Smarter diabetes management
Browse Blog Topics

Smarter diabetes management

Welcome back to What’s NEXT, the podcast exploring the future of technology. In this episode, I talk with Jeff Chang and Vikram Singh about how diabetes management platform Glooko is helping improve health outcomes for people worldwide.

You can listen to the full episode in the player below, or subscribe through Apple PodcastsGoogle PlayRSS, or your favorite podcast app of choice. We’ll be taking a quick break after this episode, but will be back in 2019.

Ryan Lawler: Welcome to What’s NEXT.

Vikram Singh: Thank you.

Jeff Chang: Thanks, good to be here.

Ryan Lawler: To start, tell us what Glooko is and what you do.

Jeff Chang: At the heart of what we do Glooko is a diabetes management platform. People with diabetes generally have to do several things. They have to watch what they eat. They may or may not have to take medication and generally they’re in some sort of communication with a doctor. Through the Glooko platform, which consists of mobile app and a web platform, we want to enable somebody with diabetes who may be tracking this variety of different activities to be able to track that on a digital platform. Before Glooko, it’s important to note that a lot of diabetes tracking used to be you would go to the doctor, your doctor would give you a piece of paper and you’d log down everything that you’ve done over a week, including how much you’re eating, how much medication you’re taking and importantly something called your blood glucose which people with diabetes generally have to measure.

Through putting that on a mobile app, our hypothesis has been that people can very easily interface with something that’s in their pocket and they carry with them all the time. We’ve also built a variety of methods to download data from the diabetes technology that you might be using and so by technology I’m referring to things like your fitness tracker. Generic things like your fitness tracker all the way down to specific diabetes technology like your blood glucose meter and for advanced users, an insulin pump or a continuous glucose monitor. Once we get all that data into Glooko, you can then seamlessly share that data with your doctor and that’s sort of our main value proposition. It’s that you and your doctor can have this way more educated doctor’s visit where you’re looking at actual data that makes sense rather than somebody’s poorly written handwriting on a piece of paper. You can have that conversation in the span of couple weeks rather than typically you’ll see your doctor every three months.

The idea is that you can much more easily get data into a platform and share that data and review that data and ultimately get to better outcomes through that whole process.

Ryan Lawler: And just to give a sense of the problem, when you talk about self-reported data, it seems to me that it’s probably not very reliable. Do you have data around that that you can share?

Vikram Singh: It’s been a common theme within the industry and well known from clinicians that if you’re going to take a log book which is what they call the journal that people with diabetes are recording their blood sugar data in, and use that log book as your source of truth, then more than likely there’s going to be several inaccuracies. The approach that Glooko has taken from the start is we want to remove those inaccuracies by integrating directly with the blood glucose meters and other diabetes devices and be able to actually allow for blood glucose data to be manually input to solve that problem for the clinician and for the patient.

Jeff Chang: I want to add something here. Disclose that I actually have type I diabetes myself and the very interesting things about what Vikram is talking about is people who, there’s a common story among people with diabetes or their endocrinologist which are kind of the diabetes specialists, people like to fudge their numbers when it comes to having doctors’ appointments. I used to do this when I was younger and there’s all sorts of ways you can do it. For instance, you can take a drop of blood, you could add some water to it and it gets lower. The reading actually gets lower. There’s a lot of this in the name of sort of trying not to get judged by your doctors. Diabetes is a very number based disease and so you check your blood glucose, you bring it in the clinic, your doctor tells you, “Are you doing that? Are you doing good?” There’s this three-month metric called an HBA1C that tells you are you doing bad? Are you doing good?

A lot of people will want to make up numbers around their diabetes. Especially if you’re a teenager or you’re a child and your parents are keeping you accountable, I think there is a motivation to do that. As Vikram said, one of our philosophies has been to only take in trusted diabetes devices data from your devices.

Ryan Lawler: Gotcha. Maybe we can talk a little bit about the scope of the problem. How many people have diabetes. How many people are at risk of it. How are those trends changing over time?

Jeff Chang: It’s a great question. Globally, over 400 million people have diabetes. Of that, about 260 million are diagnosed with diabetes and to further break that down, there’s type I and there’s type II diabetes. Type I is known to be the more genetically linked one that’s an autoimmune disease. That’s one you get, you’re sort of born with it. It’s called Juvenile diabetes ’cause you’re often diagnosed with it at a young age. And then there’s type II diabetes and type II diabetes tends to be the more obesity linked, genetically linked disease where you might see lifestyle factor influencing diabetes.

If you look at type II prevalence and specifically in the US, about 30 million have diabetes. To put that in perspective, 30 million is about 10% of the US population so one in 10.

Ryan Lawler: We talked a little bit about societal changes or trends in diabetes but how’s the technology changed? How has monitoring changed over time?

Vikram Singh: The technology, I think in the 1970s when blood glucose meters were invented and for long time I think through the early 2000s, that was the sole technology for measuring your diabetes along with insulin injections and the way that that data was used to for a clinician to actually manage a patient’s diabetes and for a patient to manage their diabetes, was through the log book. People going to be expected to write down on a daily basis what their blood sugar levels were. What are they eating. What kind of exercise have they been doing and then every three months if they make it to the doctor, that log book was going to be reviewed and within a 15 minute time span, there was an expectation that a clinician will be able to process and distill that data and make some educated therapeutic recommendation to the patient.

We all know that it’s really hard for anyone to record anything for three months straight. That remains the same for diabetes. A clinician to be able to actually process all that data was really just a Herculean task that today is still being expected of many clinicians. What has happened in the last several years and a change that Glooko has been driving is creating software that can enable the clinician and the patient to manage that data.

Up until, I’d say 2010, essentially what the market looked like was that diabetes device companies had software that would allow the data to be downloaded within a clinic and so there was many disparate technologies being used by a clinician or by a medical assistant within a clinic. The problem that created was that though there is software to get the data from the devices and allow the clinician and patient to have a conversation, those softwares were hard to use. There was abundant amount of cables used within the clinic. The medical assistants didn’t know how to use the software and so the result was that even though there was a way to access the data, they weren’t doing it and so the standard of care remained the same which is, you have a diabetes device, you have a log book but no one’s actually making data driven decisions.

Over the last eight years the movement with cloud-based technology has been, let’s move to, what Glooko has done is say, let’s standardize the diabetes device market and have a unified data model. Let’s make the data universally accessible to many constituents within the market, the patient, the clinician and diabetes device companies. And let’s make sure there’s a standardized view that everyone can look at at once. Essentially what’s happened in the last couple years is now when a patient comes into the clinic, they know that they’re going to get a standardized view of the data from the clinician. The clinician knows they’re going to be able to look at the same view of the data each time and so these conversations about how you give therapy recommendations and become much more consistent and easier to have.

Ryan Lawler: Right. From a product standpoint focused on aggregating all of that data from dozens of devices or hundreds of devices, getting a standardized view et cetera, how do you get the device manufacturers to play ball? Since to your point earlier, they probably each had their own type cord, they probably each had their own software. They each had their own way of interfacing with whatever software there is. How do you solve that probably?

Vikram Singh: I think this has been an interesting dynamic over time. In the early days the industry did not necessarily see the value in data in itself. When Glooko went and approached some companies and said, “Hey, we want to integrate with your diabetes devices.” There wasn’t a lot of pushback there. They said, okay, here’s a small company that is trying to create some cool tech, go for it. And they let us sign data licensing agreements to do that.

Our primary philosophy that we’re always going to have a data licensing agreement with a partner to get that data. We’re not going to take any sort of back alleyways of hacking into a device and getting the data. Over time, we basically started to capture the market by getting clinicians to say, “Hey, we want to use Glooko in the clinic and if your device is not on Glooko’s platform then we’re not able to use it as well and thus we’re not going to actually prescribe it to our patients.”

Though over time the device companies have now said, “Okay, we realize the value in the data, why should we give it to you Glooko?” We’ve actually been able to establish leverage and say, “Hey, clinicians all across the country and patients all the country now want our software and to play ball in the industry you need to be able to use Glooko.” And thus, we’ve been able to create some leverage in the space and get diabetes device companies to come to us and work with our platform.

Ryan Lawler: Got it. Is your primary way of getting patients on the platform through the clinicians and the care providers?

Jeff Chang: I think that’s the most natural place for a patient to get information about Glooko. We have a direct to patient offering but I think a big problem with health and health tech is that you’re going to get your 1% or 5% of people that are really motivated to download this app and to engage with it. The other 95% you have trouble getting them through the door and to even look at using an application.

We have a lot of in-clinic strategies where we want the clinician to be that starting point for a patient to know what Glooko is. And naturally, as Vikram eluded to, a lot of patients go to the clinic and they interface with Glooko maybe not even knowing about it. What happens often is they’ll bring in their diabetes devices, so their glucose meters or insulin pumps, they’ll hand that to a nurse or a staff member there and the staff member will use Glooko to download that patient data. And then they’ll generate a PDF report of the patient’s data for the doctor and the doctor will look at that and review with the patient.

We want that doctor to give the patient a handout and say, “Hey, your data’s available in Glooko. You can go home and create a Glooko account today and look at the exact same thing at home.

Vikram Singh: I think just to add, it varies country to country based on the incentives for the clinician and the patient. Glooko has a wide footprint in Europe and in Europe, clinicians are incentivized to manage the patient regardless of whether the patient is in the clinic or not. They’re able to say, when the patient comes in, “Please use Glooko software and that will facilitate our conversation when you’re not here.” And there’s economic incentives around having those conversations.

The US has been primarily fee for service forever and there’s been some slow, maybe steady changes towards risk-based and outcomes-based care. It has been much more of a challenge in the US over the last several years to say, “Clinician, please onboard a patient onto Glooko,” because the economic incentives aren’t there. But even in the last year we’ve seen new CPT codes as they say, that are going to reimburse clinicians for doing remote monitoring and so we’ve built infrastructure to get patients to be onboarded by clinicians and we expect that all digital health solutions including Glooko are going to see a lot more onboarding in the US over the next couple years.

Ryan Lawler: Okay, well that actually brings up an interesting point which is, who pays? Who’s the customer? And how do you get compensated for actually providing this platform?

Vikram Singh: Glooko lives in a multi-sided marketplace where there are many economic buyers of diabetes data and services and providers. That includes the clinician, the person with diabetes, the health system, the pharmaceutical company, the insurer, the diabetes device market and so for Glooko the challenge is, how do we deliver the best care and the best services to a clinician and a patient to have to deal with diabetes in some capacity on a daily basis? And how do we get the rest of the market to subsidize and fund those services?

There’s a variety of different business models that we have explored with varying amounts of success to get the software into clinicians and patients’ hands without charging them. Ultimately the value proposition is that we’re going to be able to allow for patients to experience better outcomes and what that should mean for the market is there is money to be saved by the insurance companies. Pharma companies are going to be able to prove out real-world outcomes and thus they will have better sales and better reputation in the market and so with that savings, some percentage of that money is going to come to Glooko and there will be economic benefits to other people within the market.

Ryan Lawler: What kind of data are you collecting? And how are you using that to drive better patient outcomes?

Jeff Chang: When you think about diabetes technology and diabetes devices, one of our goals as we’ve talked about, is creating a platform where we can get as much data as possible and really standardize that data and to be able to layer on different types of analysis, different types of decision support onto that data to create value for both the patient and for the clinician. That’s something as simple as telling somebody what time they’re in range. Or telling them, looks like Wednesday is your best day. What do you do differently on Wednesday from the other days of the week that you could be learning from? All the way up to creating a pretty complicated decision support tool we call MIDS or stands for mobile insulin dosing system which is meant to be an insulin titration algorithm.

In simple words, basically what we’re doing is we’re trying to get a patient to the right doses of insulin for themselves. We can automate a reminder that the clinician can set for you and it reminds you every day to check your blood glucose. It reminds you to take your insulin and it automatically captures that information in the Glooko app so it’s very minimal investment from you in terms of how much you have to engage with your diabetes.

Ryan Lawler: Okay. You talked about the patient side of things. What do the clinicians get out of this? How does it help them better serve their patients?

Vikram Singh: I’ll start with the in-clinic use case. A patient comes into the clinic and as I eluded to earlier, previously there was several different softwares that all present similar kinds of statistics. What was your average blood glucose over a time period? How many low blood glucose sugars did you have? How many highs did you have? But the issue was, one it was hard to actually get those reports printed and if you did get them printed, you’re looking at six, eight, 10 different views of the data which makes having that conversation hard because there’s just switching time and learning the different software.

From a purely workflow standpoint, Glooko has standardized that view so that rather than spending five minutes of the 15-minute appointment trying to get the data and the report and then interpret it, you’re spending 30 seconds to a minute doing that and you have four extra minutes to have the appointment.

It’s just added a lot of efficiency to the clinician’s workflow when they’re managing the patient and then of course as a tech company, we’ve taken a lot of algorithm base approaches to then add additional value and go beyond descriptive analytics and say, “What can we tell the clinician about the patient’s trends and patterns that you’re not going to be able to see with a simple average?” That would be patterns like what times of the day is the patient struggling? Is there any relation to exercise? Or their insulin regimen that is causing their blood sugar to go up or down? Elucidating those insights for the clinician so they don’t have to find that themselves. All these innovations are helping facilitate more effective appointments.

On the other side of the workflow is the coach that’s actually trying to have the remote patient appointment. This workflow has existed in some capacity in the US. As I said, fee for service didn’t really allow, or doesn’t really support remote coaching but in Europe it’s existed for some time. But in any case, the clinician is going to be calling the patient and saying, “Tell me about how your diabetes is going. Tell me about how your blood sugar looks.” And the patient can really say anything at that point. Is that conversation actually effective? And I think our hypothesis was, no it could be a lot more effective.

With our mobile solution and our web solution, a patient can use Glooko at home, download their diabetes data, input relevant data points and then the clinician is going to have a view of the data that they would have if the patient had come in. That conversation is all of sudden a data driven conversation that’s effective and we’ve added some population management tools so that the clinician isn’t just calling any patient arbitrarily, they can actually rank their patients by risk and say, “Okay these are the patients that I think have the most challenges right now. Let me go ahead and call them and make sure that they are getting the help that they need.”

Ryan Lawler: Okay. And just curious, how is that different from other solutions on the market? What’s unique about this?

Jeff Chang: We have an integrated system with a web platform and a mobile app and as we touched on earlier, the ability to stratify patients based on risk. We really want to enable patients to get the best care as possible by having a coach in real time or having a care team in real time be able to see exactly what the patient is doing.

On top of that we want to enable patients to have the best mobile app experience. By interfacing with the mobile app, patients are able to track what they’re doing as seamlessly as possible and share that data with the doctor. And really it’s through that interaction between patient and clinician that we want to drive outcomes.

Ryan Lawler: Got it. A lot of what we’ve been focused on is managing care patient to patient on an individual basis based on the data that you collect from them but are there things that you can do with the aggregate data with the entire community that can be used in a research capacity to drive better outcomes for the entire community?

Vikram Singh: Yes, definitely. Within the healthcare industry, I think there’s been a lot of buzz about real-world outcomes versus study based outcomes that are measuring the efficiency and the efficacy of healthcare interventions. Whether that’s a diabetes drug or a diabetes device or other kinds of programs. Every time Glooko deploys our software, we are getting a population’s worth of data that are using all sorts of healthcare interventions. Whether it’s a drug or a device or a program and we’re able to then say, “Okay, which diabetes devices, which drugs or which programs are actually helping the most? Or which kinds of people benefit most from different kinds of interventions?” And we’re able to then deliver that value to different constituents in the market and say, “Hey, here’s what you should be doing more of or doing less of.” And then allowing clinicians, pharmaceutical companies, device companies to make better decision in terms of how they design their products or services.

Ryan Lawler: What’s one controversial opinion that you have that’s really strongly held?

Jeff Chang: I think humans in general and bear with me, this probably isn’t that controversial but humans in general are really bad at embracing and dealing with very slow change. Examples are climate change for instance. Or in the case of diabetes, it’s for somebody that’s had diabetes, you really don’t see the impacts of it until 30, 40 years later because on a day to day basis you’re busy A with normal life and the complications don’t manifest themselves until further down the line.

In this particular example, I wanted to call out that I think we as a nation are not embracing the right values when it comes to primary care or to treating people. We’ve talked a lot about how we are a fee for service in the US specifically. What that means is we incent doctors to perform services on people. Once you’re diagnosed with a disease, you get a battery of blood test. You have to take medication or you do different kinds of examinations. But it really starts with that diagnosis.

In a lot of other countries, and we are experimenting more with this in the US, but not to the extent that we should, we should be focusing a lot more on preventative care. That’s everything from helping somebody not get obese earlier. Teaching somebody what the correct lifestyle should be to avoid getting obese. That’s for diabetes, that’s really, really focusing on preventing the onset of diabetes or slowing the onset of diabetes ’cause it costs the US alone $20 billion just in pharmaceutical inventions last year.

Ryan Lawler: If you weren’t working on this problem, what other areas of tech would you be interested in? What would you be interesting in pursuing?

Vikram Singh: I’m particularly interested in environmental issues and conservation. For me the oceans is a particular a passion of mine. There is a company actually that I’ve been really inspired by the last year called Saildrone that is based in Alameda. What they do is they have these autonomous drones that sail around the oceans and are able to track migratory patterns of fish populations, whale populations and also track other metrics about the ocean and are able to create value in the data set that they can then help fisheries make more important fishing decisions, help the governments make more informed ocean conservation decisions and I think it’s a really interesting combination of how technology can help conservation and also have a big economic impact and intentivize the corporations that right now are exploiting the ocean.

Saildrone and oceans tech are pretty interesting to me.

Jeff Chang: In diabetes tech specifically, there is a very small but dedicated group of people, a couple of groups of people. One is called Nightscout and one is called Loop. These are basically kind of techies, hackers who have technically hacked into some of the pumps that are on the market to put in what’s called an artificial pancreas algorithm. That’s basically they’ve built the ability for this pump to be able to deliver you insulin automatically based on what your glucose levels are. Very recently, a company called Medtronic put out a pump called the 670G that does this very thing but prior to that these communities where, and are still are developing very advanced algorithms that are able to do the same thing that these companies are doing, the only difference is, they don’t have to go through the same type of FDA scrutiny. It’s a group of people developing code, committing nightly builds for things that actually deliver insulin into your body.

There’s thousands of people in this community. It’s growing strong on Facebook. I think this kind of tech is really important because it pushed diabetes companies to see what people are valuing and to really embrace this change. These people on the forefront of diabetes technology are driving this wave of innovation and we’re slowing seeing device companies adapt to it. But it takes a long time when you have to deal with the FDA. It’s much easier when you’re a group of kind of people just doing it on your own.

Ryan Lawler: Actually that brings up an interesting question which is, when you operate in a very highly regulated environment, how do you deal with that from a product perspective in being able to innovate but knowing that you have to go through FDA clearance? You have to deal with regulations like HIPAA that you have to deal with all these other things that in general most startups might not have to.

Vikram Singh: First and foremost, I think that absolutely it adds overhead for any company trying innovate in the healthcare space. From a mentality perspective the company needs to commit to being enthusiastic about working with the FDA. If every step of the way you’re saying, “Why do we have to do this? This is making it take two weeks longer, four weeks longer, months longer,” it’s just going to diminish morale across the company. It needs to be an upfront commitment.

Our perspective through our years working with the FDA is if we show that we’re proactive with working with them, then they give us a lot of proactive guidance about what to do and how to do it. Rather than us submitting a product and then them rejecting it outright they say, “Okay, submit some MVPs of the product. Tell us about what your plans are and we’re going to help you design around those plans so that when you do actually make your submissions, there’s a much higher likelihood of approval.” I think this proactive mentality really helps with working in the space.

Jeff Chang: Yeah, one of the mantras of tech is to innovate fast and get betas out, see how people react to it. Some of the things we have to move a little bit slower on and that’s just the nature of working with the FDA. We have to make sure products are fully tested and vetted before they go out to the field. It’s something we kind of learned to embrace. Like Vikram said, if the company culture was one where it’s like, oh the FDA’s around again or we have to do this again, it wouldn’t be healthy. But I think we kind of keep in mind that we’re doing this really for the safety and protection of people and to ensure that our product is always going to be of the highest quality.

Ryan Lawler: Yeah, so how will things be different if Glooko becomes ubiquitous or widely adopted within this community?

Vikram Singh: I think we’ve eluded to this throughout the conversation, there’s going to be two major benefits. One is that patients, clinicians are going to have an easier time managing diabetes. Patients are going to have better short term and long term outcomes and clinicians are going to have a much better and a more effective experience managing the day to day of their patients’ diabetes. The economics will pan out for everybody. Pharmaceutical companies will be able to prove out real-world outcomes and thus improve their bottom line. Insurers will be able to prevent adverse events for patients thus prevent hospitalizations which is a big cost burden. And eventually as a country we’ll have less diabetes and a healthier population. Meanwhile, Glooko will have the data set to facilitate research to actually a cure for diabetes and to facilitate the autonomous closed-loop algorithms that will help people have to do nothing when they’re managing their diabetes.

Jeff Chang: For me, it means that we’ve moved away from a place where we question the value of data or where we don’t have data in our day to day diabetes management to a place where we as a society have embraced the value of being able to see your own diabetes data when you want to see it and how you want to see it. And that means we’ve moved away from the conversation of convincing people how much it will benefit them to look at their data to one where it’s the norm. You look at your data, you see what we can improve on with your data. It means that we’ve moved progressively as a society to one where we know that this should be the standard for diabetes management and let’s talk about the next thing that we could do with diabetes management.

Ryan Lawler: Well Jeff, Vikram, thanks for being here with the podcast.

Vikram Singh: Thank you.

Jeff Chang: Thank you. Good to be here.

Related Stories