Friday, September 13, 2013

For a diabetic, every step counts

Medtronic, the market leading insulin pump vendor, just published a blog entry in their blog, the Loop. Some text there makes me cautiously optimistic:
Since blood glucose is affected by so many things, we need to take a broader view of the technologies which are on the horizon. Never has it been easier to collect physiological data – perspiration, movement, heartrate, location. These aren’t diabetes devices per se, but they collect critical contextual data which these athletes are already using to improve their performance and glycemic outcomes. We’ve gone from a drought to a deluge of data.
There are four main factors affecting the treatment of a diabetic:
  1. Blood glucose measurements
  2. Carbohydrate intake
  3. Medication
  4. Physical activity
There are other factors as well, such as amount and quality of sleep, stress, other hormones, etc. But the four main factors listed above are the most important ones, and it is essential to understand the effects of each of those to the particular patient. They all affect each other.

So far physical activity has had a minor role in Medtronic's CareLink system, as well as in every other similar system. For many systems, a diabetic can enter any exercise explicitly. I'd do this for each game of football I play or for a session at the gym. But not necessarily when I'm just walking from a place to another. I may not realize it on that day, but on some days I end up being really active, and on other days really inactive. My step count from the pedometer app in my phone for the last week is 2891, 8903, 5814, 4994, 15 909, 8 804, and 3 265. It really makes a big difference whether I walk 16 000 or 3 000 steps per day. And it also affects my glucose levels the next day.



All that data is difficult to enter manually, but really effortlessly collected by an app on my phone. I hope pump system vendors would find a way to include data from various activity monitors into their reports, or allowing the data from their devices to be easily included in reports where all that activity data can be included.

Monday, September 9, 2013

Secrets

I ended one of my earlier posts about competition with a note that I should meet with Mendor, another Finnish startup working to improve the lives of diabetics. So on Friday we had a meeting.

The main goal for the meeting was, at least for me, to find out how we could cooperate instead of competing. I've waited a long time for a chance to work with something I find fulfilling and meaningful, and want to use this opportunity in the best possible way. If someone else is already doing what I'm about to do, I'd rather let them do it and concentrate on something else instead. I've got many, many ideas on how to improve the life of a diabetic. I just need to select the one I think will create the most value, and work on that.

Of course, to find out whether Mendor are planning to do what I'm planning to do, we need to be able to share our plans. But if I have a great idea for a company or for a product, is it wise to share it?

First of all, I would hate to base all my work on just an idea, and the assumption that nobody else has had the same idea. I'll much rather think that our team is exceptionally capable of executing the plan, and will come up with the best implementation of the idea. But the first goal must be to share the work between everyone interested in working in this field, and to collaborate.

Second, you need to think about what a company secret actually is. During the summer, I read Blake Masters' essay versions of his class notes from CS183: Startup, a computer science course in Stanford University, lectured by Paypal founder Peter Thiel. Peter's definition of Secret is that it's the answer to the question What important truth do very few people agree with you on?

So, to find out what Mendor agree and disagree with me on, I think it's best for me at this phase to just tell them what I think and what I'm about to do, as openly as I possibly can. I was very happy to learn that they also discussed their plans quite openly. It must be harder for them in the sense that they already have so much implemented. They are not able to ditch their main idea right away and select something else instead. However, having been around longer and having built something already also gives them an edge, they are not so afraid of competition from some fresh startup.

It turned out we shared many sentiments about diabetics, but are targeting different user groups. They believe their approach with pair measurements is the winning one, and have their hands full of work to implement Mendor Balance around that idea. I agree that pair measurements are an important first step for a majority of diabetics, but know for a fact that for a diabetic like myself that information alone is not sufficient to get the treatment to the level I want to get it.

We agreed with Mendor to keep in touch, and also already tentatively discussed the technical and commercial aspects of my solution connecting to theirs in some point in the future. It's feasible, but first I need to show there is actual demand and use case for my approach.

For me, the simplified version of the important truth very few people agree with me on is that
  1. there is a significant group of diabetics that will benefit from seeing the data of all the aspects affecting their blood glucose levels in a single user interface, and that
  2. collecting all that data can be made easy enough so that these people will actually do it.

Wednesday, September 4, 2013

One step closer to being a company

Today, we started as a team in Protomo, a local startup incubator. Four teams were selected from dozens of applicants, so we consider this to be a great achievement in itself.

We're now a team of three. Mikael Rinnetmäki and Timo Koukkari are both type 1 diabetics and contributing mainly to the technical implementation of the service. Assi Rinnetmäki will handle most of the communications and marketing, as well as managing all the administrative tasks related to actually turning the project into a company.