How to apply AI in digital Mental Health products. A framework outlining impact, risk level and time.

“AI could just replace therapists”

This statement is abundant on the internet right now. On podcasts, x posts (née tweets) and articles people are dreaming of a world where high quality therapy is available to everyone at a fraction of todays cost. While that future may become a reality I believe there to be applications for AI in digital mental health that you’ll encounter faster and that will generate a ton of value for patients.

When planning investments in AI across Big Health products I find it helpful to break them down in 3 different categories. These categories have different timelines, risk profiles and impact.

1. Generative AI tooling for personalization
2. New interaction modalities
3. Companions

1. Generative AI tooling for personalization

One of the major shortcomings in no-human-in-the-loop treatments for mental health conditions is the fact that all products are “few sizes fits all”. Take Headspace as an example, while the product is very successful and I personally find Andy Puddicombe’s voice soothing and perfect for meditation that’s not true for everyone.

This exemplifies the simple fact that we’re all different. Producing high quality content suitable for meditation, CBT or similar used to be very expensive. Especially when offered in different languages and for a range of different types of techniques.

With AI the cost of content production goes down by probably 90% or more and while we’re not yet fully there on quality, we will be soon given the current rate of improvement. And while the industry is working on foundational models we can still use the tools to launch secondary languages, experiment with new voices and improve the editing toolchain with things like Descript.

These improvements enables personalization of content unlike anything we’ve seen before. Voices that are culturally relevant, accompanying graphics and animation, available in different languages and even dialects. I believe this will make products feel a lot more relevant for many more people.

• Timing: These improvements are available today
• Risk profile: Low, output can be reviewed and verified
• Impact: Medium, we’re making the current form factor more relevant

2. New interaction modalities

Let’s be honest, screens are likely not the best modality for delivering mental health interventions.

But Nikita misses an important point here. There is nothing wrong about building an app as long you don’t have a linear relationship between screen time and outcomes. For example one of our products, Spark Direct, uses behavioral activation which encourages only using the product for about 15 minutes per day to plan activities to do in real life and then reflect on how they affected mood at the end of the day.

However, maybe we can do even better? The new advancements in voice enables us to almost completely remove the need of a screen to deliver a digital intervention. Imagine instead of tapping on a bunch of buttons and viewing content the app will ask you to put on your headphones and go for a walk while you practice. That’s a very different experience and one that I believe has the chance to have a big impact.

And it’s not only about voice either, there could be other modalities enabled by models that are still unexplored.

• Timing: Depending on implementation we’re not fully there with a voice only UI. Latency is still a bit high for example but we’re probably close.
• Risk profile: Medium. There is some risk here given that a model is live in the product, processing input and responding. It should be manageable given the output is constrained but will require diligent research.
• Impact: High. I believe this to be a big improvement compared to todays experiences.

3. Companions

This final category is what most people gravitate towards when they think about AI and mental health. I don’t blame them, in a world where we know therapy helps while at the same time being inaccessible to most people due to labor shortage and low affordability, automating therapy is a very attractive idea.

This concept has been around for some time, with companies developing therapist chatbots and similar modalities. However, now that models can reason, have long-term memory, and be trained in psychotherapy and CBT, we are closer than ever to achieving an experience that combines the right mix of effectiveness, personalization, and elegance — all at an affordable cost.

There are also some challenges though. The go-to-market strategy decides timing, safety and also which claims these products can make about their efficacy. I’m a firm believer that health care should be covered either by your employer or an insurance company, when a product is self pay you limit the impact by pricing out the people who need it the most.

That said, choosing insurance will require research, evidence and partnering closely with both industry and government stakeholders to develop a framework and understanding for how products that utilize these models can and should work.

The company that takes that on successfully will have a very large impact on the world.

• Timing: Likely years out depending on go-to-market
• Risk profile: High
• Impact: Redraws the board

Final thoughts

This framework outlines my current perspective on the application of AI in mental health and I’m actively working with my team to integrate these new capabilities into our products.

One big question is if recent developments could help establish what is called Therapeutic Alliance which is often defined as “the affective bond between patient and therapist”. Digital programs have a real challenge replicating this concept which is a key determinant of effective psychotherapy. I’d be curious to see ideas and innovation for this specific part of the puzzle.

I welcome your feedback and perspectives on these ideas as we continue to navigate the evolving landscape of mental health care.

Thanks to Dane Wetschler for providing feedback on this post.

Excited to announce that about a week ago I joined Big Health as VP Product leading an incredible team of product managers and designers.

After selling Steady to Carbon Health last year I started thinking about how software can play an even bigger role in the transformation of healthcare than it does today. We already use software in our interactions with healthcare all the time, but more often than not it’s used to connect with a doctor, schedule a visit, or share labs — a means to an end. Rarely software is the thing that actually helps us improve our health.

But then I got introduced to Peter Hames at Big Health. He opened my eyes to an opportunity that I knew about but had not fully understood the importance of, Digital medicine.

In today’s America, there are many challenges with mental healthcare. It’s expensive, has long wait times and quality is highly variable. Six in 10 practitioners reported that they no longer have openings for new patients, and nearly half said they have been unable to meet the demand for treatment.

This means that doctors often use drugs as the “solution”. Following COVID-19, roughly one-fourth of Americans took or are taking some form of psychiatric medicine. That’s more than anywhere else in the world.

But there is an alternative. Big Health is on a mission to help millions back to good mental health using an alternative (or complement) to drugs or in-person care: Software. Our apps for insomnia (Sleepio) and anxiety (Daylight) are clinically proven and loved by patients. And, have all the magical characteristics of any software solution, scalable, cheap, infinitely personalizable, and is with you wherever you go. Software alone won’t solve the crisis but I fundamentally believe it can play a much bigger role than it does today. And we’re only scratching the surface of what’s possible.

Could not be more excited to join Peter, Arun, Gabbi, Jenna, and the rest of the team to continue to build world-leading treatments for mental health conditions and to expand the role of software in healthcare.

If you’re interested in working on this problem — DM me!

Now that we have fully replaced our existing app and service with something bigger and better in every regard it’s time to retire our old app. All patients receiving care have been fully transitioned to the Carbon Health diabetes program already but if you are using our app without care, please go and download the Carbon Health app here.

The Steady app and service will shut down on April 15th, at 10am PST.

If you have any questions, need an export of your data, or similar — please reach out to us at hello@steady.health, we’re here to help.

Finally, I want to thank everyone who helped Steady along the way — our team, investors, advisors, partners, and most importantly, patients. This is not the end, rather the beginning, of a long journey toward changing the lives of people living with diabetes.

Henrik

Founder & CEO, Steady Health

General Manager, Programs, Carbon Health

Saying goodbye to Steady Health was originally published in Steady Health on Medium, where people are continuing the conversation by highlighting and responding to this story.

I’m incredibly excited to share that Steady is joining Carbon Health in their mission to make high-quality healthcare accessible to all. Together, we’re building a health platform that takes care of all your needs — from urgent care to managing chronic conditions like diabetes.

The Steady brand and service is going to remain as we integrate with Carbon, supporting Carbon Health’s home-based care initiatives and further evolving its approach to managing chronic conditions. Over the next few months, we’ll be expanding across the country, as well as launching new products and services.

This summer marks three years since we started working on Steady. It’s remarkable how much has changed in continuous glucose monitoring (CGM) and diabetes in that short amount of time. At the time of my first blog post about using data to improve my diabetes, I was still self-importing the Freestyle Libre from Europe. Today, CGM is everywhere and is reaching people with type 1 diabetes, type 2 diabetes, and prediabetes. Three years ago I believed CGM had the potential to radically improve the lives of people living with diabetes. Now, it’s obvious how revolutionary this technology is, and more and more patients are getting access to it every day.

Over the past three years, we’ve built and launched the first CGM specialized virtual diabetes clinic in the U.S. We’ve built a suite of specialized tools that allow providers to monitor and understand patients and their diabetes at a depth not possible before. And we’ve provided care for hundreds of patients in our novel model that relies on chat, video, and continuous data instead of finger pricks and quarterly visits at a hospital.

“CGM is a revolutionary piece of technology with the potential to radically change diabetes care”

I first met Eren Bali, co-founder and CEO of Carbon Health, at a healthcare founder dinner back in 2018, and we’ve kept in touch since. But it wasn’t until earlier this year that we started talking more seriously about working together. At Steady, access to care has always been one of our key goals, and after speaking with Eren I realized that Carbon Health is focused on exactly the same thing, only from a different angle. With an integrated approach combining urgent care, primary care, and chronic-condition care, we have the chance to not only reach more people but also make the services we offer more affordable so that people can actually use them.

Eren and his team have built a remarkable company over the last few years. They now have 70-plus clinics across 13 states, and they plan to grow to 1,500 locations across the U.S. by 2025. And they have worked incredibly hard during the COVID-19 pandemic with both a robust testing program and impressive vaccine distribution. When I learned that they had been thinking about diabetes and chronic care and wanted to build out a new service, we jumped at the opportunity to be a part of it.

“Diabetes is one of the most pervasive and expensive health conditions in the U.S., and we are excited to partner with Henrik and his team at Steady Health to help elevate the management of diabetes.” — Eren Bali, Carbon Health Founder & CEO

To all of the Steady members, our CGM partners, clinical team, engineers, investors, designers, member guides, operations staff and everyone else that has contributed to Steady — thank you, we could not have gotten here without you.

Now a new chapter awaits, one that is bigger and even more exciting than the last one. So if you or someone you know is excited about our mission to fundamentally change diabetes care, come join us!

Henrik and the rest of the Steady team

Steady Is Joining Carbon Health to Accelerate Access to Personalized Diabetes Care was originally published in Steady Health on Medium, where people are continuing the conversation by highlighting and responding to this story.

This is part 2 of a conversation between Henrik Berggren founder and CEO of Steady Health, a digital diabetes clinic and Eric Antonow, formerly at Metabolic. You can read Part 1 here.

Henrik: It’s happening in real-time. People sign up for our service, Steady, and they come in with a wide variety of context. Some have questions about CGM, how it works, how to best use it. So, there is a massive value that these applications layers can provide. Whether you have diabetes or not, the essential question is, “How do I extract the most value from this new data?” And that is a problem that can be solved with software.

Eric: That’s what you seem to have focused on at Steady. How can you help people use this to manage their lives? I have this data, but what do I pay attention to? What do I not pay attention to? How do I make progress? And what support do I need to do that? That’s the layer that most people are going to want, whether they have diabetes or not.

This is the core of the diet community. When people sign up for Weight Watchers or Jenny Craig, they’re for support. They want help getting through this process. In the CGM space, that will happen at the software layer, but I also bet a lot of this will happen at the community layer, which is certainly true in the diabetes community. You can go to online groups on the web and social networks that have formed to help each other manage and interpret their CGM experience. Groups like Diatribe have been invaluable.

You’re starting to see the same thing in CGM outside of diabetes. Follow Levels on Instagram, and they repost other people’s CGM screenshots and stories. I ate this bowl of oatmeal or rice, and my blood spiked like that. Those stories are going to be very powerful for users as well as a source of growth for the companies. So, there will be an application layer, a services layer, combined with a community layer, and I think that’s going to be enormous.

One of them will end up being the Peloton of CGM. They went far beyond a spinning bike. It’s a spinning bike, live classes, support, shoutouts from tier 1 instructors, along with the leaderboard and the community of other people that you’re riding with. You’re not getting a bike but a successful exercise experience. This is going to be motivational and easy to commit to this multiple times a week.

With CGM, it will be the same. I’m not just getting a device on my shoulder; I’m getting advice, guidance, and a community that motivates me to make changes. Those layers are very early but emerging right now.

There’s a whole segment of people struggling with weight loss, at least one third of the population, and they are also people aspiring to bicycle road races. So I would be very surprised not to see many distinct companies show up.

Many of the existing companies and categories will need to spin something up very quickly in response. I don’t think you can be weight Watchers or Jenny Craig and not have a CGM service. That will be true of insurance companies and hospitals and others. So, we’re on the cusp of something quite large.

Henrik: One distinction with Peloton is that this will be much cheaper to try. For $75, you get a whole month of data. That’s an opportunity to run experiments and post them on social media to understand yourself, compare what you’re doing with others, and understand more about yourself. So, the impact-to-price is enormous, and many companies will be impacted by this.

For $75, you get a whole month of data

Eric: That’s a terrific lens to look through this. This is the data you get for doing CGM for one month. Imagine taking data to your physician for your annual physical. The diagnostic quality is better in some ways than the blood labs they can order. If you do this once a year, you can, for the first time, see your metabolic response to foods and whether it’s degrading.

In the US, we have roughly 1.5 million people with type 1 diabetes, 30 million people with type 2, and 100 million people with pre-diabetes: 1/3rd of the population. The largest group has a high probability of ending up with diabetes, and now you can see where they are on that path and help them. It’s so much higher resolution than our current measure, like A1C or a fasting glucose. So, whether you’re Anthem or Kaiser or Stanford health systems, even as a diagnostic tool, I would expect to see programs launch in the next 2–5 years.

Henrik: I am currently obsessed with diagnostics. We do such a poor job today understanding whether someone is at risk for diabetes. We don’t have a good risk scale or articulation of where you are on the journey to a full diagnosis. Half of the people living with pre-diabetes don’t know their condition, and their numbers might not even show up in their labs. They might come in with good fasting glucose numbers, and a minorly elevated A1C, so many doctors will tell them there’s nothing to worry about. We are going to be able to diagnose early and put in some preventative measures that might prevent diabetes altogether.

Eric: A1C is the default measure for most doctors for risk assessment. It is just an average over months, and as an average, it can be incredibly deceptive. You could be having very severe spikes that are masked by this average. We also know from people that treat diabetes closely that the system can just crash. One year the person’s glucose number looks okay, and the next year, the system capitulates, and they have diabetes. We don’t have a ton of data on why and when that cliff happens.

The minute you have granular data predicting the move from non-diabetes to pre-diabetes or type 2, it will be much easier to diagnose early and intervene. That matters for at least one-third of the US population, and the rest of the world is in a similar boat. In countries like India and China, 10% of the population already has diabetes. And with more centralized health care, I expect they’ll be motivated to find solutions. The same is true for most European countries. This is absolutely a global problem. We’re not alone in both the risk and opportunity.

Henrik: I wouldn’t be surprised to see your insurance company pay for you to wear one of these devices for two weeks every year. It will be part of your regular checkup or labs. It will show up in the mail, you’ll just put it on, and it will be connected to their system. If something is wrong, they’ll have a specific care plan to help you.

Eric: Let’s take a moment to talk about Apple and Google. They’ve had access to some diabetes data on the phones and, in some cases, the watches. I expect to see some real innovation in the next six to twelve months, particularly as these new CGM devices come up. Do you have any quick thoughts on that?

Henrik: Apple has been rumored for a long-time to be working on glucose monitoring or a CGM. And Google has been working with Dexcom through Verily, and they’re in the process of acquiring Fitbit. So, it’s going to be interesting.

Eric: These wearable devices have mostly been about paying attention to movement. We started with counting steps, and they added sensors for heart rate, and those have gotten more sophisticated. We can detect some arrhythmias and now we have HRV, heart rate variability. That let’s them connect that to stress and sleep. All of that has very limited diagnostic value, but with continuous glucose monitoring, you are now involved in meaningful insights and real healthcare.

All of these devices, when paired with a CGM, will move into a health diagnostic category on par with blood labs. You can do it in partnership with your doctor or on your own with one of these services. It’s going to completely shift consumer health to a different space. For many people, it will make healthcare a first-person experience and responsibility. This is your body — the tools and insights are on your phone, not in your doctor’s office or some lab. So, psychologically, it feels much more in your hands than someone else’s. You don’t have to go to someone else to ask, “Am I healthy?”

Henrik: I think that notion of first-person health is incredibly powerful. The power and responsibility are shifting from mostly the physicians back to the person. I don’t know if you coined this idea, but the depth of diagnostic insight or diagnostic value in the CGM devices is so different from any health tracking we’ve had before. It’s a good proxy of overall value that comes from this data set.

Eric: Totally. A whole category will emerge called something like “CGM Health.” There will be an entirely new set of businesses in CGM Health, and the existing ones — whether they’re hospitals, doctors, dietitians, or insurance companies — will have to get into it, too. It’s the category that’s emergent.

A whole category will emerge called something like “CGM Health.”

I’m curious in the context of Steady: Do you see this behavior of first-person health, particularly as people get CGM devices?

Henrik: Absolutely. In diabetes specifically, we’ve been treating everybody with a care model that’s one-size-fits-all. If you have a diabetes diagnosis, you get a quarterly visit with your doctor for the rest of your life. This is regardless of what you’re struggling with and so on.

With this data and insight, we can be much more custom for every single person. We can spend more time with the people that are really struggling with something. Maybe they had their first child, maybe they just moved or changed jobs. All of those impact food access, exercise, and lifestyles. For people who are doing well, we can see that in their hourly data and take a different strategy. We can answer specific questions but they need less of our time.

We let them be in control of when they need help, but we can now be proactive. So, typically, we reach out to them and say, “Hey, we’ve seen that your numbers have not been as good as we want them to be for the past two weeks. What’s been going on with you?” That kind of opens the door to a different level of care and support.

Eric: You can imagine a very analogous conversation for someone who’s on a diet. If you’re trying to help someone manage weight loss, you don’t have data to see what’s going on. If you give your physician or dietician access to your CGM data, they can help monitor and support you. That’s hard to imagine right now, but we’re not far.

Henrik: That’s such a big piece. We need to help people build behavioral changes over time. The feedback and learning loop is just too long. You can think back to your behavior over the last month to try and to triangulate what you ate. It just doesn’t work.

Eric: People can debate how quickly this will happen, but I think, because of social media, at least awareness of CGM will be fast. It will be very much like Keto over the last few years. You’ll watch people posting CGM data on Instagram or wherever. You’ll see influencers or emerging fitness and diet people. They’ll start posting their CGM journey, and that will educate a few million people. It will be very niche, but it will grow quickly.

I also expect to see an “Al Roker” of CGM. In 2002, when he got gastric bypass surgery, it was much more rare. He came back to the Today Show after surgery, and everyone saw the shocking results. He was a different person. Those surgeries are roughly 10x more popular now, and Roker is responsible for making many more people aware of it. Someone high profile will do that with CGM.

I expect to see an “Al Roker” of CGM

Henrik: Yes. That’s right. Who’s going to be Al Roker of CGM? This is been a good conversation. Thank you, Eric.

This is part 2 of a conversation between Henrik Berggren founder and CEO of Steady Health, a digital diabetes clinic and Eric Antonow, formerly at Metabolic. This conversation is also available on YouTube and you can read Part 1 here.

A conversation on CGM, part 2 was originally published in Steady Health on Medium, where people are continuing the conversation by highlighting and responding to this story.

This is part 1 of a conversation between Henrik Berggren founder and CEO of Steady Health, a digital diabetes clinic and Eric Antonow, formerly at Metabolic.

Eric: Henrik, good to see you. There have been a bunch of announcements in the CGM space, and you wrote a Medium post about them. I think they’re a pretty big deal, so let’s go through them one-by-one.

Henrik: Sounds good. Let’s start with Abbott. They’re one of the biggest companies and seemingly on the trajectory to be the most innovative. First, their announcement of the third generation of Freestyle LIBRE or LIBRE 3. It was announced in Europe, and we actually don’t know how long it will take before it comes to the US.

There were three major announcements. First is a 70% size reduction compared to the LIBRE 2. The LIBRE 2 was the size of two stacked quarters; this will be only two pennies!

Eric: It’s miniscule.

Henrik: How it looks on your arm or your body is important. It can cause people to hesitate when buying a CGM device for the first time. So, that’s definitely a meaningful improvement. The other change is the LIBRE 3 is now using Bluetooth. That’s been the weakness of the LIBRE compared to some of these other manufacturers, like Dexcom or Medtronic. Previous LIBRE versions used NFC, and you had to tap it and scan it manually to get the data. This version now comes with full Bluetooth capability, making it true CGM. The update rate is now every minute compared to the previous standard of every five.

The third announcement is that the price will be the same as the current version. That begs the question of how pricing changes on existing devices and what that will do to the market as a whole.

Eric: Bluetooth is huge. It’s what separated LIBRE from the Dexcom devices. Now, you can wear this passively and have the data to show up on your phone or your watch. This competitive usability changes quite a bit. Also, NFC limited its connectivity to a smaller set of phone and watch devices, so this really opens things up, especially for Apple devices.

The market for the layer of services built on top of CGM devices gets much much bigger. Apple and Google have historically been more connected to Dexcom devices just because of Bluetooth. But with this, they have another hardware option, and because it’s significantly cheaper, it will attract an entirely new set of customers.

Henrik: These are the two biggest CGM manufacturers, and now they are at the top of their game. The delta between the two solutions is now almost zero except for price. So, it’s going to be very interesting to see how Dexcom responds, particularly when the G7 comes out.

The second big Abbott announcement was the LIBRE Sense, the first CGM device for the non-medical market. This opens up so many possible applications for blood sugar monitoring. There are athletes interested in performance optimization and people looking to lose weight.

The LIBRE Sense is specifically nonclinical. It’s a LIBRE 2, but they’ve put on guardrails to segment the regulatory market. The range of measurement is limited to 70 to 200 mg/dl. That’s totally sufficient for someone without diabetes. By doing this, they can potentially become an over-the-counter product in the United States because they aren’t serving an at-risk patient.

Again, this product announcement was for Europe, but I suspect it will quickly come to the US and be the first CGM approved for non-diabetics. That’s a landmark.

Eric: I’ve had many people ask me about getting CGM, primarily for diet purposes, and they’re absolutely willing to pay. The only issue has been access to a device. They or their doctor doesn’t want to go through the hurdles of a prescription. Outside the United States, you can buy this over-the-counter, but here, it’s a huge point of friction: you have to go through a physician, explain why you want this, and get a prescription.

I’ve had many people ask me about getting CGM, primarily for diet purposes, and they’re absolutely willing to pay.

I think this device is a great path to over-the-counter CGM in the US.

Meanwhile, there is an emergence of other consumer CGM services that are also trying to solve the access problem. And so far, they’re doing well.

Henrik: One other area that’s shifting is around access to the data on the devices. Abbott announced that they’re also launching a developer platform, and traditionally they’ve been more closed. Dexcom has performed well because they’ve been more open, allowing them to do things like power loop systems, automated insulin delivery, and a wide range of other apps. As we start to see these new consumer services get built, Abbott as a platform becomes very important.

Eric: Well, this new segment of people without diabetes also drives that. There’s less risk in giving data access here. These patients don’t have an acute health condition and they’re not dosing insulin. There’s no medical risk argument for keeping this segment closed.

Henrik: One of the core hypotheses I have around CGM is that it will parallel GPS and maps. The device itself generates data, but the next layer of services is even more powerful — similar to what we saw in GPS development. First, there were Google Maps, and then you have Lyft, Uber, DoorDash, and so many more. You start thinking about applications that will come from CGM and the glucose data that exists in our bodies.

The device itself generates data, but the next layer of services is even more powerful

Eric: That’s absolutely right. Those services are coming.

Henrik: Let’s shift to Dexcom. In terms of functionality available today, they are the top, top player — solid platform, data access, and the highest accuracy. That’s kept them ahead. The next model, the G7, is coming out soon. They’ve been developing it with Google’s Verily spin-off, and it’s going to be much smaller. It’s still going to be 10-day wear, so that’s three sensors a month versus two for the LIBRE. The biggest delta was already the costs, with Dexcom’s monthly costs being roughly $500 to LIBRE’s $75. That’s big.

Eric: Segmentation and price pressure will be interesting. The minute you have these within the general CGM consumer market, you also have segmentation around precision. Accuracy matters much less for people with diabetes, so you don’t have to win there, just place. At the same time, Dexcom can continue to point to higher accuracy and precision. For some people, that will be worth it, but the price difference will be harder to maintain.

It’s also worth noting the Dexcom devices have become much less intimidating to apply. The G5 looked like sticking a syringe into your abdomen. The G6 was a significant improvement, and I expect the G7 to be even easier for people to handle.

Irrespective of the device, this is probably the second major friction point for trying a CGM solution. People are intimidated by the idea of attaching one of these to their bodies.

It’s interesting that one of the consumer companies, Levels, made a good-looking patch that covers the actual CGM. It makes it look cool, like a tattoo or badge. It’s almost something that you’d be proud to be seen with at a gym, the way people wear Kinesio tape.

Henrik: The Dexcom G5 was basically like a gun with a needle that you had to stick into your stomach. I got sweaty palms when I looked at it. The G6 has this cute, Star Wars-looking applicator with a button, and you press and you’re done. It makes a huge difference.

Eric: Making it entertaining is great and removes some of the fear and intimidation. Let’s shift to talk about the newer consumer services.

Henrik: When I wrote the blog post, I mentioned a few of them, but more have popped up, like January.AI. Two of them, Levels and Supersapies, are capturing a lot of attention and worth discussing. At their core, they both serve healthy people. They ship them two CGM devices a month and give them access to an app and program to interpret their data. These services help them see how they’re reacting to food and exercise as well as other things like stress.

The customer targets feel slightly different. Supersapiens is focused very much on performance. It was started by a pro cyclist who happens to be diabetic. They’re one of the first companies building on this new developer platform that Abbott is launching.

Eric: To me, the future of this market is how these services differ. The interpretation layer on top of the CGM is early but the next important problem. Once they’ve given you access to a CGM, which they do solve well, the next question becomes: “How do I interpret this data?” From what I can see from Levels, they’ve done a great first version

For people with diabetes, you can find good endocrinologists or books on how to use a CGM device, but for the general population, there is almost nothing. So, the application and service layer that Levels is building is a big deal. It contains a thesis about what to pay attention to, what’s good and bad. The science is still pretty early, so it’s critical to get that guidance.

This is where the markets will segment again. You’re going to see one brand focus on elite performance, like Supersapiens is doing. It will be about optimal conditioning or preparing for a race. Then, another brand will be more focused on diet and nutrition and will target the larger market. We’ll see if that’s where Levels ends up going or if they choose something like general fitness. They can both be positioned as premium and go after very different non-diabetes problems.

This is part 1 of a conversation between Henrik Berggren founder and CEO of Steady Health, a digital diabetes clinic and Eric Antonow, formerly at Metabolic. Part two will be released shortly.

A conversation on CGM, part 1 was originally published in Steady Health on Medium, where people are continuing the conversation by highlighting and responding to this story.

Does your family have a history of diabetes and/or pre-diabetes? Sign up for our starter program, try CGM and get your results analyzed by experts. Sign up here: https://steady.health/welcome

A series of new announcements and initiatives together with the ongoing pandemic has accelerated the adoption of the CGM across several groups of consumers. I predict that “What is our CGM strategy?” will be a common question asked in board rooms and strategy off-sites of healthcare companies over the next few years. This article covers some of the big trends in the market and tries to paint a picture of what’s to come.

A year of landmark announcements

Even though CGM technology is about a decade old it is first during the last year or so it has reached significant momentum. Here are a few key recent announcements.

Abbott announces Freestyle Libre 3 and Libre Sense

A few weeks ago Abbott announced the release of the 3rd generation of the Freestyle Libre sensor. This is the first big update to the Freestyle in a number of years and will feature a much smaller device as well as complete Bluetooth capabilities. This removes the need to scan the device with a phone which is one of the main drawbacks of the Freestyle Libre system. It will also transmit your glucose to your phone every minute which makes it more continuous than any other sensor currently on the market. This is a huge step forward.

Alongside the Libre 3 Abbott also announced the first CGM ever for people not living with diabetes. It’s called the Libre Sense (What’s up with this very confusing naming???) and is targeted to athletes who wants to optimize their performance. This could be a sign that Abbott has managed to ramp up their supply chain and are starting to feel comfortable with experimenting with different groups of consumers. The announcement comes alongside a partnership with Supersapiens, a company building software to collect and interpret glucose data that is founded by people from Team Novo Nordisk, the professional cycling team. Seeing CGM finally breaking out of diabetes and reach new users is incredibly exciting.

Dexcom and Lilly partner on diabetes management platform

CGM is also getting more integrated into various other parts of the diabetes ecosystem. Lilly and Dexcom recently announced a partnership that although light on details outline tighter integration between big pharma and CGM manufacturers. To me this means that insulin companies have finally realized that CGM is going to be a key part of the management ecosystem over the next few years and that they need to integrate to stay competitive.

OneDrop acquires Sano

Earlier this year OneDrop announced that they had acquired Sano — a startup that built their own CGM but failed to take it all the way to market. It’s obvious that the OneDrop solution of providing a BGM to users is soon going to be a thing of the past, despite the nice design, and it’s really smart of them to acquire what could potentially be their own CGM hardware.

Consumer plays are blowing up

NutriSense, Veri, Lifetzr, Levels, Supersapiens. These are all companies that repackage the Freestyle Libre and sell it as their “own” sensor together with access to an app that helps track food, exercise, sleep and more. They target healthy people and promise that blood sugar monitoring can help with weight loss, sleep, energy levels and generally optimizing your health. I’m a huge believer in monitoring blood sugar but the science behind health optimization is still pretty thin according to most doctors. Hopefully this new interest in the technology drives more research and results in the next few years.

What’s next?

Last I heard there were over 30 companies somewhere in the development cycle of improved hardware. Given how complex it is to bring something like this to the market I believe it will take a number of years before we see a dramatic increase of competition on the market but its definitely coming.

I’m continuously excited about the upcoming launch of the G7 by Dexcom which is supposed to be disposable, smaller and cheaper than G6. It has been delayed due to the pandemic but should be here sometime next year.

CGM is the most impactful innovation in diabetes care since the discovery of insulin

I believe we’re in the very early innings of CGM and that it will prove to become the most important innovation in diabetes care, and health monitoring more broadly, since the introduction of insulin.

Does your family have a history of diabetes and/or pre-diabetes? Sign up for our starter program, try CGM and get your results analyzed by experts. Sign up here: https://steady.health/welcome

What is your CGM strategy? was originally published in Steady Health on Medium, where people are continuing the conversation by highlighting and responding to this story.

At Steady, one of our goals is to make wearing a CGM as valuable as possible. For our members and users of our brand new starter program we do this in a few different ways:

1. Have our doctors analyze your data and send you insights
2. Answer your questions through a secure messenger
3. Enable you to run experiments by logging events in our app

And today we’re excited to bring one of our most popular features to anyone already with a CGM, for free!

Through our partnership with Dexcom and integrations with other manufacturers we allow you to easily see the impact of food, exercise and medication on your blood sugar. Share it with friends and family, with other CGM wearers to share tips and insights and educate the world on how food affects us. We’re in the early innings of a new data-driven world where you can actually see how your life and wellbeing is impacted by your lifestyle.

How to get started

1. Download the Steady app from your app store here
2. Sign up and connect your Dexcom or Freestyle Libre
3. Tap the green +, snap a photo to log the meal
4. Check back a few hours later and see how your body reacted

If you don’t have a CGM yet but want to try it out, no problem, check out our newly launched Starter Program which gives you CGM + expert analysis and answer for 4 weeks — its a great way to try out our connected care.

We hope you find this feature useful, let us know what you think.

Get more value out of your CGM— For free was originally published in Steady Health on Medium, where people are continuing the conversation by highlighting and responding to this story.

Diabetes can be overwhelming. Diabetes care shouldn’t be. We’re making it easier to experience the magic of Steady by introducing a 5 week starter program that’ll help you take control of your diabetes.

Since we launched Steady in 2019, we’ve seen how our patients’ lives are improved by starting with a Continuous Glucose Monitor (CGM) and using that data to receive the personalized treatment from our care team.

Our new course combines the powerful insights of a CGM with real-time coaching to help you achieve meaningful goals.

We’ve heard from our members that they’re interested in doing a trial with a CGM and the Steady care team before signing up for ongoing care. This 5 week program offers all the benefits of Steady — care connected to your CGM data — for a simple $50 fee. We’re thinking of this like “diabetes care in a box” — straightforward, powerful and accessible.

• Stressed because your levels are high? Message your care team in our app
• Need a CGM? Our doctors will set you up and ship it to your house
• Want advice for a good breakfast routine? We’ll provide recommendations

Let’s say you are newly diagnosed with Type II diabetes and you’ve never used a CGM before. You’re interested in lowering your A1c, tired of finger pricks, and prefer to make changes with diet and exercise rather than medication. What can you expect from joining Steady’s 5 week program?

How our program works

The first week, we’ll get you set up with a CGM, shipped straight to your house. Our doctors will give you a prescription for either a Freestyle Libre or Dexcom G6 Pro, we’ll work with your insurance, and get the device delivered.

You’ll meet our Certified Diabetes Educator, Aimee, over video. She’ll show you how to get your CGM up and running.

At the start of the program, set your goals for the 5 weeks. Aimee will help you set a meaningful, but achievable goal for your time in range or A1c.

Then it’s time to give your CGM a try. You’ll be able to see your blood sugar in real time on your phone, updated every 5 minutes. The initial tracking period measures how your body reacts to food and lifestyle.

After tracking, you’ll get your first personalized insights report. This in-depth report explaining how diabetes is affecting you and where you need to improve to reach your goals. Maybe we should try swapping corn tortillas rather than flour tortillas at that taco shop?

Then, at the end of 5 weeks, you’ll use a CGM for another 7–10 days so we can measure your progress towards your goal and recommend areas of improvement. It’s time for a celebration and you’re taking control of managing your diabetes. You’ll wrap up the program with a 20 minute CGM interpretation with our Endocrinologist where they talk you through your results and suggest next steps.

We’re very excited to be introducing this program and helping more people manage their diabetes effectively. I believe that CGM and connected care is a powerful combination and we’re thrilled to be able to expand access to more people. If you’d like to join the new Steady 5 week program, get started here.

One more thing, from today Steady is also available in WA

The most energizing piece of working on Steady is all the people reaching out from across the US wondering when we are coming to their state. Therefore I’m extra excited to announce that we’re, starting today, live in our second state: Washington. To deliver incredible diabetes care throughout the state we’ve partnered with Dr. Sushma Reddy who’s an endocrinologist with over 30 years of experience.

I’m proud of the work of the passionate and motivated team at Steady for making our new starter program possible, and I can’t wait to hear how it might help individuals like you. As always, please share any feedback directly with me at henrik@steady.health.

Henrik

Introducing The Steady Starter Program was originally published in Steady Health on Medium, where people are continuing the conversation by highlighting and responding to this story.

Whether you have diabetes, are struggling to improve your diet, or are just curious to understand how your blood sugar changes over the course of a day, a continuous glucose monitor (CGM) can be a really valuable tool. In this article, I’ll explain what a CGM is, how it works, and what you can learn from using one. I’ll also review three different models that are currently available in the US, and review their functionality, price, and more.

What is a Continuous Glucose Monitor?

A CGM is a small wearable device that you attach to your body to measure the concentration of sugar in your blood. Monitoring blood sugar levels is critical for anyone living with diabetes, but can also give insight into how things like food, exercise, and medication affect your body.

Without a CGM, measuring your blood is both painful and cumbersome. The most common way to do this is via the finger prick method — you prick your finger, place a drop a blood on a test strip, and insert it into a meter to get a single reading. With a CGM, you automatically get an updated reading every 5 minutes. This makes it far easier to keep track of your levels throughout the day, even while sleeping or exercising.

There are three main manufacturers of CGMs on the market today: Dexcom, Medtronic, and Abbott. I’ll be comparing their latest models below.

How do CGMs work?

All CGMs on the market today use roughly the same technology. They are invasive, which means they puncture your skin to measure glucose through interstitial fluid found just a few millimeters below.

The sensor’s filament is coated with glucose oxidase that converts glucose into hydrogen peroxide, which reacts with a platinum receiver. This generates an electric signal that is then sent to a chip located inside the outer casing.

These signals are then sent to an external transmitter (either a custom device or your phone) where an algorithm is used to predict the actual glucose level in your blood. This algorithm is not 100% accurate, but has high precision and is approved by the FDA for use by both patients and doctors. In fact, it is shown to be more accurate than some finger prick meters.

What can you learn from a CGM?

At any given moment, 42 different factors can impact your blood sugar level. This means there are many things we can learn from gathering and analyzing more data. The three most impactful categories are diet, exercise, and medication, closely followed by sleep and stress. Some examples of things you can learn include:

• Which meals spike your levels, leading to fatigue and hunger?
• What time should you eat dinner to ensure stable overnight glucose levels?
• Which type of exercise has the least impact on your levels?
• What happens to your levels on six vs eight hours of sleep?

Which CGM is the best? Let’s compare the Freestyle Libre, Dexcom G6, and Medtronic Guardian.

In this section, I’ll compare the three major brands in the US today. I’ve personally worn a CGM every single day for the past three years, have tested several different models, and have used over 100 individual sensors. I am also in regular contact with all three manufacturers through my work at Steady Health.

Abbott Freestyle Libre

Let’s start with the most popular model, the Freestyle Libre from Abbott. It’s the most common device due to its sleek design and low cost, but it’s not delivering data continuously, as it requires you to actively scan the sensor with your phone to download the latest data. (It’s actually considered a flash glucose monitor, or FGM.) The sensor itself stores up to 8 hours of history, so as long as you scan regularly you won’t miss any data.

Functionality: The sensor is low profile, easy to insert, and stays on pretty well. Towards the end of each session, you might want to put on some tape to ensure it stays on. It has a 14-day wear time and a 1-hour warm-up period before it’s ready to use.

The app experience is basic but works well. It shows you daily patterns, average glucose levels, and other standard charts. It also allows you to log events like meals and exercise, but does not support pictures which makes it harder to use for food journaling.

Cost: The price is its largest benefit. Every sensor can be worn for up to 14 days and the cash price is around $60 per sensor. If you have commercial insurance, many pharmacies guarantee a max out of pocket amount of $75 per month. This means you can each sensor for $37.50.

In summary: Great wear experience, 14 day wear time and affordable, but only has basic app functionality. I consider Freestyle Libre to be a great choice for anyone that is curious about CGM and doesn’t need all the bells and whistles.

Dexcom G6

Next up is the G6 from Dexcom. This device is the most advanced CGM on the market as it sends data directly to your phone via Bluetooth. The data it provides is also available in an API so you can easily access it in raw form.

Functionality: The G6 sensor is both larger in size and higher profile than the Libre, but still small enough to not be in the way. It has more of a medical look and the adhesive around it is bigger too. It has a 10-day wear time and a 2-hour warm-up period before it can be used.

The app experience is similar to the Libre: basic, but works well. You can log events, but pictures are also unsupported.

The biggest advantage is that it does deliver data continuously directly to your phone. This means you can easily set up automated alarms and see your latest numbers via a connected watch.

Cost: The Dexcom G6 is significantly more expensive. Every sensor is about $115 if you pay out of pocket, and unlike the Libre, you also need a separate transmitter (which sends data from the sensor to your phone) that’s attached to it. Transmitters cost about $240 and can be worn for three months before running out of charge. So all in all, you are looking at $425 per month for a complete sensor and transmitter set.

In summary: Dexcom has all the features you want, including a Bluetooth connection and open API, but this is reflected in its hefty price tag. If monitoring your levels closely is important to you, it’s a great choice.

Medtronic Guardian Connect

The Guardian Connect from Medtronic is the least popular CGM of the three. It’s lagging behind both in overall functionality and experience, but has a few advantages as well.

Functionality: The Guardian sensor lies somewhere between the Libre and the G6 in terms of size, and has a relatively low profile. But, it only has a 7-day wear time and a longer 2-hour warm-up period. The Guardian also requires calibration via a finger prick measurement done twice a day; this is a significant drawback as it is both painful and cumbersome.

One big advantage of the Guardian Connect over Dexcom is that the transmitter is rechargeable and reusable.

The mobile experience is on par with the others and supports automated alarms, smartwatch support, and more. You can similarly log events in the app, but pictures are again unsupported.

Cost: Due to the rechargeable transmitter, the monthly cost for the Guardian Connect is somewhat hard to predict. Every transmitter costs $775 and a box of 5 sensors is $553. If we assume the life of a transmitter is 24 months, the monthly cost of this system is $474.

In summary: Since the Guardian Connect system requires both daily calibrations and has a significantly shorter wear-time, it ends up being both more expensive and less usable than the other models.

How to get a CGM?

To get your hands on a CGM in the US you need a prescription. If you have any type of diabetes it should be relatively straightforward — just ask your doctor during your next visit or sign up for Steady and we will prescribe one for you. If you don’t have diabetes the cheapest option is to ask your Primary Care Provider for a prescription. You can also use one of the relatively new CGM consumer services on the market, but be aware that these may cost you up to five times more per sensor than just going directly to your PCP.

I hope this overview is helpful for anyone looking to learn about CGMs, the different models, and which one to use. If you already use a CGM today and struggle with blood sugar control, check out Steady Health, as we specialize in CGM-based care for diabetes: https://steady.health/

The ultimate guide to Continuous Glucose Monitors (CGM) was originally published in Steady Health on Medium, where people are continuing the conversation by highlighting and responding to this story.