Technology in Diabetes Apps Improves, But Will It Bring a Change in Behavior?

Evidence-Based Diabetes Management, December 2014, Volume 20, Issue SP18

The announcements flow steadily from computing giants, from software startups, and from medical specialists such as Dexcom, which now sells a device that automatically sends glucose readings to tablets and smartphones.1 Each day brings some new wireless technology or smartphone software that helps patient with diabetes monitor and manage their health. Many of these improvements automate the process of data collection. Others present data in useful ways. Still others transform it directly into advice for immediate action.

All that news has generated much excitement about a general revolution in personal healthcare,2,3 but questions remain about the true impact on day-to-day well-being. Optimists expect major improvements in the not-too-distant future. Technology, they believe, will soon automate not only the laborious task of collecting data on nutrition, exercise, and blood sugar, but also the complex process of using those data to calculate optimal behavior.2 Patients who now complain of working a second job to care for chronic ailments will live easier, healthier lives.

Pessimists disagree. They predict significant limitations in technology’s ability to collect some of the most meaningful data and to make treatment decisions that take into consideration how dramatically patient responses can vary in seemingly identical situations.3 Pessimists also dispute the underlying notion that people today suffer poor health because of a lack of information

rather than an inability to force themselves to make healthy choices.4 The evidence, to date, is mixed.

On the one hand, the app stores for Google and Apple customers offer about 1100 programs designed to help patients with diabetes manage their health or otherwise cope with their disease. On the other hand, surveys suggest that only about 1.2% of all Americans with diabetes actually use any of those apps regularly.5 “Most of the apps that are out there right now do only one thing, and it’s typically not even a particularly useful thing,” said David Kerr, director of research and innovation at the William Sansum Diabetes Center in California. “The single most common thing we see right now are apps that ask you to laboriously type in lots of data before then presenting those very data back to you in the form of a pie chart or a bar chart. Apps need to be designed with ‘health clarity’ in mind. They need to be easy to understand and use. They also need to take data and put them into a personal context in order to guide the user to change behavior.”

Kerr blames a number of factors for the “shambolic” state of the health-app industry, factors that range from the general lack of design input from individuals with diabetes, to the inherent difficulty of producing something sophisticated enough to be useful, to the understandable concerns about regulation from the FDA, to the financial risk of making a world-beating application

affordable to consumers.

These are all serious concerns for app makers, who tend to succeed by quickly churning out programs and then giving them away or selling them cheap. Investing millions of dollars up front on a single app and then hoping that it proves valuable enough to recoup that cash is a risky proposition. It’s riskier still to design something that might fall under FDA scrutiny and thus may have to prove itself safe and effective before it can go on sale.

The FDA only regulates a tiny fraction of health apps—theoretically just those that work with regulated medical devices and those that “transform a mobile platform into a regulated medical device”6—but many app makers want to steer far from oversight by sticking with simple tasks, like converting data to bar charts. All that said, a few companies have taken the plunge and subjected themselves to FDA regulation by developing sophisticated apps that have proved their ability to make diabetics healthier.7

A company called mySugr offers a diabetes management app that, while still requiring data entry and producing pretty graphs, can also analyze what users tell it about food, blood readings, and other factors to help identify how medication affects users and what patterns their behaviors follow.8 The company’s software, which was originally aimed primarily at people with type 1 diabetes mellitus, also tries to motivate users with frequent challenges and by personifying their disease as a visible “monster” that can be tamed by judicious behavior.

“This is a product made by people with diabetes, for people with diabetes. Our founders had tried all the other apps before they starting building this. They knew what diabetics wanted and where software typically tended to fall short and often still does today,” said Kyle J. Rose, mySugr’s business development director, in a conversation with Evidence-Based Diabetes Management. “Programs like this need significant information to provide significant feedback, so the program tries to make input as easy as possible with passive data collection, like geo-tagging, for example, and it tries to make the process fun by turning it into a game. You get points every time you enter information, and those points help you tame your monster.”

The mySugr app called “Companion” offers some pretty sophisticated services. It can, for example, “remember” venues where users have reported experiencing health problems (say, hyperglycemia at an ice cream shop) and thus inspire the user to choose a different treat, choose a different dose of medication, or perhaps even avoid the location altogether. A simple version of the app for either iOS or Android is available for free, but the “pro” version costs $19.99.

Another app the FDA has approved, this one specifically for managing type 2 diabetes mellitus (T2DM), costs considerably more. WellDoc’s BlueStar system retails for roughly $100 a month, but it’s prescribed by a doctor and reimbursed by insurance, just like other prescription-only medical health devices. The software tool, which reduced glycated hemoglobin levels by 2 points in randomized controlled trials, utilizes the information provided by each user to learn about their body and make customized, clinically validated recommendations.9

BlueStar, like many of the most sophisticated medical “apps,” isn’t really an app at all. Only a tiny portion of it resides on user phones and computers. Most of the product resides in very large servers that can do complex analysis, and other portions reside in the electronic health record (EHR) systems that physicians use. Once a doctor prescribes BlueStar, the patient’s EHR, treatment plan, and physician’s orders are loaded into the program to custom-tailor the algorithms for each patient. Then, in most cases, a WellDoc trainer visits the new patient to help install BlueStar on smartphones, tablets, and/or computers and then explain the basics of using the system. If questions arise afterward, patients (and doctors) can call for live help.

While BlueStar allows patients to enter endless amounts of information about themselves, it also works for patients who only provide data a few times a week. “Most people simply won’t put in several hours of work per week on this stuff, but fortunately, they don’t have to,” Chris Bergstrom, chief strategy and commercial officer at WellDoc, told EBDM. “BlueStar will give you valuable feedback every time you engage it, whether it’s twice a week or 6 times a day. You’ll get more out if you put more in, but everyone gets actionable information, in the right form, at the right time.”

Fortunately, entering information into programs like BlueStar and mySugr Companion may soon become much easier, and Dexcom’s recent announcement illustrates why. Wireless communications technology has improved to the point where many medical device makers have decided to make products that can automatically share information with tablets, smartphones, and other devices. The process is under way at every type of company, making every type of device, to measure every type of variable: weight, body fat, blood pressure, heart rate, cholesterol, blood alcohol, blood sugar, exercise duration, calorie expenditure, and countless others.10

Improved hardware has certainly helped drive the trend. The components required for Wi-Fi, Bluetooth, and cellular communications have all gotten better, cheaper, and smaller over the past few years. Better software, however, is playing an even bigger role. Until recently, many efforts to sync any 2 products from any 2 companies required custom software. Some companies did form partnerships. mySugr, for example, has teamed with Sanofi on software that lets its companion app upload information directly from iBGStar glucose meters.11 But the sheer amount of effort limited the number of products that communicated. That changed this summer when Apple released Health-Kit, a program that creates formatting standards for all kinds of health data

and acts as a repository for properly formatted data—extracting it from any device that provides it and transmitting it to any program that accepts it (under the control of the user, of course).

Dexcom’s new Share device, a cradle that holds the handset from its G4 Platinum continuous glucose monitoring system, uses Bluetooth to send information from the cradle to an iPhone or iPod Touch, which sends the information on to the Cloud. From there, up to 5 “followers” can look at the data and set up the system to send them notifications. The Share system will use Health-Kit to make the patient’s glucose data available, under the patient’s control, to other apps on the patient’s iPhone. The same functionality will be built directly into Dexcom’s fifth-generation (G5) continuous monitor, which will still offer a Dexcom handset to patients who want one, but will work directly with Apple, Android, and possibly other mobile systems for other users.

“This technology offers a lot of significant advantages. Users have one less screen to carry around. They can monitor glucose privately on a standard phone rather than an unusual device that people ask about. Their loved ones can get automatic alerts about potential problems. Their doctors can get information easier. And, of course, patients can transfer information from their monitor to any compatible software they use,” Jorge Valdes, Dexcom’s chief technical officer, told EBDM.

Dexcom is also working with researchers at Stanford University on a related trial, one that explores the effect of taking information from the company’s products and transferring it via Health-Kit directly into the most popular EHR systems. “We’ve always believed that the information from our devices could produce value, but have chosen to spend our limited resources improving our actual devices to produce customized software for every potential partner or application,” Valdes said. “Having standards that work with a lot of products makes it a manageable task and allows the information from our devices to flow pretty much wherever it will do the most good.”

Valdes isn’t the only one who expects standards, like the ones from HealthKit, to precipitate dramatic new uses for underused medical information. There are, however, obstacles that may hinder such efforts. First, technology may prove unable to automate the collection of some very important types of data, like blood sugar and food nutrition.

Apple, to use a big example, generated much excitement with several moves that suggested it might include a monitor that measured blood sugar through the skin on its new watch.12 Unfortunately, barring any last-minute announcements, the first generation Apple Watch won’t have such a monitor, presumably because the technology isn’t ready for prime time.13

Food is almost as big a sticking point. It’s easy to track nutrition information for home-cooked meals and to scan bar codes on packaged foods and even to use guides for chain-restaurant meals. But although researchers are looking for solutions,14 there’s no way to snap a picture of a meal at a friend’s house or the little restaurant on the corner and get reliable information about carbohydrates and sugar.

Second, even if such things existed, they might do little good for most diabetics, who are disproportionately old, disproportionately poor, disproportionately afflicted with mental health issues, and, thus, disproportionately unlikely to have an app-capable cell phone, let alone a collection of cutting-edge devices.15,16 Third, lack of information isn’t really the big problem for most people with chronic ailments. Nearly everyone with T2DM realizes they could minimize their health problems by eating a low-carb diet, exercising, and maintaining a healthy weight.

“I believe that information technology can produce significant benefits to human health, but I also believe the marketing hype has created unrealistic expectations,” said Kerr, the research director at Sansum Diabetes Center. “Linking a scale and a treadmill to an iPhone and using the data to make a bunch of those bar charts won’t transform lives. But the potential is there. Providing clear

information, when it’s most important, and providing it in a way that maximizes the motivation for action, can make people healthier. It’s just going to take us longer than a lot of people expect to put it all together, and the answer will hinge more on the psychology of motivating people than on simply providing easy information.”References

1. FDA approves Dexcom SHARE, the first remote mobile communications device used for continuous glucose monitoring (CGM) [press release]. San Diego, CA: Business Wire website; October 20, 2014. http://www.businesswire.com/ news/home/20141020005879/en/FDA-Approves-Dexcom-SHARE%E2%84%A2-Remote-Mobile-Communications#.VE5oCaDF9IR.

2. Marsden R. Could our smartphones soon be diagnosing diseases via HealthFit and Google Fit? The Independent. http://www.independent.co.uk/life-style/gadgets-and-tech/features/

could-our-smartphones-soon-be-diagnosing-diseases-via-health-kit-and-google-fit-9638755.html. Published July 30, 2014. Accessed October 28, 2014.

3. Farr C. Health developers, doctors want to see more from Apple’s watch. Reuters. http://www.reuters.com/article/2014/09/11/us-applelaunch-health-idUSKBN0H62IA20140911. Published

September 11, 2014. Accessed October 27, 2014.

4. Belluz J. The Apple Watch won’t make you healthier. Vox. http://www.vox.com/2014/9/10/6128193/will-the-new-applewatch-improve-your-health-wearable-technology.Published September 10, 2014. Accessed October 27, 2014.

5. By 2018, 24 million people will use diabetes mHealth apps, report says. Nunviun Digital Health. http://nuviun.com/content/news/By-2018-24-Million-People-Will-Use-DiabetesmHealth-

Apps-Report-Says. Published March 26, 2014. Accessed October 27, 2014.

6. FDA website. Mobile Medical Applications. http://www.fda.gov/MedicalDevices/ProductsandMedicalProcedures/ConnectedHealth/MobileMedicalApplications/ucm255978.htm. Updated June 4, 2014. Accessed October 28, 2014.

7. Wood J. Mobile medical apps: which ones are regulated? Digital Pharma blog. http://digitalblog. exlpharma.com/2014/02/mobile-medical-apps-which-ones-are-regulated/. Published February 28, 2014. Accessed October 27, 2014.

8. Companion App. MySugr. http://mysugr. com/companion/. Accessed October 28, 2014.

9. Bluestar website. https://www.bluestardiabetes.com/. Accessed October 28, 2014.

10. Pai A. 23 health and wellness apps that connect to Apple’s HealthKit. MobiHealth-News website. http://mobihealthnews. com/36870/23-health-and-wellness-apps-thatconnect-

to-apples-healthkit/. Published September 30, 2014. Accessed October 28, 2014.

11. Johnson S. Bring it your way? MySugr. http://mysugr.com/blog/mysugr-ibgstar-integration/. Published April 9, 2014. Accessed October 27, 2014.

12. Apple creating an iWatch that could monitor glucose. DiabetesMine website. http://www.diabetesmine.com/2013/08/apple-creatingan-iwatch-that-could-monitor-glucose.html.Published August 5, 2013. Accessed October 27, 2014.

13. Isaacs L. What happened to the GlucoWatch Biographer? Diabetes Monitor website. http://www.diabetesmonitor.com/glucose-meters/what-happened-to-the-glucowatch.htm. Published

January 3, 2012. Accessed October 28, 2014.

14. Lynch TW. Can your smartphones help you count calories? Reviewed.com Refrigerators website.http://refrigerators.reviewed.com/news/can-smartphone-photos-help-you-count-caloriessri-

ceres-app. Published April 16, 2014. Accessed October 28, 2014.

15. Statistics about diabetes. American Diabetes Association website. http://www.diabetes.org/diabetes-basics/statistics/. Updated September 10, 2014. Accessed October 27, 2014.

16. Diabetes and socioeconomic status. The National Bureau of Economic Research website. http://www.nber.org/bah/summer07/w12905. html. Accessed October 27, 2014.