In 2005 or so I was approached by a couple of guys in IT working for BellSouth Corp (now AT&T) about collaborating on a redesign of my mobile blood sugar meter data collection device known as 'GlucoMON'. They had been tasked by their CIO, Fran Dramis, to investigate what it would take to move remote health device data onto their DSL network as a means of filling the pipe and better competing with the higher speeds offered by cable networks.
Over the course of about a year we came up with several designs to do just that. Several approaches to integrating meters to the cloud via the DSL router were investigated. We also considered a bluetooth adapter to a PC dongle. When you start working on a specific challenge like boosting traffic and value of the DSL network it's pretty easy to get lost in the technological possibilities.
Finally, we had a candidate to propose for funding and on top of the build estimate we had to add a real-world pilot study. I also proposed the idea of just getting into a pilot study using our existing device which at the time was a cool little number based on a rechargeable battery operated meter jacket with 2-way remote control over the paging network. Swapping out devices later would be a trivial task if the intervention protocol proved to be effective.
"To hell with a new product everyone. We have a huge expenditure for our retirees and employees with diabetes."
Due to the development costs and time delay of building a new device, it was decided that we would follow my suggestion and just get into the field as quickly as possible with our existing wirelessly connected meter case to determine if a DSL focused hardware development was even a viable option. I'm not sure who made the brilliant decision on their side but I suspect it was Fran.
From there, I began work on a detailed protocol which would be called the Active Diabetes Management pilot. My approach was to focus on data driven education for people with type 2 diabetes rather than an emphasis on charts and graphs based on blood sugar numbers. Until then, my focus had been strictly on type 1 diabetes.
As I soon learned, these are two entirely different challenges. What they had in common however was the requirement for dead-simple blood sugar meter data collection.