Big Diabetes Data Requires Big Analyses

If you’re anything like me, you’re currently clocking data on your CGM, your insulin pump, your BG meter, and any other number of devices including mobile apps for diabetes, fitness, or menstrual cycle, and wearable fitness devices like Fitbit.

It’s easy to look at one post-meal high and make a judgment call.  But it’s really hard to look at months worth of data and try to pull out patterns to really improve overall BGs and health.  Websites that integrate with my OmniPod and CGM (and are Mac compatible…don’t get me started on this…Gah!) only have the capability to really track BG and carb counts well.  But we all know it’s the type of carbs, not the number of carbs that really matters.  Also, was I especially active on a particular day?  Stressed from a big meeting at work?  Having PMS?  So many variables to consider!

big diabetes dataSince I already have all the BG and insulin data on my pump, meter, and CGM, (that I’ve laboriously collected!) I literally fantasize about just uploading those items to a single program online and then using apps of my choice to input details about my other “life variables,” such as food, exercise and activities, moods, monthly cycle, etc…..and finally (here’s the kicker) have the apps sync their data with the existing pump/meter/CGM data online in the same program!  Perfect!  Easy!  Right?  No way!

As you’re probably well aware, most of our diabetes meters, pumps, and CGMs have proprietary software and/or limited relationships with other diabetes-device companies.  So, based on who manufactured our devices, we’re all pretty much limited to one or two platforms for viewing the data…and those options sadly don’t integrate with apps we’re using to track our food, fitness, etc.

My current work-a-round solution for viewing OmniPod and CGM data (on a Mac) is Diasend, however even here you need a Clinic ID# (or to register as a non-US citizen) or your CGM tab will be grayed out.  I also use the mySugr app for logging (btw…I love mySugar), Google Cal for my monthly cycle and HRT, and just started tracking activity and sleep with a Fitbit Flex.  I make it work but it’s still me piecing together data from four different locations.

fitbit open apiFortunately (and just in the nick of time if you ask me) the US is at the beginning of a wave of personalized, data-centric healthcare, sometimes called the Quantified Self.  A lot of new data collection platforms designed for non-PWDs (like Fitbit and Lose It!) are using open APIs, which means they share and can integrate data.  After years of finding work-a-rounds and “making do” I feel like the current big data trend in healthcare is finally going to make my fantasy a realty…in the very near future.  So, everyone put down your proprietary diabetes devices for a second and raise your glass!  Here’s to hoping!


Super interested in the Quantified-Self movement like me?  Here are a few really cool recent articles:

Will An App A Day Keep The Doctor Away? The Coming Health Revolution (Forbes)

Solving America’s Big Health Care Challenges With Big Data (Huffington Post)

How Patient Generated Data Changes Healthcare (Information Week Healthcare)

7 thoughts on “Big Diabetes Data Requires Big Analyses

    • Hello! Loved your site and your white paper! Just commented on another post recently that it will only take one D vendor to go open API and the rest will follow :) Also loved your “stack” model. Sounds similar in theory to the project I currently work on in genomics where we use an annotation pipeline to essentially “translate” genomic annotations into the same language, enabling researchers to conduct comparative analyses.

  1. Pingback: My Fitbit Flex: A Delightful Surprise | cyberneticdiabeticcyberneticdiabetic

    • Katy, I considered that as well…but couldn’t justify it (especially since I’d have to buy one with an “old” operating system!). I’m trying out virtualbox now, which is free, to run Windows 7 on my Mac. I’ll post how/if it works :)

  2. First off, I think I’ve found my favorite new D-blog. I love your OmniPod man in the header.

    I’m just starting to try to use Diasend w/ my Dexcom G4 and OmniPod & my Mac, and I’ve got Dex uploading w/o a problem, but OmniPod doesn’t seem to be recognized. Any tips you can give for getting the uploader to recognize it?

What do you think?