paradigm of continuous and proactive care. Is it possible? Yes. Is it easy? Certainly not … But if one message resonated upon leaving the presentation, it’s that that vision isn’t as far away from reality as we might think. As Dr. Kraft concluded, “Honestly, the future in many ways is already here …”

9. Mr. Chris Risdon (Adaptive Path, San Francisco, CA) delivered a fascinating lecture on how to shape behavior through design, creating apps and services that both fit into people’s lives AND prompt them to act. Through a number of stories, he covered examples of how leveraging psychology and empathy – “embracing the squishiness of life” – can influence behavior in significant ways. He pointed to the Red Cross’s short-term text capability for the Disaster Relief Funds during the 2010 Haiti earthquake as a prime example of a mobile system that capitalized on human impulse (feelings of sadness, sympathy) rather than rational long-term thinking (much harder!). The glucose sensing smart contact lens was mentioned as an example of working products seamlessly into lives though he did not comment on its development – still, it was great to hear accessible glucose data highlighted as an area of unmet need. Ultimately, this talk had many broader lessons for diabetes, whether it be for general product development (“every design decision influences the user”), reflecting on what works and what doesn’t (“most digital solutions require people to opt in rather than fitting seamlessly into their lives”), and the value of actionable insights (analytics that influence the immediate future vs. reflecting on the past). This was a very insightful talk and we highly recommend product designers take a look at the full write-up below.

10. IBM’s Cognitive Lab drew a long line to experience a spectrum of impressive computing applications: better health insights through the Watson/Under Armour partnership; automatic personality profiling and finding the perfect gift (just enter a twitter handle and Watson analyzes it); personalized food, mixed drink, and TED Talk recommendations; robots that can play rock, paper, scissors against humans; and more. The Under Armour section of the Lab showed off the Watson-powered UA Record app, which compares a user’s exercise, nutrition, and biometric information to peers: “During the past seven days, your average workout duration was 37 minutes. Men your age average 48 minutes per workout” or “Men your age work out an average of 1.7 times per week. That 4.3 less than you.” The social comparisons are paired with compelling educational messages like “90% of people who lose weight and keep it off for at least five years exercise at least an hour a day” or “An American Cancer Society study shows spending an excessive amount of time sitting (at least 6 hours per day) decreases a person’s life expectancy – regardless of whether they regularly exercise.” This is a potent combination – objective health statistics that compare users to peers + evidence-based health outcomes – and we wonder how it will play out in diabetes with the Medtronic and Novo Nordisk partnerships (e.g., Patients like you tend to do X in this case”). There was no mention of either diabetes partnership in the Lab, though several stats reminded us why computers like Watson are needed in healthcare: medical data are expected to double every 73 days by 2020 and 80% of health data are unstructured, which makes them invisible to current approaches. Watson can read 40 million documents in 15 seconds, including unstructured data and images.

  • There is a lot of hype around Watson and some are skeptical if the computer can live up to the promise; we’ll get a first look this summer when Medtronic and IBM launch a hypoglycemia prediction app. IBM is investing in Watson Health in a big way, most recently acquiring Truven Health for $2.6 billion – the company has data on the cost and treatment of more than 200 million patients. IBM has acquired four companies since it created the Watson Health unit last April, bringing the total spent to more than $4 billion. WOW. We salute IBM for taking an open ecosystem approach with Watson and hope it results in more actionable data for patients, more time for doctors to do what humans are good at (social intelligence, human connection), and big savings for the healthcare system (finding patients at risk, recommending the right treatments at the right time, etc.).

11. An excellent session on wearables argued that the biggest barrier to uptake is quality evidence, not cost, product design, or regulatory concerns. Pharma and health systems are showing greater interest in wearables (with some exciting early data in heart failure), but evidence will be lacking until wearables are used as part of a medical intervention. Validic CTO Drew Schiller shared fascinating early data from I Get Better/Brocton Medical Center, which gave 31 post-op heart failure patients a blood pressure cuff, a connected weight scale, and an activity tracker. Brocton would normally expect eight of the 31 to be readmitted in 30 days; zero were actually readmitted, saving the system over $200,000 immediately ($28,000 per readmission). The details of the program were not discussed, but we assume it included some remote monitoring component. Said Mr. Schiller, “That is the proof that is going to help leverage change and use of wearables.” Garmin’s Allison Swelin highlighted increasing pharma interest in using wearable devices in trials, with one partner expressing hope for cost savings and efficiency (“Every day a trial is late to market, they lose $2 million in revenue”). Unfortunately, Mr. Schiller said pharma trials still use wearables as a secondary input (if at all), and the field will only move once a drug or device is paired with wearables in a packaged intervention. “With those proof points, the industry will explode. It goes beyond ‘this is nice’ to ‘this makes real financial sense.’” We love the idea and believe there is so much potential in diabetes to wrap drugs with connected wellness devices (activity trackers, scales), connected glucose monitoring, and remote-enabled care. Ms. Swelin said that payer discussion have “started to pick up” in the last 12 months, and employer wellness programs really see wearables as a must-have now.

12. FDA Associate Director of Digital Health Mr. Bakul Patel opened his presentation with a telling gesture – “FDA should NOT be thought of as a barrier. We are your partner!” It was an appropriate introduction to a valuable lecture on the Agency’s forward-thinking approach to regulating digital health technologies and its effort to adjust to the challenges of working in this rapidly changing space. The presentation itself did not share updated thinking beyond what Mr. Patel discussed at DTM 2015, which centered on the Agency’s  risk-based approach to digital health: promote innovation, patient engagement, and safety. Balancing these goals, in his view, is about figuring out the amount of risk one can take to reap the ultimate benefits. He specifically highlighted the importance of designing products that fail safely, sharing a tolerance for imperfection that we could not have imagined coming from the FDA a few years ago: “Digital health doesn’t have to be perfect. Nothing in the world is defect-free. We need a system that works within those bounds. That’s the reality.” It’s terrific to hear this open mindset and it is clear that the FDA understands the need not to overregulate health technology. Dr. Patel did not talk specifics about diabetes, though we have seen this area moving in the right direction with rapid automated insulin delivery progress (MiniMed 670G pivotal complete + 80% of patients in the FDA continued access program), the rapid approvals of Dexcom’s G5 and MiniMed Connect, and several new cloud-connected meters coming to market (to name just a few).

Honorable Mentions

 1. NY-based startup Hero Health impressed us in the exhibit hall with its new smart appliance that stores up to 10 different pills at one time, intelligently dispenses them on schedule, and sends reminders and tracks use with a mobile appsee a video here. The Wi-Fi-enabled device is pretty expensive at $499, but it could help with several massive problems we see in people with diabetes and obesity: (i) taking the right pills at the right time, particularly those on polypharmacy (diabetes, blood pressure, lipid, etc.); (ii) receiving reminders to take pills (the app sends users notifications); and (iii) tracking what pills are actually taken (the app notifies caregivers). The device stores up to a month’s supply of 10 pills, which are loaded into one of ten cartridges and slid into the machine. The medication plan is set up on the device, and after pushing a button, it will dispense the right combination of pills like a vending machine (e.g., blood pressure + vitamins + statin + diabetes drugs). Reps said that some people are buying them for their loved ones, which allows them to get notifications like, “Mom took her 8 am pills. The device just launched, meaning the cost will hopefully come down over time.

  • Another promising approach we’ve seen for polypharmacy overwhelm is Pill Pack. This intriguing startup simplifies the user experience of taking oral medications through very convenient packaging, technology, and personalized service. Patient medications come in a roll of individual tear- packs organized by date and time; are mail delivered every two weeks; and refilled automatically. This video really shows it best. The focus is on oral medications, and given how many pills people with diabetes take, this has a lot of potential to help with adherence and taking doses correctly.

2. NeuroMetrix announced that its next-gen Quell Wearable Pain Relief Technology began shipping at the beginning of March and that an upgrade program for first-gen users will launch “next month.” We first heard about the newly FDA-cleared product at CES 2016 and today got details on some of the new features: (i) extended battery life (~one day longer); (ii) the ability to control Quell through a mobile app (now available on both Android and Apple devices); (iii) “more sophisticated” sleep tracking technology – see below; (iv) more detailed therapy tracking over three months; and (v) an alternative stimulation mode that gives users a fast gentle pulse instead of constant vibration for more customizable pain relief. We got to play with the sleep-tracking features of the app in the exhibit and were intrigued by the wealth of data NeuroMetrix is collecting (sleep duration, sleep quality, body position, time out of bed, and repetitive leg movements). Accurate sleep tracking is valuable for diabetes and we hope that Quell can give users insight into this aspect of health – that said, we’re not entirely sure about the everyday utility of metrics such as sleep time spent on left side vs. sleep time spent on right side. [Perhaps for patients with chronic pain, correlating sleep position with next-day pain might provide insight?] Ultimately, we continue to hear that Quell is helping a TON of patients with neuropathy and reviews have been glowing. Reps in the exhibit hall, too, expressed great enthusiasm for what has been a “really positive response” so far, and we look forward to seeing the impact of the next-gen devices on sales in coming quarters.

3. Admetsys’ Mr. Jeff Valk shared updated plans for the company’ s in-hospital closed-loop system: (i) CE Marking by mid-year 2016; (ii) a Scandinavian launch in late 2016; (iii) an FDA PMA submission in early 2o17; and (iv) a US launch in late 2017. Talk about ambitious goals! He laid out concrete go-to-market plans – no partner, initial launch in critical and perioperative care (particularly in cardiovascular surgery), and an initial cost of ~$250/day – sharing that the company is hoping to raise ~$12 million in the near future to fund the aforementioned goals. We learned that Admetsys’ approach now titrates IV dextrose (previously, they reported glucose) and insulin in response to plasma glucose measurements at five-minute intervals. Work has been very stealth to date (we have yet to see clinical trials outside the research setting and this remains a major question) though the company has shown encouraging results in three small studies (n=43): 97% time in the tight range of 80-125 mg/dl with no hypoglycemia (<70 mg/dl). [Of course, IV glucose/insulin delivery makes this sort of control very achievable in controlled settings.] Overall, we have yet to be convinced that Admetsys is the real deal considering how little we still know - our immediate questions concern regulatory hurdles, liability concerns, the challenge of convincing hospitals that the cost-benefit analysis comes out in their favor, and the challenge of convincing providers to trust the system. Edwards backed out of hospital CGM and Medtronic has not reported any progress with its Sentrino in-hospital CGM in Europe. The work is unquestionably needed – we are not aware of any others trying to commercialize an in-hospital closed-loop system – though the road ahead seems equally steep. See our coverage below for more.

4. President Barack Obama delivered a passionate and inspiring keynote before a PACKED auditorium (and thousands of remote viewers), advocating for the technology industry to reexamine its relationship with the government and participate in solving the nation’s biggest problems – “I expect you to step up and get involved because the country needs you”. To that end, President Obama identified three areas where SXSW attendees are primed to make a difference: (i) helping the government run better through digital platforms and technology (e.g. simplifying processes such as applying for healthcare, student loans, and Veterans’ benefits); (ii) tackling “big problems in new ways”, such as through the Precision Medicine Initiative’s advanced biomedical research; and (iii) employing big data, analytics, and technology platforms to make civic participation easier (e.g. voting, countering violent extremism). He also showcased the government’s dedication to bridging the cultural gap between government and technology through the US Digital Service, a team of talented designers and engineers from around the country tasked with “making government services simple, effective, and efficient.” Beyond briefly referencing the PMI and the challenges in launching healthcare.gov, President Obama did not address the government’s plans for technology in healthcare in detail. Still, we think its fantastic to see the nation’s leader advocating so tenaciously for private-public partnerships and bringing the best tech talent into government. The implications have huge potential in healthcare and we hope to see the next president continue the momentum.

  • In a subsequent panel on the Precision Medicine Initiative, FDA’s Mr. Taha Kass-Hout shared an update on PrecisionFDA, a secure, cloud-based online platform launched last December for testing and developing next-generation sequencing (NGS) genome assays. According to Mr. Kass-Hout, PrecisionFDA now has over 1,000 members worldwide including biotech companies, pharmaceutical companies, regulators, patients, and clinicians. Further, one-third of total members live outside the US, and one-third of US members are from Silicon Valley. Mr. Kass-Hout also discussed the PrecisionFDA Consistency Challenge, which the White House announced a few weeks ago to evaluate and improve the precision and accuracy of DNA test results. We’ve seen a rapid proliferation of consumer and professional testing kits in recent years, but these tools are not useful in the clinical setting without established benchmarks for precision, accuracy, and links between genes and different therapeutic options (e.g., you are more likely to benefit from exercise, you are an exercise non-responder). We are encouraged by the FDA’s leadership in promoting innovation and rigor (and healthy competition!) to improve genomic testing, and hope to see reliable, easily interpreted, and truly actionable products emerge as a result.

Themes

1. The line between digital health and medical devices is getting much blurrier. Wrist-worn devices can now measure heart rate, which can be used for wellness purposes (resting heart rate) or medical purposes (heart rate variability, detecting a heart murmur). FDA has taken a very hands-off approach to wellness devices like Fitbit, though the lines are getting much fuzzier with heart rate (and probably other new sensors). This is a particularly important debate for wellness CGMs like Sano (expected to launch this year as of the last update). Data privacy is equally a big concern in digital health (wellness devices do NOT fall under HIPAA), and ethical questions are increasingly coming up (e.g., heart rate data could tell when people are sexually active, etc.). Federal agencies spanning these areas include FDA and FCC, and speakers said they are not always on the same page.

2. Two of the most frequent replies to hard questions at SXSW: “Well, we just don’t know what motivates people with chronic disease” and “We need more research on behavior.” Let’s get this research done!! What can we learn about motivation from patients who really take care of their diabetes? What can we learn from patients who didn’t care about their diabetes and then took control of things (what Dr. Bill Polonsky calls, “Something clicked”)? More broadly, is motivation THE missing link in diabetes, or is it really a combination of factors that prevent patients from doing well: cost, side effects, life circumstances, stress, etc.? The silver bullet in many talks this weekend was often positioned as, “If only we knew how to motivate people with diabetes, we’d have perfect outcomes.” That’s probably not true, though a greater understanding of motivation will undoubtedly help drive much better patient-clinician conversations. What’s the path forward and who will spearhead it?

  • diaTribe.org has written a bit on motivation, and Adam has been thinking a lot this week about “perceptions of time” – are some “future-oriented” individuals more likely to take care of their diabetes vs. “past-oriented” and “present-oriented” individuals (different people have different orientations)? See this book and this TED talk for more on this, and please let us know if you have thoughts on this topic!

3. Digital health needs more robust evidence and far more actionable data to see greater uptake. How can it be combined with traditional drugs and devices? While technology is clearly outpacing peer-reviewed literature, we hope more can be done with pilots and real-world data to prove digital health’s efficacy. We also believe digital health needs to be combined with diabetes drugs and devices in a bigger way (apps, connected wellness devices, telehealth) – these interventions do not need to exist in siloes, and probably shouldn’t if they are to be deployed effectively.

4. Pharma is showing greater interest in wearables – how should clinical trials harness digital health (activity trackers, connected scales, apps, electronic dashboards)? Digital health could help improve the understanding of how drugs impact daily life, collect valid real-world data, enhance the efficacy of drugs, reduce clinical trial costs, and even speed time to market. How could Apple’s Research Kit and other software enable virtual trials that eliminate wasted time on travel and in-person logistics? What is the killer app for clinical trials? If the current pharma model cannot continue (as some believe), what will the clinical trials of the future look like?

5. There is a growing tension as consumer technology comes to healthcare: rising patient expectations, clinical skepticism, product design tradeoffs, and internal company conflicts. The merging of tech and healthcare can be synergistic, but it also brings tradeoffs that are not always easy for companies: consumer advertising claims vs. FDA label claims; medical apps forced to keep pace with consumer apps; consumer design principles vs. medical device design; clinicians demanding published data vs. tech’s push to launch minimum viable products and iterate.

6. There is still resistance to pay for chronic care management. Many comments on reimbursement use the words “transition” and “changing” to refer to the shift to value-based payment – this is generally good news for digital health. Near-term, however, many speakers were pessimistic: everyone realizes chronic disease is a problem, everyone knows someone needs to pay for it, and no one wants to step up. Will value-based payment solve this problem? What new business models will be compelling in 10 years?  

7. EMRs are still universally despised, with little optimism they will get better any time soon. Rock Health’s former managing partner Malay Gandhi was blunt on the timeframe for EMR improvement: “I wouldn’t expect it for at least ten years. We blew about $30 billion. I’m sorry, doctors don’t have the tools to pull in data from hundreds of apps. I wouldn’t expect it any time soon.” Ugh. Is this a solvable problem? We hope Washington is thinking hard about pushing it forward.

8. We’re noticing more subscription and pay-for-outcomes business models and wonder if this will come to diabetes in a bigger way. Will all devices and drugs be paid based on outcomes in ten years? Twenty years? Will payers reimburse for a single bundle of diabetes products (e.g., drugs + devices + education + wellness devices)? Another quip from Mr. Gandhi put it well: “Bundled payments is where it’s going to be. Selling your widgets at high margins – that business is dead. It’s the service model now.” This certainly bodes well for patients.

9. A growing trend in healthcare is conversation-based artificial intelligence (AI), something more companies seem to be pursuing in diabetes. We’ve been interested to play around with the Lark app for nutrition, which offers the “feel” of talking to a person, but in a completely automated fashion. This strategy could be pretty compelling if the apps can sustain use – it will be all about providing enough value relative to the user costs of interacting with them. We are hopeful!

10. Virtual reality (VR) was all the rage here at SXSW, with some hints of huge potential in healthcare and diabetes. USC’s Dr. Leslie Saxon discussed an immersive VR experience that provides a glimpse into the reality of living with diabetes-related vision loss (the technology was the winner of a recent USC/IEEE “Hackathon”). We see this as a particularly powerful tool for inspiring glycemic control in teenagers with diabetes. We could imagine many uses for the technology, including patients interacting with an avatar version of their doctor; patients visualizing their own diabetes progressing; more interactive education or motivational gaming; more entertaining exercise; physician training; treatment for mental health; and more. The nice thing about some virtual reality system is low-cost: Samsung’s Gear and Google’s Cardboard both insert a standard smartphone into inexpensive <$100 goggles, enabling virtual reality viewing. Pretty cool stuff and we hope it really comes to fruition in healthcare.

Detailed Discussion and Commentary

Smashing “One Size Fits All” Diabetes Care At Last

Panel Discussion

Kelly Close, MBA (The diaTribe Foundation, San Francisco, CA); Manny Hernandez (Livongo Health, Mountain View, CA); Scott Johnson (mySugr, Vienna, Austria); Justin Wright, PhD (Eli Lilly, Cambridge, MA)

This panel discussion offered a sharp and realistic assessment of the many challenges and opportunities in diabetes care. Moderated by our own Kelly Close, the wide-ranging conversation featured Dr. Justin Wright, Mr. Manny Hernandez, and Mr. Scott Johnson in front of a packed room, many of whom had a connection to diabetes. Among other things, the panelists discussed the value of CGM (“revolutionary”), the lack of enthusiasm for endocrinology in the medical community (“Who is going to make diabetes care an engaging pursuit?”), the need for interoperability between devices and platforms (“burdensome”), the importance of putting patients first (“the thing we lack is the personal story occurring in a patient’s home at 11 pm at night after a crappy week … understanding the patient journey and the user experience – that’s where the investment needs to be made …”), the need to generate actionable insights from existing data (“We need to free our human caretakers to be more human because as much automation and algorithmic magic we’re going to do, I still need … human contact in my diabetes care”), and how they think about the utility of retrospective vs. real-time data (skewed toward the latter). Below, we have distilled some of the most notable quotes and some of the most prominent themes of the evening – we’ll be updating this piece later this week.  

Themes and Quotable Quotes

  • Do we even know how to personalize care in diabetes? Do we understand the specific behaviors and motivators and barriers different patients face? This may be the biggest obstacle to smashing one-size-fits-all care. Scott Johnson emphasized, “We don’t really entirely understand the challenges to personalizing diabetes care. I have yet to meet someone with diabetes who doesn’t want to take care of themselves. We all want to live a healthy life but there is a knowing-doing gap and we don’t understand it very well. We need to spend some time and energy figuring this out. It’s not the mechanical part of taking shots that we have trouble with – it’s all up here [points to head]. I think it’s about incorporating elements of smiles to make diabetes suck less.”
  • If you could invest in only ONE of the following, which would you choose: better patient tools to individualize therapy, better physician tools, or better payer tools?
    • Patient tools came out ahead in this discussion, with a majority of panelists citing the self-managed nature of diabetes. The most memorable analogy came from Dr. Wright, who described the serious need for individualization that is impossible to predict in clinical practice: “We’re seeing a convergence of the life sciences, of clinical innovation, and of engineering best practices. The thing we lack is human perspective. We aren’t capturing the personal story that is occurring in a patient’s home at 11 pm at night after a crappy week. Understanding the patient journey and the user experience – that’s where the investment needs to be made.”
    • That said, panelists noted that clinicians are overburdened and need more help managing patients. Better provider tools could make a serious difference in individualizing care – speeding clinical efficiency, pointing out problem areas, flagging patients that need help vs. those doing fine, etc. In other words, technology can help titrate limited clinical bandwidth to match where a population of patients needs the most help.
      • “I think there should be smarter ways to crunch my diabetes data and help my doctor reach insights about my life with diabetes. He doesn’t have time to analyze all of my diabetes data and it’s unrealistic for me to expect him to do so. We need smarter systems for him to come to those conclusions while I’m with him, so that he can provide me with lifestyle changes that will impact my data by the next time I see him. But really, he NEEDS smart systems to help him do that. He can’t be asking me, ‘What did you have for breakfast five days ago?’ That doesn’t make sense when we have all these tools available.” – Mr. Johnson
    • Ms. Close emphasized the importance of technology tools that are engaging and trustworthy: “We’ve seen technology enable other fields to scale – healthcare and chronic disease has been more elusive. Not only do patients and doctors have to have access to new solutions and tools, they have to trust them and want to use them every single day. That means we have to create solutions that are as addictive as the iPhone.”
  • To truly individualize therapy, patients need more clinically meaningful insights. The field is facing some important challenges: (i) what is the right data to collect; (ii) how does it affect blood sugar; (iii) how do we make sense of it for each person; and (iv) how do we present it to people with diabetes and providers in a way that helps them make sense of diabetes and take action.” No panelist could point to an amazing example of this happening within diabetes when asked. We think Dexcom has taken some encouraging steps with Clarity, Abbott has done good work with FreeStyle Libre’s AGP download report, Medtronic is working on next-gen CareLink reports and the hypoglycemia prediction app with Watson, Glooko is trying to track insulin dosing titration, etc. The technology exists, and panelists pointed out that bigger questions related to how much burden it will place on patients to collect the data, how much providers need to be in the loop, and what the regulatory bar is for truly compelling patient decision support.
    • “I would describe personalized medicine through a persona – let’s say her name is Maria. She’s the kind of person that thinks about everyone in her family before herself. She’s struggling with her diabetes control but nobody knows and it goes completely unnoticed. It only gets worse and worse. The thing is that we have the tools and mechanisms to passively acquire data and understand that data, so we can know when Maria needs help and can know that she’s the type of person that would not ask for help in the first place. We need to apply that kind of thinking to diabetes.” – Mr. Hernandez
    • “Wearing a CGM has been revolutionary for me. And for those of us who adopted the technology early, it’s easy to forget that it’s still a very young technology...CGM also just collects a ton of data; so much data that I can’t make sense of it. That can be feeding into a system that can run invisibly in the background, smartly combining it with other pieces of data that we’re not properly collecting yet, making sense of all of that – that’s what personalized healthcare should be.” – Mr. Johnson
  • But to truly individualize diabetes care, do we need to shift out of the “retrospective data review” model of therapy change to a “real-time data analysis and notification model?” Panelists did not have time to debate this, but it’s worth thinking about! Shouldn’t patients get advice in the moment rather than three months later? Doesn’t it make sense to understand cause and effect in the moment vs. trying to recall what happened at breakfast two weeks ago? The value of retrospective data today is mostly for physicians in appointments (patients don’t really download); where could that go once more devices are cloud-connected and algorithms can automatically analyze it? For which patients is real-time data analysis unrealistic and unnecessary? For which patients is it vital? When will reimbursement catch up?
  • Device interoperability was called a “low hanging fruit” in the quest to personalize diabetes care. Ease of data downloading and interpretation in the clinic is critical, and panelists stressed that standardization would play a big role in efficiency. The panel felt that report standardization was more of an industry problem, and commentary elicited a notable response from Dr. Wright: “From an industry perspective, I can say that we hear this. We’re in the lab putting this stuff together and the debate that we have about which software to utilize goes on for hours and hours, and it’s never perfect. The analogy I like to make is to roads and transportation. After all, the roads are all the same but the cars are different and customizable. We need to get to the same place in diabetes.” We agree that an ideal scenario would be a standardized output that could be followed with proprietary industry outputs on page two, three, etc. that could differentiate products. More on this in our coverage of February’s AACE CGM Consensus Conference.
  • Panelists also wondered if technology could negatively influence the “humanity” of clinical interactions. Indeed, they stressed that personalized care does NOT mean a transition to automated systems that replace clinicians – rather, personalized care should ideally free providers to work more closely and intimately with patients. Panelists recalled appointments with clinicians where the provider asks a question, types into the EMR, asks another question, types into the EMR, etc.; it’s like playing ping pong against yourself! Technology is a good servant but a bad master, and we hope it increasingly moves into the background in healthcare (even scribes alone would help!).
    • “I would love to see more work moving into computers and allowing caretakers to do ‘human’ work. Fundamentally, computers are good at being computers and humans are good at being humans. We need to free our human caretakers to be more human because as much automation and algorithmic magic we’re going to do, I still need a lot of human contact in my diabetes care.” – Mr. Johnson
    • We have to take the whole person with diabetes into consideration if we really want to achieve personalized care.” – Mr. Hernandez

Full Panel Transcript

Ms. Close: If you could invest in one of the following which would you choose and why? Patient tools? Physician tools? Or better payer tools? And let’s assume access isn’t a problem.

Mr. Johnson: Respecting that I have to choose one, I would have to choose patient tools. However, creating a tool isn’t enough. You have to empower the people using the tool to use it correctly. I think one of the most exciting ways to do that is peer-to-peer support, and I think that’s a scalable way in today’s world.

Mr. Hernandez: I find it extremely hard to pick one, and I would say yes to all three. Having to pick one, I would say physician tools. An overwhelming number of people with diabetes are diagnosed by their primary care physician, and practices there are far from current. There is a campaign going on right now led by the AMGA – the American Medical Group Association – called Together 2 Goal to put best practices together in one box. This doesn’t sound like a revolutionary concept, but sadly it is. We need a level of basic care as groundwork. 

Mr. Wright: We’re seeing a convergence of life sciences, clinical innovation, and engineering best practices. The thing we lack is human perspective and the personal story that is occurring in a patient’s home at 11 pm at night after a crappy week. Understanding the patient journey, the user experience, the investment needs to be made there.

Ms. Close: Absolutely. Understanding the patient story the patient story is key and we’re definitely not where we need to be on peer-to-peer support.

Mr. Johnson: We’re definitely not where we want to be with peer-to-peer support. However, that’s an unfair answer because peer-to-peer support isn’t easy. It’s a scary world for healthcare providers.

Ms. Close: A lot of people who were diagnosed decades ago like us would say that things like peer-to-peer support have improved the most. The next important piece is making sure it is acknowledged and recommended by physicians.

Mr. Johnson: We need a good clinical education foundation.

Ms. Close: Tell us about a technology or tool that shows a lot of promise for personalizing diabetes care.

Mr. Hernandez: For the first time in my memory, I’m starting to see things like what a company like Companion Medical is doing, which is trying to turn an otherwise dumb delivery device like a pen into a smart delivery device via Bluetooth connection. It sounds basic, but for the first time the innovation in the insulin delivery isn’t just on pumps but on devices that are used by the majority of people on insulin. 

Dr. Wright: What that ladders up to is having algorithms in the background and understanding what data we’re collecting. There are several talks this week referencing algorithms, and as we build data sets and pull this together, these adaptive algorithms are going to bring data together. This is one of the most exciting things over the next five years.

Mr. Johnson: Wearing a CGM has been revolutionary for me. And for those of us who adopted the technology early, it’s easy to forget that it’s still a very young technology. A CGM allows me to wear a small sensor that feeds data to a receiver – in this case my iPhone. It alerts me if my blood glucose goes above or below range. It’s also just collecting a ton of data; so much data that I can’t make sense of it. That can be feeding into a system that can run invisibly in the background, smartly combining it with other pieces of data that we’re not properly collecting yet, making sense of all of that, that’s what personalized healthcare should be.

Ms. Close: We need to move more towards standardized communications protocols.

Mr. Hernandez: Like any of you in the audience who have T1D you know that if you show up at your doctor’s office, you hand over all your devices and then go through the fascinating process of finding all the right cords and cables. If you’re lucky, then you might get your data downloaded. It’s very primitive.

Mr. Johnson: I would love to see more work moving into computers and allowing caretakers to do “human” work. Fundamentally, computers are good at being computers and humans are good at being humans. We need to free our human caretakers to be more human because as much automation and algorithmic magic we’re going to do, I still need a lot of human contact in my diabetes care.

Mr. Wright: If you go back 15-20 years and you’re a patient in North Dakota, you’re alone trying to figure it out. Peer-to-peer and our ability to connect people across the planet is the best opportunity for scalability and for building a network of information and people.

Ms. Close: It’s cool to think about. Advocacy has come a long way and patients feel strongly, which is what we are seeing in the #WeAreNotWaiting campaign. We are seeing more enthusiasm at FDA and changes in interoperability. Maybe it’s slower than we want it but it’s still a step in the right directly. I’d also like to see a greater focus on mental health, which I think would translate into fewer long-term complications. A lot more could be put into mental health. Overall, when you think about the ideal world, what does fully personalized care look like? How optimistic can we be?

Dr. Wright: There has been a transition of the medical home. The question is where does this really exist? Honestly, the medical home is everywhere. As we continue to invest in tools that are adaptive – whether from the bed to the kitchen to the car – it’s about having the tools and connecting people to those tools. Those are investments we should continue to make.

Mr. Johnson: I think we are finding ways to take some of the very mechanical aspects of a data driven disease and translate that into human care. We can abstract all the layers of data and make it a little more visible and present it in ways that resonate with the individual. The common scenario at endocrinologist appointments is that the doctor will look at your log book and ask about high or low glucose numbers from five days ago. How many of you can remember what you had for breakfast five days ago exactly? Neither can I. We need to abstract some of the more mechanical data parts of diabetes, and turn as much as we can into a more human centered care picture.

Mr. Johnson: At the end of the day, each device has its own software and its own data and its own report. You see blood glucose on one report and insulin delivery on a separate report. None of these devices talk to each other, and it’s very burdensome to put this together.

Dr. Wright: From an industry perspective, I can say that we hear this. We’re in lab putting this stuff together and the debate that we have about which software to utilize goes on for hours and hours, and it’s never perfect. The analogy I like to make is to roads and transportation. After all, the roads are all the same but the cars are different and customizable. We need to get to the same place in diabetes. So it’s on all of us.

Dr. Wright: I could still probably personalize the system around me but is it truly adaptive and truly personalized? It’s a system in the background adapting to me hour-to-hour and day-to-day.

Mr. Hernandez: I would describe it through a persona – let’s say her name is Maria. She’s the kind of person that thinks about everyone in her family before herself. She’s struggling with her diabetes control but nobody knows and it goes completely unnoticed. It only gets worse and worse. The thing is that we have the tools and mechanisms to passively acquire data and automatically upload the data, so we can know when Maria needs help and we can know that she’s the type of person that would not ask for help. It doesn’t need to be people with diabetes that come to her aid; it could be others that love her. Perhaps her neighbor says, “let’s go for a walk.” So we need to apply that kind of thinking to diabetes.

Dr. Wright: I am reminded of the conversation we were having last night. With the way technology is progressing into healthcare, we’re currently going in the wrong direction. Doctors are buried in their laptops but we need them to be focused on us. We need them to ask us how we feel.

Ms. Close: I wanted to talk about how devices and drugs are analyzed in RCTs where we control for a lot of what happens in real life. But that doesn’t tell us a lot about real life itself. Should we be taking a look at how commercial products get tested?

Mr. Johnson: As an individual consumer, it’s my feeling that many of the companies in diabetes would love to be much more agile, but are a bit restricted by the systems and rule they have to play by. They would love to iterate more quickly. There is a good reason for some of those rules but they need to progress more and loosen up so we can move faster.

Mr. Wright: For years we’ve operated under the notion of vertical segmentation. We’ve attempted to make one product that satisfies the entire patient or customer population. The notion of horizontal segmentation is where the product meets a number of patients, but takes new tools to understand patients in new ways.

Ms. Close: It is cool to watch how Lilly has done this. They’ve worked with IDEO on Trulicity – their new GLP-1 agonist – and you see the results. It’s 7% of Lilly’s revenue and 50% of its growth. That gives you more resources to go back and invest in R&D.

Dr. Wright: To arrive at products and solutions is a journey. It requires new models of innovation and reinforces that you can get to a more personalized level.

Mr. Hernandez: The reality of trials is that they miss that we are a lot more whole than our type of diabetes or an A1c. We are people before we are people with diabetes, and when you leave the very controlled and safe environment where that trial is taking place, people are constantly making decisions and changes that are not the way the product was conceived of and tested for. Our friend Jeff Hitchock who runs an organization called Children with Diabetes likes to say that “life is lived off label”. Minority groups are severely underrepresented in clinical trials. This is particularly important given the fact that if you look at type 2 diabetes, that’s where the biggest incidence lies. Do we know which particular therapies are the best for those folks? We need to start looking at more real world situations.

Ms. Close: We can use our voices as patients, too. There is a lot more we can do to make suggestions and make things like clinical trial recruitment sound more interesting. The next thing I want to ask about is face-to-face interaction. How do we scale these solutions?

Mr. Johnson: I keep coming back to the same point. I think there should be smarter ways to crunch my diabetes data and help my doctor reach insights about my life with diabetes. He doesn’t have time to analyze all of my diabetes data and it’s unrealistic for me to expect him to do so. We need smarter systems for him to come to those conclusions while I’m with him, so that he can provide me with lifestyle changes that will impact my data by the next time I see him. But really, he NEEDS smart systems to help him do that. He can’t be asking me, “What did you have for breakfast five days ago?” That doesn’t make sense when we have all these tools available.

Mr. Hernandez: If a particular patient has extremely good glycemic control, does that person need to see physician for anything beyond just refilling their prescription? Probably not. Would that time and energy be best spent focusing on where physicians can have the most impact, on where people are falling through the cracks?

Dr. Wright: The social media tools we have right now are one of the most promising aspects of the future. I just heard that 150 million Americans have an iPhone. The number of social media channels is huge. You can find someone that you are comfortable with via video or via chat or something else.

Ms. Close: I want to go back to what Scott was saying about getting your data to doctors so that they can analyze it and are not going to need to think about retrospective data. Will we really get to this place where we just use real-time data and not retrospective data?

Mr. Johnson: The question is, is retrospective data even necessary? I think it is necessary but not something that I should have to worry about as a person with diabetes. It should be something that the computer analyzes in the background to make smarter adjustments to my treatment going forward. I want as many layers of abstraction between me and my retrospective data as possible. Do you know anyone whose only job in life is to take care of their diabetes? I don’t. And that’s not the way it should be.

Dr. Wright: On the product development side, I would point out that retrospective data is valuable. It helps us design guardrails and helps us assess the systems we need to build. That is of immense value.

Mr. Johnson: Is it possible that other people’s retrospective data could help me? Yes. There is immense value in the data and we have no idea how to use it correctly. We’re only collecting a of what’s possible to collect and we’re not close to making sense of it yet.

Ms. Close: We know that doctors’ time constraints are not going away, and large, consolidated networks aren’t moving in reverse. What do you think will help healthcare providers the most?

Mr. Hernandez: We are even applying one size fits all to say we need doctors to see patients the same number of times. Some patients need more support to be empowered, and others don’t need as much. We need to help healthcare providers understand how to customize this care to fit their time constraints.

Ms. Close: It is a crisis that there is the same number of endocrinologists as when we were all diagnosed, and there are 20 million more diagnosed patients in the US alone. We need to make diabetes care an engaging pursuit for you doctors and help them be successful. I don’t see these numbers changing or enthusiasm increasing around the endocrinology profession.

Dr. Wright: I think a lot of it starts with education on the physician side. I think one of the things I’ve struggled with is “who’s the customer?” The patients are. I think that’s something we can improve on a population level. The patient is actually a customer who shows up and I think there’s a great opportunity to embed this idea that the patient is the customer.

Mr. Hernandez: If you were to get the same customer service experience that you get in healthcare with your cable provider or phone company, you would ditch them for somebody else. When you are living with a chronic condition you are kind of stuck – you can’t say, “screw you I’m going to Verizon” – these are decisions that many of us cannot afford to make. We are left as the glue that has to hold all the pieces together on top of all the things that managing a chronic disease entails. One month ago, I placed an order for new pump supplies, and they never showed up. Nobody called or reported why, so I called the 1-800 number and got an automated system that told me the company is still waiting on materials from my doctor and hadn’t gotten any faxes yet. As it turns out, they faxed it to another office. Those are super low hanging fruit opportunities. The time spent on that – that raises my blood sugar. You are making my diabetes worse!

Ms. Close: It’s interesting as we think about community health. Overall, I’m hearing a lot about value-based healthcare. How do you think about smashing one-size-fits-all diabetes care from a payer perspective?

Mr. Johnson: This may seem like a simple view, but sometimes it seems like it’s difficult to get supplies from payers. Some of us are limited in the number of test strips we can have on a daily basis. This doesn’t make any sense because the more tests I do, the more likely I am to avoid complications. It seems like the way payers are thinking about this is that by the time I encounter a complication, I’ll be somebody else’s problem. However, more personalized care means giving me the ability to manage my life. Overall, I’ll be a cheaper person to take care of.

Mr. Hernandez: At Livongo for example, we are working on addressable hassles. People need test strips to test their blood sugar, and the number they need varies. They don’t need to fight over that. Rather than charging for individual strips, we have a subscription model. You don’t even have to bother talking to the 1-800 number, more strips will just show up on your front doorstep. Eliminating those kinds of hassles will reduce burden across the system.

Ms. Close: You have to know where you are to figure out how you can get there. There is so much movement around democratizing numbers. The device I use is not approved in the US – FreeStyle Libre – but I just wave the reader over the sensor and this is how I get my blood glucose. This is really hard to get approved in the US. We’re not advertising for this company but this is just to say that patients are definitely getting their voices heard. With that in mind, my next question is: Who is doing the best job of personalizing care? Who is doing the least impressive job?

Dr. Wright: A few days ago, I was just looking at a graph of the insulin absorption profiles of all the vendors that make insulin. The release time varied so much from hours to minutes. This is after 50 years that we’ve achieved this highly customizable approach. We need that bridge back to everything we’re discussing.

Audience Q&A

Q: I’m from the UK and live in a very different treatment model than in US. Do you guys know anything about biometric measurement and moving wearable tech into clothing? That seems like it could be really useful for diabetes. Sports are very difficult to manage and we cannot predict where our sugar is going to be. My second question is whether you you think preventative medicine is actually the right way to approach all of this? A lot of money is made from drugs. Is there ever going to be a focus on preventative medicine?

Dr. Wright: We opened our innovation center in September of last year. Our first purchase was a sewing machine, and we are training a bunch of guys on how to use it. We are moving towards full product systems. For years it’s been all about the pharmacology, the biotechnology, the molecule, but ten years from now it will be about the full product system. I’ve seen this across other spaces in wearables and technology deployed in new ways. It’s absolutely coming.

Mr. Johnson: It makes me think that we’re not making use of that fact that our smartphones are collecting a lot of data that we could be using in the diabetes world. There are a lot of things that we – “we” meaning the big picture – need to fit into our world. We need to convince people to use the data that’s out there. The challenges that many are facing us are: (i) what is the right data to collect; (ii) how does it affect blood sugar; (iii) how do we make sense of it for each person; and (iv) how do we present it to people with diabetes in a way that helps them make sense of diabetes on a day-to-day basis. I think it’s happening but we’re not there yet.

Ms. Close: This speaks to the importance of personalizing diabetes care so you can be met where you are. To the second question, I would say that the US has something to learn from other cultures in terms of lifestyle, diet, exercise, and nutrition. We are spending so much more on heart attacks, strokes, kidney dialysis, and long-term complications. It’s not even drug spending that is the majority. We want to see much less spending on the long-term complications of diabetes and instead focus on optimizing health.

Dr. Urmimala Sarkar (UCSF, San Francisco, CA): I am a physician at the SF General Hospital and am a diabetes researcher at UCSF. As you’ve been talking, I’ve been thinking about my patients, many of whom have not been included in this discussion. All my patients are publicly insured and many do not speak English. They are not well represented in clinical trials, which I think is a very important priority. Shouldn’t my patients be better represented in patient advocacy? They are not asked for the input that goes into diabetes apps and as a result, they see much of the customization and empowerment in apps as complexity. This is not fair. So I want to challenge all of you to incorporate these patients into your thinking.

Ms. Close: We totally agree. I would say that I don’t think anyone is thinking about highly complex applications of technology. We’re really trying to make things simpler.

Mr. Johnson: Speaking more on the social media side of my experience, we actually find it very difficult to get people living with type 2 diabetes to engage. We don’t know where that world is in the social media space.

Ms. Close: In the words of the great academic Carol Dweck, maybe they aren’t engaging yet and more needs to be done. It might have something to do with the stigma of diabetes. Justice Scalia had diabetes, but no one knew that until after his autopsy. I would love to see more from researchers on what they are doing to understand this.

Mr. Johnson: So I’m curious for your thoughts. How can we do a better job of engaging these groups?

Dr. Sarkar: I think it’s about bringing people to the table. I think my patients would have a hard time finding a blog. I don’t think you can overstate the stigma effect. When you do usability testing with apps, who are the patients in those studies? When are we bringing in my patients? These are patients who have trouble taking two medications a day. We need to bring them to the table.

Mr. Hernandez: I think your comment is absolutely spot-on. Smashing the one-size-fits-all model is fundamentally not just for engaged and empowered patients and truly for everyone. If an experience or tool demands more from a patient, then that person is going to fall flat on his face regardless of whether or not he is engaged. I hear you about the importance of finding the cases that fall through the cracks and making sure that we cater to them. Otherwise, this system is starting to crack and then it’s going to collapse if we don’t address it. For a magical experience, patients can’t see all these layers of technology and complexity and algorithms.

Ms. Close: It also comes back to how are we funding all of this. We hear there is no more money coming from the NIH, and have to think more about public private partnerships. There has never been this kind of wealth creation in US, yet there has also never been this level of health inequality. If we can’t figure out how to change things it’s on us.

Q: When you think about all the advances in diabetes care, it seems like there is an increased risk of disassociating the primary care physician? How can we help them take advantage of the data out there?

Dr. Wright: I think there’s an opportunity. One of the things we are doing at Lilly is looking at novel innovation models. It lets us bring a host of partners to the table. We wouldn’t have had this ten years ago. Sanofi is in the building next door and I’ve had many conversations over lunch with them. It takes new and novel partnerships to accomplish this.

Ms. Close: Industry has to be behind interoperability as well. The device I’m using spits out a one-pager for my doctor. But industry says that’s Abbot’s thing. Having industry get behind patient voices and amplifying those will really help.

Q: Two things that I’ve heard in the past few weeks: (i) that the most common cause of amputations is diabetes; and (ii) that these amputations are normal. How do we do public education?

Mr. Hernandez: I’d love to talk more at length afterwards. That is a painfully true reality. Blindness is another place where we are not well enough. Short of being born blind, diabetes is a leading cause of blindness.

Ashley Boyd (Mad*Pow, Boston, MA): I’m really excited about this panel. A comment is that I feel we are missing something really important. I live with type 1 and diabetes sucks. What are you doing to address motivation? There is so much we can gain there. Not just do patients have to take medication every day, but they also have to maintain their management long term. How are you addressing motivation and incorporating those insights into products or services?

Mr. Johnson: Well, I think that one of the big things is that we don’t really entirely understand the challenges to personalizing diabetes care. I have yet to meet someone with diabetes who doesn’t want to take care of themselves. We all want to live a healthy life but there is a knowing-doing gap and we don’t understand it very well. We need to spend some time and energy figuring this out. It’s not the mechanical part of taking shots that we have trouble with – it’s all up here [points to head]. I think it’s about incorporating elements of smiles to make diabetes suck less.

Ms. Close: I think companies are working to engage people and starting to break out of silos. We are all in diabetes-centric models. A lot of us are trying to learn from Silicon Valley and successes in technology. Adam Brown has written a ton about this at Diatrib.org/motivation. Getting this thinking out is really important. There are many researchers looking at this that need more funding. Adam and I interviewed them asking what they could do with $5 billion, and they said they just need $50,000 to finish their studies.

Dr. Wright: A comment I’d add is that I’ve had conversations on medication adherence with many smart people. The conversation always ends with motivation. That’s where the investment needs to be made: truly understanding human behavior around motivation.

Q: Physicians want to spend more time with patients and patients want to spend more time with physicians. Payers want better care, too, and we have devices that are getting there. However, the population isn’t really doing better. So when do you see this all converging and really helping patients?

Dr. Wright: I think we talked about a number of things today, such as developing technology that can run in the background that will make a difference. We need to have a lot of touch points. As we evolve though, I have confidence that people will get what they need.

Mr. Johnson: I honestly don’t understand where the disconnect is either.

Ms. Close: Now for our lightning round, what do you think is the single most important thing the audience should take away from this panel?

Mr. Johnson: This a tremendously big thing to try and unwind and it should excite all of us working in the space.

Mr. Hernandez: That we have to take the whole person with diabetes into consideration to give personalized care.

Dr. Wright: The solution is going to come from all of us, so take action. If you have an idea start a company, start a blog, buy a sewing machine, get a 3D printer.

Ms. Close: We also need to better engage the payers because we need more investing here. We also need to better engage healthcare providers and show gratitude for the people making the system better.

Healthcare Innovation in Austin: Hype? Or Reality?

One Drop: Navigating diabetes, together

Jeff Dachis (Founder, One Drop)

Jeff Dachis shared that the One Drop iPhone app’s Bluetooth-paired meter and unlimited test strips subscription service will ship this October (pending FDA clearance and approval). The monthly bundle is expected in the range of the cost of a co-pay for those paying out of pocket (presumably ~$30-$50) and insurer discussions are ongoing to hopefully offer a free version. In new news, the subscription will also include 24/7/365 diabetes expert coaching (CDE or other qualified professional) via in-app text chat, plus a 12-week in-app structured educational program for diabetes and prediabetes rooted in behavioral science. The meter timing is behind that shared last April, which expected a launch in “early 2016” – that was clearly an ambitious timeline, and in seeing the meter up close for the first time, One Drop has made several tradeoffs to keep the costs in line (e.g., thumb-drive sized, no color touchscreen as the original artist rendering indicated). An “experienced” manufacturer is making the strips and technology inside the meter; we’re not sure whom. The business model is highly ambitious – a startup offering unlimited strips and a Bluetooth meter with expert coaching for a single low monthly fee – though we hope it is possible, since many users could benefit from greater access to SMBG.

Since launching last April, the free diabetes logging app on iPhone has gathered 15 million user data points from over 500,000 user sessions in 650 cities and 140 countries. That translates to an average 1.5 minutes each and up to two minutes each in the last month. It’s impossible to back those metrics out into number of active monthly users (we assume in the ~5,000-15,000 range), though Mr. Dachis highlighted several new features that are driving recent uptake (comprehensive food and restaurant database, auto-medication logging and reminders). One Drop is currently a featured app in the Apple app store (among just 20 apps), a notable vote of confidence from Apple. The app has a middle-of-the-pack 3.5 stars and 73 reviews overall vs. mySugr’s dominant 4.5 stars and 2,185 all-time reviews (to be fair, mySugr has been in market since 2012, while One Drop has only been in market for ~11 months). The unlimited strips subscription service could certainly accelerate uptake, though obviously a lot has to go right on the business and regulatory fronts for that to happen. LabStyle Innovations launched a similar self-pay BGM subscription model today for its Dario system in the US, though it is more expensive than OneDrop’s hope for unlimited strips in the range of a co-pay.

Overall, we like One Drop’s focus on giving users insight into their diabetes (glucose, food, medication, exercise), reducing manual entry through a Bluetooth meter and HealthKit integration, and adding accountability to a human within the app (either a patient or CDE coach). We believe many apps and connected meter can and should succeed in diabetes and we hope One Drop can follow through on its promises. More details below on the business model, app, meter, SWOT analysis, and our questions.

One Drop Business Model

  • One Drop is working with insurers to offer a low-cost or free option for the Bluetooth-enabled meter, unlimited test strips, and in-app expert coaching. Users that elect to self-pay can expect a monthly subscription fee “in the range of a co-pay” via credit card (presumably ~$30-50). The plan is to launch the “subscription diabetes care” concept this October, subject to regulatory approval this summer. A single monthly payment of $30-$50 – including unlimited strips and coaching – for self-pay users would be remarkable, but can the economics work out for a startup of One Drop’s size? Is the coaching scalable? Will insured patients cross-subsidize those without insurance? 
  • Similar to capitated payment, insurers are expected to pay One Drop per member per month; in exchange, users get the One Drop service and insurers presumably get access to the data. Mr. Dachis immediately countered privacy concerns: “Insurers have a vested interest in keeping you healthy so you make less claims. If they can offer suggestions, intervene, and offer additional wellness opportunities (based on your data), they make more money when you’re healthy. We’re working with insurance partners to help keep their members well.”
  • LabStyle Innovations launched a similar self-pay BGM subscription model today for its Dario system in the US, though it is more expensive than OneDrop’s hope for unlimited strips in the range of a co-pay. LabStyle has tiered pricing based on the length of commitment (one-month vs. three months vs. 12 months). At the most expensive one-month commitment option, LabStyle is offering strips for $0.40 each (i.e., $20 per month for 50 strips, $120 per month for 300 strips); in the least expensive scenario (12-month commitment), LabStyle’s price declines to $0.25 per strip (i.e., $12.50 for 50 strips, $75 per month for 300 strips). Assuming the average intensively managed patient tests five times per day (150 strips per month), LabStyle’s Dario would cost from $37.50 to $60 per month, depending on the subscription selected.
  • Will diabetes technology companies increasingly move towards a service subscription model? Can One Drop can pull it off? This self-pay pricing business model makes economic sense – most BGM’s would be happy to sell strips for $12 per box, especially knowing they have visibility to the recurring revenue via subscription. Of course, $12 is also below most co-pays. Glucose monitoring startups always seem to struggle with reimbursement, and even Sanofi’s iBGStar had a tough time when it launched. Data is of course valuable and multi-billion dollar companies have been built on collecting and analyzing it – can One Drop do the same, particularly with Mr. Dachis’ previous Big Data expertise?
    • Livongo has talked since last May about launching a similar consumer-facing subscription service for its cellular-enabled BGM, unlimited trips, and web platform. Where is it? The plan was to roll this out in late 2015, though we have not heard any updates from the company. Upending the business model of a longstanding industry is of course difficult, and it probably takes many sales cycles to gain the confidence of self-insured employers and payers. The economics are also challenging for BGM startups, where scale really matters for driving down margins. Presumably these companies are using contract manufacturers are the contract terms will only improve once scale reaches a critical mass.
    • Bigfoot is also planning to offer a subscription for its automated insulin delivery system; more on that here.

One Drop Meter Details

  • We got our first in-person look at One Drop’s meter, which is about the size of a thumb drive, has a black-and-white backlit screen, and single button. The meter is significantly stripped down from the iPhone-like color touchscreen device in the original artist rendering we saw last year, which is not a surprise. We continue to see a trend of down-featuring diabetes hardware and up-featuring the paired apps (e.g., OneTouch Verio Flex, Accu-Chek Connect, Insulet’s next-gen PDM) – it’s a smart way to iterate more quickly, capture better margins, and leverage the phone to offer a better user experience.
    • Mr. Dachis shared that the strip and meter contract manufacturer is very experienced in diabetes. We’re not sure who it is, but wonder if One Drop can achieve enough scale and margins to make the business model work.
  • One Drop is still putting a major focus on sleek iPhone-like product packaging, a leather meter case, customized colors, and hardware that doesn’t look like a medical device. Said Mr. Dachis,If you love your diabetes gear, you’ll use it. Right now, the equipment is awful and expensive. Empowering people with things they love is better than disempowering them with a lack of information or things they hate.”

One Drop App Details

  • One Drop just completed a pilot of its 24/7/365 coaching service with “universally positive” reviews (size not quantified, but the slide showed three reviews from people with type 2 diabetes). This will roll out into the free app soon to gain experience, though once the meter launches, the coaching will likely become part of the monthly subscription service. One Drop will also launch an in-app 12-week educational program, which sounds like a parallel to what WellDoc is doing with BlueStar.
    • Scalability is always a question with coaching, though Mr. Dachis emphasized that it’s “near-time,” asynchronous communication. The experts can respond in near-time, offering potential for a much larger ratio of users to experts. We wonder if the company’s approach can scale to many thousands of users and beyond; presumably some users will take larger advantage of the coaching than others. Given the company’s technology background, we could envision machine learning and AI handling some of the heavy lifting.
  • One Drop is currently a featured app in the Apple app store (among just 20 apps), a notable vote of confidence from Apple. Mr. Dachis noted several recent updates to the app:
    • Food library, restaurant menu searching. The app remembers frequent meals and adds up the carbs from multiple food items. The search does return a lot of results, though it is better than trying to guesstimate carbs when eating out. One Drop still encourages taking pictures of meals, something we agree is very useful for food awareness.
      • Easy food entry is a major obstacle in diabetes and health apps – pictures are simple but don’t offer enough data, while manual entry is too time consuming to be sustainable for the vast majority of users. One Drop has done a good job of integrating with a broad database, though we hope the entire app field figures out something better; search results are generally too broad and still require a fair amount of food awareness and selecting the option of best fit.
      • As an side, we’ve been pretty impressed with what the Lark app is doing with conversation-based AI. We are hearing about more companies pursuing this route in diabetes, which offers the “feel” of talking to a person, but in a completely automated fashion. It could be pretty compelling if the apps can sustain use.
    • Medication plan with reminders. Users can set up their medication plan in One Drop for a wide variety of diabetes drugs (e.g., Lantus 20 units at 10 pm; Januvia at 11 am), and One Drop allows pump patients to enter basal rates. The app reads data from HealthKit (e.g., Dexcom G5, Verio Flex, Accu-Chek Connect), though the CGM data visualizations are not informative (each glucose value appears as a pink dot, so CGM data is just a mass blob of dots).
      • One Drop users “love the medication reminders,” which Mr. Dachis told us is “driving huge use of the app.” The app sends notifications to take diabetes medications (e.g., Januvia), and One Drop is notably adding depression, hypertension, and lipid pills.
  • “People are finding their use cases for One Drop.” Mr. Dachis said the app is improving on the different cases that users would download it for: glucose, medication, food, and activity. In the last month, the average use session has increased to two minutes, up from the cumulative 1.5 minutes since launching last year.
  • One Drop send users motivational notifications and reminders, which can total up to 8-10 per day. Users can turn this feature on or off , though we always wonder about sustained engagement with so many buzzes. Some reviews in app store really appreciate the notifications, though we imagine they could be overload for some users.

One Drop’s Strengths, Weaknesses, Opportunities, and Threats

Strengths
- Apple Watch and Health Kit integration, unique manual logging approach with dials

- Community stream to congratulate fellow users and share tips

- Useful highest” and “lowest glucose moments” analysis, unique Map insights

- Motivational notifications, medication reminders, comprehensive food library

- Very frequent app updates to iterate product

Weaknesses*
- Data logging is mostly manual unless using HealthKit compatible devices

- Colored drop data visualization can be confusing

- Currently cannot do pattern recognition; show time-in-range; or offer search feature of previous data (e.g., “last time I ate pizza at this location”)

- No bolus calculator

- Dexcom CGM data visualization in app much harder to interpret than in Clarity

 

Opportunities
- Bluetooth-enabled meter with a more consumer feel

- Low subscription price for unlimited strips, especially for those without insurance

- Coaching, in-app education, particularly for type 2 diabetes and prediabetes

- Most people with diabetes have neither comprehensive data nor insights from it

Threats
- Challenging track record for most BGM startups: regulatory delays, achieving scale, broad reimbursement

- Payer relationships take time to develop

- Patient inertia to switching BGMs, particularly those already covered by insurance

- Crowded diabetes app and connected BGM marketplace, particularly mySugr, Glooko, Roche Accu-Chek Connect, J&J Verio Flex, Livongo, LabStyle

* OneDrop shared with us that many of these are on the near-term product roadmap

Close Concerns Questions

  • Regulatory: Is a launch by October possible if the meter has not been submitted to FDA yet? We have seen a lot of variability in the timing of connected meter clearances – Roche’s Accu-Chek Connect had a one-month review, while LabStyle Innovations’ just-launched Dario had an ~24-month review. What will One Drop’s regulatory process look like?
  • Strip manufacturing and accuracy? At this point, One Drop has not revealed who will manufacture the strips or meter, though we assume it is an experienced contract manufacturer.
  • Cost? Can the company live up to its goal of unlimited strips at a cash-pay price in the range of an insurance co-pay?
  • Switching costs and payer traction? It’s hard to get users to switch off their current meter – will One Drop’s product design and value proposition be compelling enough? Will the company secure relationships with big payers?
  • Burden of logging? Right now, the One Drop app, though unique in design, puts some burden on users to manually log glucose values, exercise, carbs, and insulin. The company has done an admirable of making this as easy as possible, and certainly, HealthKit integration and the connected meter help. However, One Drop’s Big Data goals are only as good as the information users actually log. Will the app sustain engagement over time? In practice, we’ve always gotten fatigued from diabetes logging apps – will One Drop be compelling enough for users to stick with?
  • Crowded app/meter marketplace? Approaches similar to One Drop (in one way or another) include Livongo, mySugr, Glooko, Accu-Chek Connect, OneTouch Verio Flex, and others. Can One Drop stand out?
  • Big Data hype vs. reality?Like many digital health companies, One Drop hopes to aggregate all the logged data and provide users with insights. Big Data has a lot of potential, but has yet to live up to the promise in these early days. We are hopeful, but can’t help but interpret One Drop’s Big Data enthusiasm cautiously.
    • One Drop founder Jeff Dachis has significant experience in building businesses on Big Data. Previously, Jeff was CEO/Founder of the Dachis Group, which built a proprietary social analytics platform to monitor over 10 billion social behaviors in near real-time. The company was acquired by Sprinklr in 2014 for an undisclosed sum. At the time, the combined companies had raised over $95 million in venture capital, acquired 11 companies, and served over 50% of the Fortune 500.
  • Expecting too much from great design alone? Patients are of course fans of better-designed products, but this is only a piece of the puzzle. Sanofi/AgaMatrix’s iBGStar is a perfect example – the concept and design was excellent, but the product did not succeed commercially. Some of that stemmed from Apple’s move to the lightning adaptor (iPhone 5), though as we understand it, Sanofi also ran into reimbursement hiccups. Great design is necessary but not sufficient for commercial success in diabetes.

MedTech Exhibit Hall

Yes Health

UCSF just announced a partnership with Yes Health, a new mobile app for diabetes prevention. We learned more at Yes Health’s sizeable exhibit hall booth this week – its differentiators include a consumer-facing business model ($39 per month for four months), a personal coach, an all-mobile offering, and a four-week free trial. The Yes Health mobile app enables tracking of meals and exercise, and serves as a platform for communication with personal health coaches who offer tailored recommendations on nutrition, fitness, and mental wellbeing. Yes Health is available for free on the Apple app store; after a four-week trial period, users can opt to enroll in the 16-week core program that includes personalized coaching and chatting 14 hours per day. The out-of-pocket cost is a fairly reasonable $39 per month ($155 for 16 weeks). After completing the core program, users can continue to use the app for maintenance for $15 per month. As part of its new partnership with UCSF, researchers will gain access to behavioral and biometric data, while Yes Health gets UCSF diabetes content, expert advice, and advertising. This partnership represents an early vote of confidence for the very young Yes Health, who soft-launched in October 2015 – the feedback so far has been positive from its private beta trial last year and early app store reviews since it’s general market roll out (a small sample of six reviews on the app store, all 5/5 stars).  Management told us average self-reported weight loss and increased activity thus far is close to the Diabetes Prevention Program target.

We are excited to see another player in diabetes prevention, particularly given the bio of founder Alex Petrov, who has built consumer brands at PayPal, Safeway, and Nestle. The big questions we have are: (i) can the app sustain engagement? (ii) can it prove outcomes? (iii) how large is the market for consumer self-pay diabetes prevention?; (iv) is Yes Health’s coaching scalable and is the business profitable at this price; (v) how much harder is generating awareness and scaling in a direct-to-consumer model, especially given the volume of available apps? Overall, we hope many technology-enabled diabetes prevention programs can be successful – 86 million Americans will require it!

  • As part of its new partnership with UCSF, Yes Health will create an “opt-in” function for users to share their de-identified behavioral data (diet, exercise, sleep patterns) with UCSF Diabetes Center researchers. The hope is that this data can provide insights regarding how behavioral and environmental factors impact prediabetes and inform future clinical trials. Further, UCSF plans to couple the behavioral data with biometric data from blood and DNA samples to create a deeper understanding of diabetes etiology, risk factors, and the impact of various interventions. As part of UCSF’s precision medicine program, the end goal is to develop individualized prevention and treatment programs for type 2 diabetes. In return, Yes Health will gain access to UCSF diabetes content and experts for improving its app, and will also benefit from advertising through UCSF media. This partnership represents an important early vote of confidence for Yes Health, a potentially critical move for a new company in the crowded app space.
  • Yes Health is a diabetes prevention program that provides personalized, real-time fitness and nutrition coaching via a mobile app. Each new Yes Health user is assigned a health coach, the major distinguishing feature of the program (we got Agatha). The health coach is available 14 hours per day and provides motivation and customized recommendations on nutrition and fitness activities; users can also upload meal and exercise photos to receive additional feedback. In addition, the app provides a number of suggestions for healthy meals, fitness activities, and well-being exercises such as meditation, deep breathing, and self-care. The app will also eventually integrate with health trackers and Apple HealthKit.
  • Yes Health is available for free on the iTunes store and app store; after a four-week trial period, users can opt to enroll in the 16-week core program for a total of $156 ($39 per month). The cost is fairly reasonable for something with one-to-one personalized coaching, and we wonder if it is scalable at this price.
  • While similar to Omada Health’s program in its overall goal to prevent type 2 diabetes, Yes Health is differentiating itself on two key fronts: (i) marketing and selling directly to consumers (vs. through the healthcare delivery and reimbursement system); and (ii) it is all-mobile with instant 1:1 coaching feedback, while Omada utilizes a web platform for detailed analytics and peer-to-peer support). Omada also has a mobile app, though we have not tried it and can’t speak to its functionality or usability. We like the idea of an all-mobile offering, which has convenience and passive data acquisition advantages (HealthKit, activity tracker integration). Also, harnessing mobile activities like picture taking, texting and calendaring that we all do every day for health may aid behavior change. The DTC model obviously has pros and cons: less hassle dealing with the healthcare system and sales cycles, though a larger ocean of offerings in the app store to overcome. The monthly $39 fee is fairly reasonable for a personal health coach, though it will alienate some who cannot afford it. The four-week free trial is a great option to get patients hooked and we hope that generates interest in the product. We wonder how consumers will perceive the value of Yes Health’s mobile app and whether they will be willing to pay for it; expectations are high for apps and the reluctance to pay even $0.99 is continually shocking. Ultimately, we think there is plenty of room for Yes Health to find its niche in tech-enabled diabetes prevention – the need has never been greater to reach millions of at-risk users.
  • Yes Health soft launched in October 2015, with positive early feedback. Its private beta trial last year and early app store reviews since it’s general market roll out have been positive (a small sample of six reviews on the app store, but all are 5/5 stars). Management told us self-reported average weight loss is close to DPP (Diabetes Prevention Program) target and average active minutes are at DPP target. Yes Health’s ongoing mobile marketing campaign are also showing high click through and conversion rates. We wonder how many people are using the app (our guess would be somewhere in the thousands), and how many continue the 16-week core program to completion. For context, Omada Health’s successful Prevent program has enrolled >40,000 participants since its founding a few years ago and boasts an impressive >80% completion rate. 

Close Concerns Questions

Q: Can the app sustain engagement?

Q: Can Yes Health prove meaningful health outcomes (reduced BMI, lower A1c, diabetes prevention)?

Q: Do the pros of a direct-to-consumer model outweigh the cons? How big is the market for self-pay diabetes prevention? Is the path to widespread awareness challenging for a consumer-facing health app vs. an insurer- or employer-sponsored program?

Q: What is the maximum ratio of patients to health coaches? Is this sustainable and profitable at the current price? Will Yes Health lower its subscription fee as it grows?

Q: Q: Will Yes Health eventually seek reimbursement through traditional payer and employer channels (like Omada Health)?

Q: How much upside is there from the UCSF partnership?

Plenary Sessions

We the People: Healthcare and the 2016 Election

Jane Adams (Senior Director, Federal Affairs, J&J, Washington DC)

J&J Senior Director of Federal Affairs – and, as we learned, type 1 patient (!) – Ms. Jane Adams gave a wide-ranging talk on the current healthcare climate in the context of the 2016 election that happened to touch on two majors topics relevant to diabetes: (i) the ridiculous bureaucratic red tape muddling the process of Medicare coverage of CGM; and (ii) the huge resistance among US payers and government to pay for any kind of chronic care management. This was among the most valuable talks we’ve ever heard on healthcare financial incentives, and it was fantastic to hear her honesty and perspective as someone who lives with type 1. Ms. Adams clearly appreciated that CGM can provide critical information for making therapy decisions (continuous data, alarms, trend information) and alluded to her personal experience with the “life-saving” technology. More poignantly, she even shared that her mother (of Medicare age) also has type 1 and has had to foot the bill herself. She did not dig into the weeds of Medicare’s primary objection to covering CGM (i.e., adjunctive labeling) but did stress that there is HUGE federal resistance to paying for delayed savings of any kind in healthcare – “Congress doesn’t invest in the long term. It invests in the right now … They are more than happy to pay for a kidney transplant but won’t pay to avoid it in the first place.” We’ve long seen that the challenge with government and payers alike is convincing them of cost long-term savings – this requires rigorous and costly trials to begin with and Ms. Adams stressed that overcoming this inertia and bias is going to prove very challenging. Ultimately, we salute Ms. Adams for driving awareness about this issue at SXSW and in Washington – she was one of the more charismatic speakers we hear on healthcare incentives and we include some of our favorite quotes below.

  • “Diabetes and obesity are bankrupting the system,” stressed Ms. Adams. She acknowledged the significant public health epidemic on our hands and the silent nature of the disease. Indeed, Ms. Adams got particularly animated during Q&A when an audience member stood up to call out the skyrocketing prevalence of prediabetes and diabetes. He lamented that no one is “waving flags and paying attention to this tsunami” despite the pending financial crisis. Ms. Adams had what we felt was a very simple and appropriate answer: “That is absolutely true!”
  • Despite the urgency, Ms. Adams stressed that addressing the diabetes epidemic costs money. The BIG question: Who is going to pay for it? She acknowledged there is a significant disconnect in government – everyone wants a solution but no one wants to pay for it. Part of this question gets into politics as well – what is the moral obligation of the federal government to help keep its people alive? Ms. Adams did not have an answer but stressed that the issue of financial incentives in healthcare is as much a political one as a practical one.
  • What might change the tide on capital hill? Patient advocacy. “Patient advocacy at the FDA is like nothing I have ever seen before,” said Ms. Adams. “Patients are not only storming Washington but are coordinating their efforts to impact regulatory agencies. This is real and this is happening.” Certainly, we’ve seen this in diabetes around the November 2014 FDA-Patient Dialogue on unmet needs in diabetes and continue to hope that efforts on CGM advocacy and other fronts come to fruition.
  • Ms. Adams expressed a great amount of frustration regarding the bureaucratic red tape muddling the process of Medicare coverage of CGM. As a reminder, bipartisan support of the Medicare CGM Access Act of 2015 has grown following the reintroduction in the Senate and House of Representatives in late March. Judging from Ms. Adam’s commentary though, it sounds like the 2016 election cycle could cause more delays and complicate this picture in a big way.
  • Ms. Adams was one of the more charismatic speakers we heard this weekend and we’ve included some of our favorite quotes below:
    • “Congress doesn’t invest in the long term. It invests in the right now. We’re talking about complications down the line and the fact of the matter is that Congress is very reluctant to pay for preventative care or anything that will keep us out of the emergency room ten years from now. We all know it doesn’t make sense but they want to see the cost savings is very important. I feel very strongly that we struggle with this every day.”
    • “It’s not that no one will pay for it. It’s that no one wants to pay for it. We need solutions in chronic care. Behavioral health management is a huge area right now and the more you can connect to digital health, the better the solutions. The reality is that diabetes has an emotional price, an intellectual price, and a financial price. I think the government is recognizing that they have a responsibility to help on this front but the question is how much and when. When I point out that Medicare doesn’t pay for CGM, it seems ridiculous. However, when you price it out and realize that it would cost the government $1 billion as a whole, then you begin to understand why they ask whether to put that money there or in Alzheimer’s.”
    • “How are we going to pay for preventative care? Who is going to pay for behavioral modification? We keep talking about surgical or medical interventions to treat disease. How about preventing in the first place? This is largely behavioral and it’s about education and impacting behavior which is difficult. I think there are many organizations who recognize this coming tsunami. However it costs money, takes time, and it’s not just here in the US. If you look at the global map of diabetes, it’s going to bankrupt countries across the world.”
    • “Patient advocacy at the FDA is like nothing I have ever seen before. Patients are not only storming Washington but are coordinating their efforts to impact regulatory agencies. This is real and this is happening.”

A New FDA: A Partner for the Digital Future

Bakul Patel (Associate Director for Digital Health, FDA, Silver Spring, MD)

FDA Associate Director of Digital Health Mr. Bakul Patel opened his presentation with a telling gesture, making a show of loosening and removing his tie to match the casual SXSW dress code – his message was clear and was met with raucous applause: “FDA should NOT be thought of as a barrier. We are your partner!” It was an appropriate introduction to a spectacular lecture on the Agency’s forward-thinking approach to regulating digital health technologies and its effort to adjust to the challenges of working in this rapidly changing space. The presentation itself did not share updated thinking beyond what Mr. Patel discussed at DTM 2015 – greater leniency on low-risk products to enable faster innovation – though his remarks underscored just how seriously the FDA is thinking about the evolving regulatory climate. Dr. Patel preached that the Agency is taking a risk-based approach to digital health that is intended to promote innovation, patient engagement, and safety. Balancing these goals, in his view, is about figuring out the amount of risk one can take to reap the ultimate benefits. He specifically highlighted the importance of designing products that fail safely, sharing a tolerance for imperfection that we could not have imagined coming from the FDA a few years ago: “Digital health doesn’t have to be perfect. Nothing in the world is defect-free. We need a system that works within those bounds. That’s the reality.” It’s terrific to hear this open mindset and it is clear that the FDA understands the need not to overregulate a quickly moving area with high potential. Dr. Patel did not talk specifics about diabetes, though we have seen this area moving in the right direction with rapid automated insulin delivery progress (MiniMed 670G pivotal complete + 80% of patients in the FDA continued access program), the rapid approvals of Dexcom’s G5 and MiniMed Connect, and several new cloud-connected meters coming to market (to name just a few).  

  • Mr. Patel focused on the FDA’s recent mobile medical app guidance that details the regulation of medical devices, apps, and accessories – namely that some platforms are of such low risk that they will no longer require regulatory oversight or will be regulated under a lower risk classification. Mr. Patel acknowledged that the Agency is still figuring out “exactly where to draw the line,” which is getting trickier as wellness devices become more advanced.
    • In Q&A, Mr. Scott Johnson alluded to this vulnerability in his own words: “I speak with the perspective of someone with diabetes. You might not hear this often but I genuinely feel the work you’re doing on a daily basis. You certainly don’t hear that enough and this open conversation speaks to how far you have come. I have to acknowledge how valuable this is. Thank you.” Indeed!
  • Mr. Patel encouraged those in the room to pursue frequent and extensive interaction with the FDA website. Indeed, he urged entrepreneurs to approach the Agency proactively to seek out advice and insight – the process has long been opaque and it’s a mutual effort (FDA sharing information AND industry asking for information it wants) that will move the conversation forward. Transparency is clearly at the forefront of Mr. Patel’s agenda, and we’re hearing more and more companies talk about early and often interaction (Dexcom definitely wrote the book on this, though we think many have learned from their example).

Designing Behavior Change at Scale

Chris Risdon (Adaptive Path, San Francisco, CA)

Mr. Chris Risdon delivered a fascinating lecture on how to shape behavior through design, creating apps and services that both fit into people’s lives AND prompt them to act. Through a number of stories, he covered examples of how leveraging psychology and empathy – “embracing the squishiness of life” – can influence behavior in significant ways. He pointed to the Red Cross’s short-term text capability for the Disaster Relief Funds during the 2010 Haiti earthquake as a prime example of a mobile system that capitalized on human impulse (feelings of sadness, sympathy) rather than rational long-term thinking (much harder!). The glucose sensing smart contact lens was mentioned as an example of working products seamlessly into lives though he did not comment on its development – still, it was great to hear accessible glucose data highlighted as an area of unmet need. Ultimately, this talk had many broader lessons for diabetes, whether it be for general product development (“every design decision influences the user”), reflecting on what works and what doesn’t (“most digital solutions require people to opt in rather than fitting seamlessly into their lives”), and the value of actionable insights (analytics that influence the immediate future vs. reflecting on the past).

  • “Every design decision influences the user,” said Mr. Risdon. He acknowledged that this concept is not novel – “we all know this” – but stressed that it rarely translates into action at the design stage. He argued that many manufacturing decisions are based on defaults and urged those in the audience to think outside the box.
  • “We now have more direct relationships with products and services. A relationship invites influence.” Traditionally, Mr. Risdon suggested that other individuals (family, friends, co-workers) had the largest influence on our decision-making. However, in the digital age, he noted that our physical proximity to devices provides a new avenue through which industry can impact consumers. As such, Mr. Risdon stressed that the value of digital solutions to consumers isn’t just in the data – it is in the fact that they believe the product brings them closer to a goal. Mr. Risdon concluded that the most successful digital solutions will leverage this relationship between a person and their technology to craft an interface that helps a person feel as though they are succeeding. This could explain why CGM is so challenging for many – it makes them feel like they are failing.
    • Granted, the above is particularly tough in diabetes where short-term behaviors are only distantly linked to long-term – and usually negative – outcomes (e.g., you shouldn’t eat a doughnut because of the consequences many years down the line). The challenge is moving away from this negative reinforcement paradigm where patients take action to avoid punishment. How can digital health solutions focus on rewards and immediate positive feedback? What short-term reinforcement can we employ that help patients feel that they are extracting value as opposed to simply ‘postponing the inevitable’?
  • “Collection -> Story -> Communication.” As we’ve heard from many in the diabetes space, Mr. Risdon stressed that collecting data is just the first step in behavior change. Putting that data in the context of a relevant story and communicating that context is what shapes behavior in practice.
  • “Math … WTF?!” Mr. Risdon expressed frustration with the fact that most digital devices collect data without adding consequence to it. “We tell people they are averaging 7,000 steps. But is that good or bad? We tell people their heart rate is 70 beats per minute. But is that good or bad? We’re giving people data but are we making it easy on them?”
  • “Framing makes a HUGE difference.” Mr. Risdon stressed that we can tell very different stories from data depending on how we choose to interpret it. He cited loss aversion phenomenon – when the same change in price framed differently (as a loss vs. as a gain) has a significant effect on consumer behavior – as one example of this tendency. The overwhelming message was that providing data in itself does not help individuals make sense of the world – rather, it is designers’ job to craft a story around data that is consistent with consumers’ goals.
  • Prospective insights > Retrospective Insights. “We need to influence the future immediately. We need to give users a sense of what will happen if I engage in a behavior, not confirmation of what just happened.” He raised the example of an app that identifies when an individual walks into a sandwich shop and informs them immediately of the consequences of various decisions (e.g., water vs. coke; cookie vs. fruit). Mr. Risdon suggested that influencing decisions in real-time takes advantage of human impulse as opposed to retrospective analyses that rely on patients making a mental note and recalling that information the next time they walk into the same shop.
    • We thought that the sandwich example could be applied to diabetes fairly readily. What if the same app told you ahead of time what eating a cookie might do to your glucose level? Or provided encouragement when you walked into a gym for what 30 minutes of exercise could do for CV risk?

MIT #HackMed

#DIYPS

Dana Lewis (#DIYPS, Seattle, WA)

Type 1 patient Ms. Dana Lewis discussed her experience spearheading the OpenAPS artificial pancreas project, sharing new news that 40 people (as of this weekend) are now using the open source closed-loop system. It was an informative session as Ms. Lewis detailed the desperate need that inspired this grassroots action: the ceaseless fear of hypoglycemia, the overwhelming cognitive burden of diabetes decision-making, and the bureaucracy that has slowed what should be relatively straightforward innovation. She was impressively persuasive, recounting the journey that has brought the OpenAPS project to where it is – from the insight that initially brought her to hack her CGM (“I wanted louder CGM alarms and industry was unwilling to accommodate my n=1 problem”) to the early iterations of the DIYPS algorithm that enabled remote monitoring to the final form that has now closed the loop. As a reminder, the OpenAPS hardware consists of a Medtronic insulin pump, Dexcom CGM receiver, a raspberry pi device for running Linux OS, a CareLink USB stick to enable communication to the pump, and a battery pack. Ms. Lewis did not share new updates per se though she was certainly a fantastic advocate for those with diabetes (on a public stage, no less) and for patient-led innovation. DIYPS has unquestionably been a part of the lead community that has pushed the FDA and industry to move faster, and the hassle factor of using this hacked together devices underscores one thing: people with diabetes are thirsty for devices that make their lives easier

  • “I don’t want companies to design a product for me. I want them design a base product that allows for customization that I can do it myself.” We loved this perspective on patient choice. Ms. Lewis stressed that designing a product for everyone’s individual needs is impossible – what is possible is giving patients the ability to make those determinations for themselves and address their individual pain points. (e.g., adjusting volume/tone of CGM alarms, who they want to share data with). This goes both ways, however – too much customization adds complexity for the average patient. We would argue for products that are so simple and so effective and so useful that no customization is needed, or they adapt automatically to someone’s physiology (automated insulin delivery).
  • “When we started hacking, industry was ten years behind us and wasn’t listening to us at all. They would say, ‘This is the product. We know best. We’re designing for the average patient.’ However, over the past few years – partly because of social media and pressure we’ve been able to bring – we’ve really started to have a conversation.”
  • Ms. Lewis stressed that the continued interaction between DIY hackers and industry will define the future of medical innovation: What kinds of collaborations are possible? How can industry work in a regulated environment? How can DIY Hackers work safely and leverage industry’s expertise and resources? Ms. Lewis shared that these important conversations are already happening, and it’s become increasing clear that industry is taking user innovation as an indicator for value that needs to be created.
  • Asked about her biggest learning from hacking her CGM, Ms. Lewis was quick to answer: immediate feedback. “Before this DIY work, I did not log anything … The reason I was willing to log this data is that I was gaining real-time updates, real-time valuable feedback. When I didn’t put in the data, I didn’t get as good insight. In terms of behavior change, you need that feedback. That’s what we need – not the kitchen sink. It’s hard to do something big at once but doing something small is reasonable.”

Admetesys

Jeff Valk (CEO, Admetsys, Boston, MA)

Admetsys’ Mr. Jeff Valk shared updated plans for the company’ s in-hospital closed-loop system: (i) CE Marking by mid-year 2016; (ii) a Scandinavian launch in late 2016; (iii) an FDA PMA submission in early 2o17; and (iv) a US launch in late 2017. Talk about ambitious goals! He laid out concrete go-to-market plans – no partner, initial launch in critical and perioperative care (particularly in cardiovascular surgery), and an initial cost of ~$250/day – sharing that the company is hoping to raise ~$12 million in the near future to fund the aforementioned goals. We learned that the company’s approach now titrates IV dextrose (previously, they reported glucose) and insulin in response to plasma glucose measurements at five-minute intervals. Work has been very stealth to date (we have yet to see clinical trials outside the research setting and this remains a major question) though the company has shown encouraging results in three small studies (n=43): 97% time in the tight range of 80-125 mg/dl with no hypoglycemia (<70 mg/dl). [Of course, IV glucose/insulin delivery makes this sort of control very achievable in controlled settings.] We have been following Admetsys since last year but there is still uncertainty on many fronts: regulatory hurdles, liability concerns, the challenge of convincing hospitals that the cost-benefit analysis comes out in their favor, the challenge of convincing providers to trust the system, and a poor track record for moving even CGM into the hospital. The work is unquestionably needed – we are not aware of any others trying to commercialize an in-hospital closed-loop system – though the road ahead seems equally steep.

  • Mr. Valk shared that the company’s closed-loop in-hospital system will run ~$250/day. Will Admetsys be able to convince hospitals that the cost-benefit analysis comes out in their favor? Management has previously suggested that a modeling study (Sadhu et al., Diabetes Care 2005) will support their claims, though we saw with Edwards’ abandoned in-hospital CGM with Dexcom that it was challenging to change clinical practice and convince buyers of the technology’s value.
  • Studies in the “real-world” hospital setting are still needed. Initial findings are promising but we are looking forward to results from the recently completed a 40-patient safety/efficacy study of the system with a fully featured prototype. Results have not yet been published.
  • Another big question in our mind is whether hospitals will trust this system. Mr. Valk suggested anecdotally that nurses are actively clamoring for this system and that workplace integration will be straightforward. Clinical inertia to new devices is always a question, and we wonder if some market research could support this point. As a reminder, part of the reason cited for Edwards discontinuing its in-hospital CGM program was the “effort needed to change clinical practice.” Clinical inertia does tend to be tough to reverse though we hope nurses and hospitals will jump on-board.
  • Mr. Valk did not refer to pivotal trials plans though we assume these are in the works or ongoing if an FDA submission is on the horizon. Considering how long it has taken the Agency to warm up to closed-loop work in the outpatient setting, we can only assume that Admetsys has been in close contact with regulatory officials.
  • Apparently, Admetsys has outpatient plans in the works. This was the first we had heard of this. Mr. Valk referred obscurely to “data in the outpatient setting” that the company is not yet ready to share. IV glucose/insulin delivery is obviously impractical in the outpatient setting, and we’re curious about exactly what he was referring to. Does the company have a subcutaneous system? The comments struck us as a bit off the cuff, and we wonder whether this is really consideration set considering the STIFF competition (to say the least).
  • For context, today’s timeline was a bit behind the guidance originally shared in mid-2015 for CE Marking in early 2016. We’re not quite sure what was responsible for the delay.

Digital Health and Outcomes: Where's the Evidence?

Panel Discussion

Christine Lemke (Co-founder/CPO Evidation Health), Malay Gandhi (Managing Partner Ensemble Labs), Mike Payne (Chief Commercial Officer, Omada Health), Stephanie Lee (Senior Technology Reporter, BuzzFeed)

Omada Health’s Mike Payne shared several gems on the company’s digital diabetes prevention program in this panel: (i) it has treated “about 50,000 patients” so far (up from “more than 40,000” in February); (ii) it expects five more publications in peer-reviewed journals this year (building on four published to date); and (iii) “it’s a possibility that Medicare will start reimbursing soon” for Prevent. The latter was amazing to hear, though “possibility” and “soon” of course do not provide specifics on when it could happen. As a reminder, CMS does pay physicians to watch and manage patients remotely, but it’s a measly $42 a month. Omada has always prioritized publishing outcomes, and we especially loved this comment from Mr. Payne: “In 5-10 years, we believe that the currency of selling stuff in this space is holding up peer-reviewed publications when you walk in the door.” Omada is investing “millions of dollars” in clinical research over the next 1-2 years and will continue to price its Prevent program based on outcomes (per pound lost or maintained). Mr. Payne said Omada is also doing continuous A/B testing to tweak reminders and subject lines (a huge advantage of digital), while new machine learning work is taking human bias out of hypothesis formulation – that sounds fascinating and we wonder what novel questions it might come up with. Omada has about 300 employees now, up from 185 employees as of September. More broadly, this session included a lot of relevant commentary for diabetes companies, particularly this quip from Rock Health’s former managing partner Malay Gandhi: “Bundled payments is where it’s going to be. Selling your widgets at high margins – that business is dead. It’s the service model now.” We hope device and drug companies are listening! More quotable quotes below.

  • “There are two major things to focus on early once you develop something in digital health. The first is engagement. The quick death for digital health startups is you can’t engage people. That means the product sucks and your company is dead. Once you get to engagement, then you talk about outcomes... If you have no data for your product, you better stand on the shoulders of someone who does (i.e., like Omada and the DPP).Malay Gandhi (Ensemble Labs)
  • “There is no reason why device companies cannot have a service model...Funding models have to change. There are the chronic care management codes, which CMS just lit up. Those pay physicians to watch and manage patients; it’s a small amount, $42 a month. That’s a lot of money CMS is making available to manage patients outside the office. Then there is DME reimbursement for devices that help with that monitoring. Can device companies monitor that? I think the answer is yes. As they scale to millions, it will takes clinicians and data science, learning algorithms, and maybe FDA approved decision support.” – Mike Payne (Omada)
  • “Digital health has a huge role to play for seniors; they are clearly responsive to these integrated solutions. The next-generation of telehealth is where it’s at: highly data-enabled, care-team wide, smarter devices linked in with telehealth. We’re starting to treat a lot of seniors with Medicare Advantage. We think it’s a possibility that Medicare will start reimbursing soon for what we do. Seniors face loneliness and isolation, and we think we’re helping with that problem. But someone out here should focus squarely on loneliness and isolation.” – Mike Payne (Omada)
  • “The average size of VC rounds is going to go up over time. You have to fund studies, and not just single studies. We’re investing millions of dollars in clinical research over the course of 1-2 years. We have to keep evolving the program to stay ahead. We have to be relevant to medical professionals and show sub-populations.” – Mike Payne (Omada)
  • “The most important capability in a digital health company if you want to have impact is design. Design in healthcare has sucked for a long time. It’s focused on the physician. For great user experience and behavior design you need world class people. The big trap is that everyone here thinks they do this. People who are really user experience designers are rare.” – Mike Payne (Omada)
  • “It’s a mistake to believe that digital health can solve all the problems in healthcare. We’re talking about basic issues like food scarcity and education; I would never sit up here and pretend that technology and apps are going to solve all of society’s problems. Some are more basic.” Malay Gandhi (Ensemble Labs)
  • “There is a tradeoff between legal and regulatory risk and business opportunity. You have to figure out where you fall. We have a healthy tension at Omada. Get a pharma person that is conservative in how you think about claims. Then go get consumer person. We have a healthy tension at Omada. Our chief brand officer was never in healthcare and wants to make claims that with FDA would never fly. We end up in a good spot.” – Mike Payne (Omada)
  • “One of the biggest risks to this sector is information security. Omada has invested millions in information security architecture. We have a 24/7 security information center with humans watching it. You have to raise VC money for information security. Or you can partner and use Amazon web services, which is HIPAA compliant. If Omada had a major data breach, it would cast a pall over the whole sector.” – Mike Payne (Omada)
  • “We took a page out of the book of pharma. You should have a KOL strategy and a scientific advisory board. If you do more trials, don’t put your name on the paper; it should be a high profile academic.” – Mike Payne (Omada)

Virtual Physicians: The Future of Healthcare

Panel Discussion

Andrew Thompson (CEO and President, Proteus Digital Health, Redwood City, CA), Jay Iorio (Director of Innovation, IEEE, Los Angeles, CA), Leslie Saxon, MD (Executive Director and Founder, USC Center for Body Computing, Los Angeles, CA), Todd Richmond, PhD (Director of Advanced Prototypes and Transition, USC Institute for Creative Technologies, Los Angeles, CA)

During a panel on the future of virtual physicians in healthcare, Dr. Leslie Saxon (USC Center for Body Computing, Los Angeles, CA) discussed the incredible value of “virtual Dr. Saxon” for personalizing chronic disease education, delivering empathy and motivation, and arming patients with useful scientific knowledge. Virtual Dr. Saxon is an animated “character” or avatar of Dr. Saxon that has been programmed to express sophisticated verbal and body language and provide a range of tailored content; patients can access the tool via mobile phone for follow up appointments, questions, and continual coaching. Dr. Saxon noted that for diabetes care specifically, virtual physicians provide personalized educational content to teach patients about glucose control, medication doses, and other relevant guidance. In addition, Dr. Saxon believes that all patients should have access to the same content that guides doctors in clinical decision-making. To that end, her USC team is adapting data from peer-reviewed journals into a straightforward format for use in virtual care appointments. We love this incredible mission to inform patients on the rationale behind treatment recommendations, and hope it could offer considerable motivation and empowerment. Dr. Saxon also shared another striking application of virtual reality (VR) in medicine – an immersive VR experience that provides a glimpse into the reality of living with diabetes-related vision loss (the technology was the winner of a recent USC/IEEE “Hackathon”). We see this as a powerful tool for inspiring glycemic control in teenagers with diabetes, as this age group often struggles with maintaining motivation to prevent complications lying decades in the future.

  • Throughout the panel, speakers highlighted other impressive applications of VR beyond chronic disease care delivery: Physician training, communication therapy for autism spectrum disorder, and treatment for PTSD and other mental health disorders. We are fascinated by the far-reaching implications of this technology in medicine, and hope to see more VR-enabled healthcare tools come to market. 

Virtual Health: Is it Real or Just Fantasy?

Panel Discussion

Alexander Vo, PhD (Executive Director, University of Texas Medical Branch, Galveston, Texas), Julie Hall-Barrow, PhD (VP Virtual Health and Innovation, Children’s Health, Dallas, TX), Mary Jean Strength, MD (Pediatrician, Pediatric Health Care Associates, Peabody, MA), Nora Belcher (Executive Director, Texas eHealth Alliance, Austin, TX)

A valuable panel addressed the “explosive” supply and demand of virtual healthcare services, arguing that the future success of the sector will rely on developing guidelines to ensure quality and continuity of care, as well as establishing credibility through involving providers and publishing strong clinical data. Multiple panelists attributed the rapid proliferation of telemedicine to shifting consumer expectations, underscoring the expansion of the millennial “on demand” mentality into the healthcare domain. Dr. Mary Jean Strength (Pediatric Health Care Associates, Peabody, MA) touted the potential of telemedicine to meaningfully improve the patient-provider relationship through reducing the logistical inconveniences associated with the traditional “brick and mortar” doctor’s office. To facilitate quality and continuity of care in the virtual health arena, Dr. Strength is in the process of establishing one of the country’s first clinical integration networks (CIN), a group of physicians collaborating to develop standards to improve care quality and limit hospitalizations. According to Dr. Strength, the goal is to unify practitioners beneath a single “clinical umbrella” so that care delivered via telemedicine will offer the same quality as care delivered in-clinic. Further, this standardization will ensure that patients do not receive conflicting treatment plans from different doctors; we think this will be critical as patients begin remotely accessing providers from a wide geography.

  • Ms. Nora Belcher (Texas eHealth Alliance, Austin, TX) provided a legislative perspective on the topic, noting that the success rate of virtual health legislation will rely on demonstrating clinical legitimacy. She asserted that there is a perception among regulators that telemedicine is more akin to “internet prescribing” than it is to true quality healthcare; combatting this will require the support of clinicians, as well as reputable journals publishing strong data to reassure people that telehealth models can deliver quality care. We whole-heartedly agree with this notion – it seems that virtual care is here to stay, but the presence of standards, clinician support, and reliable data will determine its niche in the wider landscape of healthcare delivery.

Imagining The Future Of Personalized Medicine

Panel Discussion

David O’Reilly (Chief Product Officer, Proteus Digital Health, Redwood City, CA), Ms. Jenny Hapgood (VP Product Management and Marketing, Chrono Therapeutics, Hayward, CA)

Mr. David O’Reilly and Ms. Jenny Hapgood aimed to get “beyond buzzwords” in talking about digital health, reflecting on the hurdles to patient-centric medicine. Their presentation began with a video detailing the pitfalls of too much health data, as current digital health solutions do a solid job of collecting data but the generation of actionable personalized insights remains a pipe dream. The panelists were concerned that the field is not moving quickly enough in this direction – they drew an analogy to self-driving cars (“it just feels inevitable”) but suggested that the same feeling of inevitability has not been captured in healthcare. Of course, that doesn’t mean there are zero companies thinking along those lines – they cited Omada Health and Twine Health (app-based health coaching system that can relay blood glucose data to caregivers; we have not heard of them) as examples of diabetes startups that do a great job avoiding information overload and creating digital tools that “fit nicely into patients lives.” Still, the overwhelming message was that our ability to harness digital health remains relatively rudimentary – indeed, we have a ways to go before achieving the “self-driving car equivalent” in digital health.

  • Dr. O’Reilly did not share any updates on Proteus’ work or pipeline but did provide an overview of the company’s effort to create a digital medicine. As a reminder, Proteus makes an FDA approved/CE-marked ingestible sensor that transmits medication adherence and biometric data (using stomach fluid for battery power) to a body worn patch and Bluetooth enabled devices (video here). We last heard from the company at CES 2016, where Head of Product Management Ray Lee attracted our attention with a brief reference to structured glucose testing. We’ve been following Proteus for more than five years and continue to believe that addressing diabetes or obesity would be a major win for the adherence tracking company. Current applications are for heart failure and hypertension.
  • Ms. Jenny Hapgood provided an introduction to Chrono Therapeutics’ smoking cessation solution, which delivers nicotine replacement therapy though the skin via a wearable pod that can be worn anywhere on the body. The injection system does not use needles with the idea of discreetly delivering the necessary medication to help smokers quit. It’s a neat solution and while applications to diabetes were not mentioned, we can certainly imagine the potential to deliver insulin through the device. Prometheon is one company we are aware of working on an entirely needless insulin delivery patch – see our coverage from May 2015 for more.

Inner Space: Bioelectronics and Medicine's Future

Panel Discussion

Kate Rosenbluth, PhD (CEO, Cala Health, Burlingame, CA), Walter Voit, PhD (Assistant Professor of Mechanical Engineering, University of Texas at Dallas, TX) Sree Sreenivasan (Chief Digital Officer, Metropolitan Museum of Art, New York, NY), Monsef Slaoui, PhD (Chairman of Vaccines, GSK, Brentford, UK)

This fascinating panel discussed how “bioelectronics” – miniaturized implantable devices that connect to neurons and control their electrical signaling patterns – could define the future of chronic disease treatment. Panelists in the session –Drs. Walter Voit (UT Dallas), Kate Rosenbluth (CEO, Cala Health), and Moncef Slaoui (Chairman of Vaccines, GSK) – mentioned the application of bioelectronics to diabetes (!) in the course of introducing us to this virgin field of research. The basic concept is surprisingly simple. Functions and organ systems of our body are, to a significant extent, controlled by electrical signals that travel along nerves; a “bioelectronic” device would thus treat disease by sitting on nerves before they enter organs, deciphering the electrical signal, and correcting inappropriate patterns. It is not entirely clear to us how this might be utilized in diabetes treatment (i.e., how good electrically controlling the pancreas impact diabetes?) though multiple speakers did allude to the promise in “diabetes and obesity.” Hmmm … certainly, the work is VERY early stage and the road to managing the computational complexity in itself – let alone the regulatory, pricing, safety, and efficacy questions – figures to be long and arduous. At the same time, hope was far and away the central theme of the session (see below for some of our favorite quotes) and it was striking for us to see just how much enthusiasm exists for this early arena [and as we learned, much of this enthusiasm is actually grounded in a relatively firm research foundation.] Panelists offered the optimistic prediction that clinical technology that interfaces with nerve fibers is only on the order of “years away, not decades!” A bioelectronics system would certainly be a spectacular advance for people with chronic disease, and we hope a product like this comes to fruition some day. The application to diabetes remains a big question mark (i.e., would this be an implantable, miniaturized artificial pancreas?) though the transformative science certainly allows us to dream …

  • The fascinating discussion yielded a number of quotable quotes. We bring you our favorites below:
    • “With this technology, we can treat type 2 diabetes. We can treat obesity. We can accelerate the healing of bone fractures. We have identified a number of devices that could play this role. There are a huge number of hurdles but it’s exciting.” – Dr. Moncef Slaoui
    • “The vision here is medicine without pills, when and where you need it. It’s about both temporal and spatial sensitivity. It’s coming.” – Ms. Kate Rosenbluth
    • “You have to be optimistic and you have to be resilient to discover anything. I began working on the malaria vaccine on my first day at GSK 28 years ago. It was approved in July 2015, 27.5 years later. It wasn’t because we were lazy. It was because it was complex. You have to consider the timeframe of things in medicine. We are interfacing with human biology of which we understand 0.1%. But my message here is that this can work.” – Dr. Moncef Slaoui
    • “The ability to apply neuroscience and computer science to medicine requires unprecedented integration – integrating aerospace research with engineering with data science with wearables.” – Ms. Kate Rosenbluth
    • “If some fundamental human function is going to be controlled by a device, it MUST be fail-safe. 99.99% performance is not good enough.” – Dr. Moncef Slaoui
    • “I think in this field – belief, faith, vision, and resilience – are very important. Hang in there. You’ll get your own bioelectronics implant soon.” – Dr. Moncef Slaoui

New Cancer Tech: Is Easy Detection A Reality?

Panel Discussion

Elaine Schattner, Kevin Conroy , Luke Timmerman, Sam Hanash MD PhD

This panel discussion on cancer screening provided what we felt were three really outstanding insights into parallel challenges in diabetology and oncology. We were struck by the frustration on multiple levels (regulatory, cost, fragmented ecosystem) that echoed many themes we hear in diabetes. Remarks underscored immense dissatisfaction and reminded us that the exasperation we experience in diabetes (e.g., “Why aren’t things moving faster?!?”) is not unique to the field but is characteristic of healthcare broadly – indeed, is anyone in healthcare satisfied with the status quo? We have reproduced below our three major takeaways:

  • “We spend billion of dollars on cancer drugs that can extend lives a few weeks but not on diagnosis that can totally change the game.” Panelists expressed significant frustration with the money spent on late-stage care that could be spent instead on diagnosis. The parallel to diabetes prevention struck us quite strongly here. With the advent of genomic data, panelists stressed that we have enormous predictive potential but lamented our focus on “disease” care rather than “health” care.
  • “One drop of blood gives enough data to fill a laptop! Our challenge is sorting through that.” Dr. Sam Hanash stressed that the oncology field has figured out how to gather data though the challenge remains finding ways to use algorithms to generate actionable recommendations. Simply collecting the data is only the first step; no one has the time or desire to sift through data and make sense of it. His hope is that the future will see better use of algorithms to recognize patterns and make diagnoses.
    • Harnessing big data remains a challenge in diabetes, too, and we have long said that it is time to move past the optimism and get to the pragmatism and really dig into the important unanswered questions. How do we convince patients to pay? How do make behavior change something people want to do? How do we show payers value? How do we consolidate our disparate data? To an extent though, we found it reassuring to realize these questions have yet to be solved in oncology as well … these are TOUGH questions.
  • “One thing that payers are sure of is that new technologies cost them money.” Panelists alluded to huge barriers to insurance coverage in some areas of oncology screening (e.g., colon cancer) despite the life-saving success that has been seen in other areas (e.g., prostate cancer). What needs to change? Panelists argued a need to show improvements in clinical outcomes and tying those outcomes to economic benefits, expressing HUGE frustration over the cost of running the long-term trials that are needed to convince payers on this front. Sounds familiar

Rethinking Healthcare Through Design Thinking

Panel Discussion

Stacey Chang (Executive Director, Design Institute for Health at the Dell Medical School, Austin, TX), Betsy Fields (Senior Lead, IDEO, Palo Alto, CA)

Design thinker Mr. Stacey Chang argued that changing the healthcare system will require new and creative training for young doctors, plus an environment that values such training. He noted that Austin is perfectly poised to support this transformation, as its brand new University of Texas Dell Medical School (enrolling its first class this fall) and county hospital (under development) will provide a rare opportunity to build an integrated delivery system from the ground up. Dr. Chang noted that the new medical school is unique in that it does not carry “legacy or historical baggage”; thus, it is free to design an innovative curriculum and business model around the needs and opportunities of 21st-century medicine. Some of the fundamental shifts Mr. Chang hopes to see in Austin include: moving from acute care to preventive care; focusing on communities instead of industry; serving low-income populations; and eliminating outdated perverse incentives. Talk about implications for diabetes and obesity! Mr. Chang’s team is working to design a “naturally malleable” healthcare system in Austin, even down to movable walls in new facilities that will support learning and evolution over time – this is also a hallmark of the Design School at Stanford, which has many such flexible work spaces. While Mr. Chang expressed optimism for this model’s potential, he acknowledged that design thinking in healthcare is incredibly difficult: the system fiercely resists disruption. Stakes are high in healthcare, and failure often leads to huge losses of time, money, and even human lives. As a result, the system rarely invents from scratch and instead focuses on low-risk, incremental modifications. Co-panelist Ms. Betsey Fields said the key to cracking this rigid “sick care” system lies in redefining society’s mental model of health and healthcare delivery. Ms. Fields advocated for moving beyond traditional healthcare institutions to design for health where “life is happening”: at the grocery store, at home, during commutes, and at work and school. We wonder what grocery store diabetes care and prevention would look like, or how employers could better help people with obesity. We are always thrilled to see designers bringing creativity to healthcare and wonder what care delivery models could make a major difference in diabetes and obesity...

 

 

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