We Are Not Waiting: CGM in the Cloud (Part 1).
Waiting, when it comes to diabetes, frustrates the hell out of me. According to the “cure in five years!” mantra that rang out constantly when I was diagnosed in 1986, I’ve been waiting for a cure for almost 30 years. Currently, I wait (impatiently) for the Animas Vibe to become available to US patients. I wait for the Dexcom Share application to become available. I wait for doctors to call me in for appointments and on the phone with mail order pharmacies and on and on … lots of waiting.
I hate waiting.
But people aren’t waiting anymore. There’s a whole movement in the diabetes community embracing that very concept. And today (and tomorrow) I’ll be taking an in-depth look at how amazing people in this community are taking their diabetes data into their own hands. We are not waiting, indeed!
Today, John Costik, one of the founding members of the CGM in the Cloud Facebook group, an engineer, and diabetes dad to Evan, talks with me about the #wearenotwaiting movement and how he was inspired to make CGM data bend to his needs. (This is a long post, but his perspectives are awesome.)
Kerri: What is your connection to diabetes? And can you tell me about the CGM in the Cloud group, with some background on the We Are Not Waiting movement?
John Costik: When Evan was diagnosed, it felt like the floor to the lives we knew had vanished. The grief, anxiety and denial were all very real, and once we had our hospital training and sent on our way, the reality of it all sank in. Type one management is hard! But my wife, Laura and I are both engineers; we saw, almost immediately, that processes can be improved, data can be collected and analyzed. If we could make life (even just a little) more like life without T1D, we had to try. After researching pumps and CGMs around Christmas of that year, we decided we wanted to start on the [Dexcom] G4 as soon as possible. The DOC was very helpful – your first and second impressions posts, along with other reviews, made it pretty clear that this was an amazing device.
We danced with the insurance company, fortunately not for too long, and we had Evan’s CGM in hand in late February of 2013. I decided to start tweeting about it – it was a very real shift to a more optimistic attitude in all of us.
— John Costik (@jcostik) February 23, 2013
I was hooked, and the anxious father in me never wanted those numbers and trends to be more than a glance a way. Less than a week after getting the G4, using files provided with Dexcom Studio (libraries), we had a windows laptop pulling the data and sending to a simple Google doc for Laura and me to pull up throughout the day. I built a simple trend and number app for our iPhones that pulled this data – sitting in a dock at work, Evan’s BGs were a glance away.
— John Costik (@jcostik) April 2, 2013
I also wanted to cover the “gaps” in data – recess, walks around the ponds at daycare; arguably the riskiest times for a low to creep up on Evan. I then began to look into using a cellphone to read the G4, and send the data to a cloud service that any number of devices could pull from. I took family medical leave starting in April 2013: Evan’s honeymoon was ending, pump therapy was beginning, and Kindergarten prep wasn’t going to be as simple as we’d expected (plus, Laura and I were exhausted). I set aside a few weeks in May to see if I could make the cellphone chat with the G4. Fortunately, I didn’t let my technical shortcomings keep me from trying, and I figured it out pretty quickly. An Android app that can read G4 data became a thing!
— John Costik (@jcostik) May 13, 2013
— John Costik (@jcostik) May 14, 2013
This led up to an email from the wonderful Lane Desborough. We got talking, I shared my simple Windows uploader with him, and he began working on “Nightscout” – a glance-able BG chart that could be viewed throughout a home. After several months of testing, improving the Android app to work more reliably, I shared the code – Lane continued to develop Nightscout, with assistance from Ross Naylor – I leveraged the chart code in our own “Care Portal” and grabbed a Pebble watch to play with.
— John Costik (@jcostik) July 16, 2013
We continued to work on our apps and tools, more folks on twitter began to notice – and it became very clear that people really wanted this tool. It was life changing for us, and other parent’s immediately saw the value, the hope for a less complex, safer, healthier life it can bring. Freedom!
The “CGM in the Cloud” facebook group started from these early twitter interactions and helping just a few other d-parents looking for a better way. We have wonderful tools, but they can do more! Jason Calabrese, Jason Adams, and Toby Canning deserve the credit for scaling this system.
Jason Adams started the group, because he knew that we were not the only ones tired of waiting.
A simple tweet that started something bigger:
— Jason Calabrese (@jasoncalabrese) August 18, 2013
#wearenotwaiting was coined by Howard Look, d-parent and CEO of non-profit start-up Tidepool. It was the call to arms for the first “d-data exchange” hosted by Tidepool and DiabetesMine, just prior to their Innovation Summit in November 2013.
— Tidepool.org (@Tidepool_org) November 21, 2013
As a movement, it is all about doing more & not waiting for:
- Anyone else to step up and change the standard of diabetes care.
- To cut through the old proprietary systems of big medical.
- To take ownership of our data.
- To use the combined data (BGs, Nutrition, Insulin, and Biometrics) to unravel the unique mystery that
- everyone’s T1d is.
- To go out on a date with a spouse, without T1d’s shadow tagging along.
- So many things!!
— John Costik (@jcostik) May 14, 2014
Kerri: What kinds of discussions have you seen taking place in the group?
JC: Wonderful Testimonials! I can’t express how amazing it is to see so many people “taking back” from T1D. Parents out on dates, children riding bikes, going on sleepovers – these events were either put off or filled with anxiety and fear over the constant “???” Removing the mystery makes it possible. It makes averting dangerous lows possible… To see others experience what we experienced, it is still overwhelming.
HELP! And lots of it freely given… New folks learn, and teach others, the combined learning of hundreds, now (potentially) 1,500+ people. How do you pack this setup together? How does your child carry it? What phone works best? What cables do I need? How do I compile the code? I see the full spectrum of technical prowess in the group members, but no one should ever (and I hope hasn’t) feel that there are any questions off limits. I know, as I’m typing this, that there are at least 20 conversations going on covering how to install the pebble watch to the best Nintendo 3DS case to stuff this setup into. The future! Don’t like ???’s We can see through them – Want less lag in your CGM data? That’s coming.
— John Costik (@jcostik) June 21, 2014
The discussions in the group also point out the very real shortcomings of a DIY system – all of a sudden, instead of just worrying about a sensor problem or being out of range – you have cell service drops, weak wifi, bad cables, phone battery life, all these new points of failure… but we work through them, because not one or all of those new concerns can topple the improvement in life the “CGM in the Cloud” brings.
Kerri: How is this group moving current diabetes technology into tomorrow’s tech space?
JC: CGM in the Cloud gives us a look at the future of connected devices – a space that consumer products are starting to fill, but medical devices lag. It’s understandable: regulatory delays, walled-off device ecosystems that, by design, keep you tied to a single device maker, and device hardware focus (not the integration or software) induce this lag (among many other things). CGM in the Cloud bridges this gulf in time – simply knowing how to “talk” to a device and get the same data we see on a receiver screen, that’s all we need to get started. The remaining technology is there to be bolted on and consume the available data, to display it in such a way that unobtrusively integrates it back into our lives. What works for me may not work for someone else, but that’s fine – there’s no limit to what or how we use or access the data.
Light bulbs that can turn themselves on, change color, audio systems that can wake up the neighborhood, an app that can call someone when conditions merit. Standards, and organizations like Tidepool will make this fly – I would much rather read a standards document than decode byte arrays. I would love a single platform or application that lets me talk to all of my medical devices – not three separate and poorly designed apps that I hate so much I never use anyway…
Publishing protocols and using standards will benefit device makers – I wonder how many new G4 Platinum systems have been sold because CGM in the Cloud exists? I know people that were on the fence between Enlite & G4 – and CGM in the Cloud was the deciding factor (I know people that have switched out right – Insulin suspend or Remote monitoring … remote monitoring was a bigger benefit to them)
Kerri: Is this group only for Dexcom G4 users? Or is someone dabbling in Medtronic/Abbot space?
JC: I “hacked” the G4 because it is the best CGM product for us, available to us – d-parents, PWDs – anyone that uses a medical device to keep them alive & well should be able to use device that suits them best. The company name on the device should be irrelevant – I have zero brand loyalty when it comes to CGM, BGM, and pumps – I will use the best product available, and that’s it.
If the Enlite is the best product for someone, and they want to use the CGM in the Cloud tools, they can – we just need to know how to get the data from a 530g to the cloud, preferably wirelessly, and the rest is ready to go. I actually think Medtronic would be doing themselves a favor by letting us know how to read the data in this way. As I mentioned, I know people that have switched or picked the G4 because they can see the data virtually anywhere in the world.
I hope these companies understand that something as “simple” as internet accessible, real-time CGM data makes a big difference for a lot of people.
Kerri: Do you fear the FDA? The CGM companies? Anything? (Or is part of the movement to also #techwithoutfear?)
JC: I don’t fear the FDA or CGM companies … anymore. My biggest concern, and why it took nearly a year to make the code open source & available on github, is the fear of litigation from individuals. Covering those bases was extremely important. Doing all this work to make our (my family) lives better, only to risk financial ruin would’ve been, well, awful. Licensing, disclaimers, LLC’s – they can cover you pretty well, but it’s the larger CGM in the Cloud community that offers the most help. The programs that CGM in the Clouds use are part of an open source repository owned, not by me or anyone else, but to the group – it’s open to any and all contributors. Ben West has taken on curating the group code, and that code is as much yours as it is mine.
The FDA is coming around – the group will “pre-submit” to the FDA at some point, and if their new guidance is any indication, we could eventually see a “CGM Uploader” app in the “Google Play” store.
In fact, I have several apps and tools that remain private, and these regulatory changes would make sharing them a no-brainer.
Dexcom has been largely silent, but always aware of what we’re up to – I see very little reason to be afraid of them. Every interaction with them has been great. I hope they like what they see!
Kerri: I’ve downloaded the “CGM in the Cloud” high level set up. How easy is it for members to get one another suited up and running on a remote device?
JC: Pretty easy! From phone advice to actual setup help (Rajat Gupta is amazing, I think he’s helped over 60-70 people get it set up), it’s all there, and Facebook group posting style can make it challenging to find an answer – so a new post is always okay.
Laurie Schwartz, Jason Adams, and Jason Calabrese are the group admins, and they’ve done a great job guiding users and staying very current on the posts with their very sage advice. As much as any of us that wrote the original code may have done, they’re the ones that carry it up and on to a level I never would have thought possible. [Editor's note: Look for more from Laurie tomorrow!]
Kerri: John, why is this tech important to you?
JC: It makes life better, it gave Evan a school year with but a single BG below 60. ￼It gives Evan non-diabetic A1c’s – with pizza and cake still on the menu! And hypo and hyper a mere 1% for over a year. (Caveat: Evan eats anything and everything, which allows very good timing and 100% pre-bolus capabilities – solving for specific foods by collecting the data, analyzing and improving the bolus strategies)
It lets Laura and I go on dates, and actually pay attention to something other than diabetes! Evan and Sarah can play for hours outside, without mommy and daddy hovering or interrupting constantly. As I’ve said before, it takes back some of what type 1 took from us that day in August.
Technology is only as good as the good it does for people. If it doesn’t make life better, easier – skip it,find something that does. For us, and our use, it goes beyond merely seeing BGs all the time. The increased awareness, even at its most passive, helps us understand diabetes a little better, and it gives a CWD or PWD a team of people that understand it as well. Our school nurse is amazing, and she used a custom site (which will end up in the open source repositories this summer) to view his BGs, log treatments, and view those treatments on the same Chart. Her own intuition about Evan’s diabetes was key to a virtually hypo-free year.
— John Costik (@jcostik) April 28, 2014
I hope the shared awareness continues to lighten the burden for Evan, and helps prevent burnout as the ￼tasks ramp up. I told him, if we haven’t cured it, I will always be happy to mind his diabetes if he needs a break from it – whether 13 or 53. Technology like this enables me to do just that, without actually impeding or limiting Evan’s ability to enjoy life.
Kerri: And lastly, how can the greater DOC support this movement?
JC: Share use cases – how would you make it better, how could it make your life better? YMDV, and no system will meet all needs all the time, but we can try – the beauty of a crowd-sourced tools like this -someone else probably wants what you want, and if you’re alone, there’s still someone more than willing to help make it happen for you.
The DOC can spread the word, and help build the community. Together we can show the device makers what products we really want – if a group of 1,500 can get some attention – the stronger our voice, the better. It’s hard to find better motivated people than the DOC – what we do to keep ourselves, a child, a spouse, a friend, (and so on) happy and healthy comes from a place of great love. I suspect our best ideas, our biggest innovations come from the heart, and not a desire to make money.