Of course we’re ALL ABOUT the diabetes device hacking and data-sharing efforts going on throughout our D-Community these days — the collective push for open platforms to allow more meaningful access and use of our devices and data known as the #WeAreNotWaiting movement.
One individual working hard on this front is type 1 Doug Kanter in Brooklyn, NY, who several years ago founded a diabetes data software startup called Databetes. We’ve loved being in touch with Doug over the past few years, and most recently welcoming him to our DiabetesMine D-Data ExChange event held in conjunction with the American Diabetes Association’s Scientific Sessions in June.
Today, we welcome Doug right here at the ‘Mine as he shares more about his own diabetes story and what led up to his Databetes efforts that have captured our community’s attention.
A Guest Post by Doug Kanter
“Your type of diabetes will be cured in the next five years, 10 years tops.”
I still remember my first endocrinologist saying that to me way back in 1994, eight years after my diagnosis with type 1 diabetes at age 12. He was convinced that the product evolution from insulin pump to artificial pancreas would be a fast one.
In the 20 years since, promises of a coming cure have remained on the horizon. While the pace of change remains frustratingly slow, I also find it important to remind myself how much I’ve benefited from advances in diabetes technology. I’m grateful that the same endocrinologist I mentioned above was also an early adopter of insulin pumps and put me on one 27 years ago. And in recent years, my Dexcom CGM has also become an essential part of my treatment.
Yet I don’t need to remind readers here that challenges with this technology remain. My main motivation for creating Databetes and working to design software for patients was frustration with the current batch of software from device manufacturers. The lack of access to all my data, the lack of interoperability between devices, and the lack of software for the Apple platform all contributed. I also wanted to help bring the same level of smart design and functionality that we’ve grown used to on the consumer internet and apply it to the diabetes sector.
While I’ve been thinking about this topic for a long time, I began Databetes development in earnest on Jan.1, 2012. That day I started a yearlong experiment to test the core idea behind Databetes, that aggregating all my diabetes-related data in one place would allow me to improve my own health.
That year turned out to be the healthiest of my life, with my A1c readings improving nearly a full point.
My love for data and design was making a difference on my life with diabetes, but it wasn’t always my obsession.
In Pursuit of Flexibility
I’ve always tried to take advantage of the benefits of diabetes technology. For me, the best thing about going on an insulin pump was the flexibility. I never liked the regimented schedule that came with taking shots of both short-acting and long-acting insulin, requiring me to dose and eat at specific times throughout the day. With a pump as part of my treatment, I was able to follow my interest in photography and begin a career as a photojournalist in New York City. Despite the unpredictability of covering news, I was always able to manage my diabetes. My bosses never had to consider my condition when assigning me to handle a story, even during such major events as 9/11. As a result, my pictures for news wire services were seen around the world.
In 2003, I pursued my interest in working internationally and moved to Beijing. During my eight years there, I covered stories around China for major news organizations including Business Week, Bloomberg News and the Financial Times. A highlight of my time there was photographing the 2008 Olympics for the New York Times. My whole time overseas, I was able to keep my blood sugars in check by seeing my doctors and restocking supplies on trips back to the U.S. a few times a year.
A Year of Tracking Everything
In 2011, I decided to return to New York, end my career as a photographer and focus on Databetes development. To help with this transition, I enrolled in a two-year graduate program at NYU called ITP. One of the things I found interesting about ITP was its focus on humanizing technology and learning by making things, not taking tests.
As I learned the basics of how to code, I began to focus on data visualization. I explored new ways of making sense of medical readings by importing my own diabetes data. An early project I created was “Insulin on Board,” a visualization of 100 days of CGM and insulin pump data. I wanted a better way to look at my eating habits and their effect on my blood sugar levels. While none of my doctors have ever pressured me to adopt a low-carb diet, I was interested in exploring whether the days I ate less carbs were also the days with the best control. I also wanted see my insulin data in a way that factored in the drug’s latency, showing me when it was actually “kicking in” rather than when I took it. I often take a staggered series of small bolus doses and wanted a visual representation of the aggregate effect.
The yearlong self-tracking experiment that I described earlier became the basis for my thesis. Throughout all of 2012, I kept track of every blood sugar readings from my glucose monitor and CGM, every insulin pump dose, a description of every meal I ate, meal photos and location data. I also trained for and ran the Philadelphia Marathon, tracking the exercise with a FitBit, Nike FuelBand, a heart rate monitor and the RunKeeper mobile app. I saw a 40% reduction in my insulin basal rates when I was at peak training compared to when I started. After completing the marathon the weekend before Thanksgiving, I stopped running for a few weeks to recover. That change, in combination with stress from finals week at school, resulted in significant increases in my insulin rates in December.
Completing this project meant that for the first time in a quarter century of living with diabetes I had a complete picture of my year in diabetes. I decided to design a way to help me make sense of the 91,251 CGM readings and thousands of other data points. What were my trends throughout the year? Was my control in the winter better than in the summer? How did the start of the year compare with the end? What was my best day, what was my worst and why? How do you define the worst day anyway, based on average blood sugar or the volatility of the readings? How did eating at restaurants affect me differently than home cooking? These were some of the questions I wanted to explore. I did this with several visualizations that filled both sides of a poster.
Since 2012 I continue to self-track, but with a little less intensity. I still analyze my CGM readings as well as log my exercise and meals. I follow groups like Quantified Self and think they’re doing great work. I also remain intrigued by new activity-tracking apps like Moves (until they get acquired by Facebook and reverse their policy on data sharing).
As I completed graduate school, Databetes won awards from NYU and received early funding from the Dorm Room Fund. This has allowed us to scale development. Our first products are all patient-facing software that builds on lessons learned from my self-tracking experiments. We aim to make the same process of self-management easier for other patients. Much of our focus is on mobile, making the data actionable when and where patients need it. We also prioritize merging lifestyle information, such as nutrition and exercise readings, with medical data. This approach provides patients with the context they need to understand and respond to changes in their readings.
Our first software release is a mobile app called Meal Memory (available on Google Play, coming soon to iOS). We began with a focus on nutrition after talking with dozens and dozens of patients. Consistently, managing food was listed as the biggest problem patients faced. Meal Memory is designed to make the process of recording both what you ate and its effect on your blood sugar as easy as possible. Logging a meal starts with a photo. Users can then enter a carbohydrate estimate and a pre-meal blood sugar. Two hours later, we send an alert asking for a post-meal blood sugar reading. Comparing these readings gives the patient a sense of how well they are balancing their meals and medication.
Meal Memory also is designed for those of us who are creatures of habit, often eating the same meals at our favorite restaurants or at home. When a patient eats a meal again, all their past information is actionable and can be used to better manage that food this time.
Beyond the details of each meal, we also wanted a way for patients to look at their overall eating habits. Our meal log appears as a photostream. The pre- and post-meal blood sugar readings are color-coded and layered on top of the photos, allowing a user to quickly scroll through and see how often they’re in-range after eating.
The Future is Open
Databetes has started with a focus on self-management software for actively engaged patients. We are also speaking with health care providers about developing clinician-facing tools to help doctors more efficiently manage diabetes data and integrate it into treatment regimens. There is terrific potential in this technology to encourage and facilitate behavior change.
Affecting our development is the issue of open access to device data. The diabetes patient community is focused on this, and rightly so. The #WeAreNotWaiting movement and Tidepool have done a great job of communicating how important this issue is and pushing for device manufacturers to change their approach. The CGM in the Cloud group has shown that there is considerable market demand among users for new services. Beyond the obvious potential for improving health outcomes, it also seems like simple good business sense for these companies to address user demands. A rising tide could lift all boats here, helping everyone from patients to doctors to device manufacturers do better.
Open data will allow Databetes to design products that are even easier for patients to use, enable better analytics tools, power improved feedback loops and ultimately translate data into actionable insights. We’re encouraged by the signals we’re seeing from many industry players. The entry of major tech firms like Apple, Google and Samsung into health care will also have a big influence, giving us hope that there is light at the end of the tunnel on this issue.
In the years since starting Databetes, I’ve learned to see diabetes from many different points of view beyond being a patient. Despite all the challenges of working in this complicated sector, I remain encouraged that better use of existing technology can help ease the burden of diabetes and improve our lives until that cure arrives.
Doug, you are our hero! We look forward to seeing your Databetes vision materialize, and are just as excited to be part of the #WeAreNotWaiting surge as you are!