Tidepool - Acquiring, Storing, and Opening Data for Improved Diabetes Care

Review by John Walsh, PA, CDTC, Ruth Roberts, MA

Who does your diabetes data belong to? You - obviously. You're the one who created it and needs it for diabetes management. Your glucose readings, your insulin doses, your carbs and meals consumed, your exercise events, your stress levels, your devices. How can you take charge of your data and benefit from it?

If you have diabetes and are technologically savvy, and pay attention to the latest devices and features, you have an insulin pump, a CGM and one or more blood glucose meters. You gather tons and tons of useful data. You have all the cords needed to plug into your computer and download your devices into proprietary software to view that device's data. You know how to use each software program to print the charts, graphs, and other useful sheets from each program, and your physician is as savvy as you and likes to analyze all of this information. Does this sound like you?

Not unless you're a rare breed. Downloading devices remains a major impediment for most users. The Helmsly Foundation estimates that only 5% of people with Type 1 diabetes even attempt to access their data. Either the systems are too difficult to use or the software to view the data is poorly designed. As a result, only a small portion of the valuable data on devices gets used to improve control. A major obstacle is that each device has its own proprietary way of doing things, so unless all your devices happen to download to Carelink you cannot get your data from your devices into a single piece of software to really analyze what's going on.

It's even worse at most clinics. Your compliant and dedicated healthcare professional has to accommodate every possible system that the insurance companies and Medicare steer their patients toward. Multiple devices, multiple connections, multiple screens, multiple all becomes very confusing and deters those in the medical field from even attempting to download information. Because of this, your data remains in various data silos hidden or locked away from the person who should really have easy access to it - YOU.

This is where comes in. Started by Steve McCanne and Howard Look, software moguls whose teenage daughters have diabetes, Tidepool aims to make data collection and utilization far easier. Their goal is to create an application infrastructure that allows the collection in the cloud of data from multiple devices and vendors. Data is then pushed toward an opensource application of YOUR choosing on you or your doctor's browser. It can then be viewed and analyzed by you or your healthcare professional.

The goal is to no longer need different software packages or websites to look at your data. All your information will be automatically transported and accessible when and where you want. Ideally, you would simply walk into your house or physician's office or pass by your cell phone, and your devices start sending your latest data to a computer or straight to the cloud with HIPAA compliance fully ensured. Your data and therapy recommendations could even be entered directly into your physician's electronic medical record. The opensource community has the option to collect data from different devices like brand X pump and brand Y CGM and display both on the same cell phone screen. You, your family, and your doctor can see everything that is going on in real time.

Getting the Data

In a perfect world, all the vendors would gather your information in the same exact way and let you easily access your data. While this is possible, it will probably not happen until a proven system is developed by one company and put in place. Once the system is available and shown to work well, other device and software companies have the option to jump on board.

The data from each device would be moved to a cloud storage system from where it can be accessed from a desktop or smartphone browser. Because the Tidepool system has already translated the proprietary data and reformatted it, any opensource software system can access the data in a familiar way using helpful application programming interfaces or APIs. The system has different "clouds" for each vendor device but as long as the same open APIs are used, accessibility will not be an issue.

After data like carb counts, insulin doses, BG levels from meters and CGMs, and activity levels become accessible, an app similar to the early version of Tidepool's "blip" take over. Blip is described as "a single-page application that runs in a web
browser, provides easy-to-use yet powerful visualizations of diabetes and activity data, and anchors an on-line conversation using social networking so the PWD, their family, and their care providers can all interact online with low effort and high efficacy, capturing the "learning moments" of diabetes management when and as they occur." Take a look at the Blip intro video

In the early stages, apps will be able to give you any of the collected information and provide graphs or visual connections between the devices and activities. In addition, automatic analysis and suggestions from the apps would be great once FDA hurdles are addressed to allow this functionality.

Is Data Ownership Legislation Needed?

Federal legislation might speed up this entire process. Consider that you and your insurance (also bought by you) buy the devices you use, you collect your data (and hopefully download it), yet you usually can't access your raw data. And despite numerous discussions between diabetes device manufacturers and software companies over many years, no agreement has occurred to set up a unified diabetes database.

To correct this logjam, a legislative bill might simply require:

  1. that ALL diabetes medical data belongs to the individual who generates this data, and
  2. that this personal data must be readily and rapidly accessible to its owner at all times in a useable format from an internet site.

Some additional costs will be incurred in making data more accessible, but consider all the clinical time wasted during downloads or attempted downloads by medical assistants, nurses, and physicians, plus all the additional costs generated as a consequence of uncontrolled or erratic glucose levels when millions of peoples data is never downloaded or looked at.

Open Source

Open Source development means that Tidepool will not keep its code or procedures hidden from the development community. This availability allows other people (coders, designers, enthusiast) outside of their organization to participate in the development and improvement of the system. Open source development has led to the creation of some great software (think Thunderbird, Firefox, OpenOffice), and the diabetes community will likely benefit from this type of developmental infusion.

Where it's at now

Steve and Howard already have some great applications started. Quick demos have shown how information from multiple devices can be presented and accessed on one screen. Another application, named Nutshell, allows you to keep information about a meal, including the food eaten and the subsequent impact it has on your glucose. This information, tagged appropriately, can later be accessed the next time you eat the same meal so you know exactly how to handle it. Simple, intuitive applications like these are the key to getting more users to actually analyze their data.

Companies Currently on board:

  • Asante, makers of the Snap Insulin Pump, was the first insulin pump company to open their protocols up to the Tidepool system.
  • Dexcom became the first continouous glucose monitor company to work with the Tidepool system.
  • Medtronic allows Tidepool to gather data from Carelink System.
  • Animas has allowed access to the data in Vibe and Ping insulin pumps via the Animas IR Cable and the Tidepool Uploader