Pixels: Your Personal Diabetes Big Picture

Over the last several weeks my husband and I spent our evenings working on a proposal (below) we submitted to the Target Simplicity Challenge.  The whole point of our idea is to simplify and customize diabetes analysis and management for patients and doctors.  Check it out!  Let us know if you’d use a site like this, or what features you’d like to see to manage your diabetes.

Pixels

This is an initial mock-up to help convey the idea. Click for a larger pdf version.

 

PROBLEM:  Diabetes management requires continuous collection, tracking, and analysis of vast amounts of data, however current digital interfaces for diabetes data have four crucial faults.

1)   Web-based programs that pull data directly off of diabetes devices such as glucometers, insulin pumps, and continuous glucose monitors are often designed by the device manufactures and are largely proprietary.  Therefore data is siloed, increasing both access and analysis difficulty.  For healthcare professionals, this means patients come in with data either manually entered on paper, printed from the multitude of proprietary web-based platforms, or still stored on the diabetes device and requiring download assistance.

2)   They usually track only two or three variables, medication (e.g. insulin), blood glucose (BG) levels, and sometimes carbohydrate intake.  This limited number of variables is grossly lacking in context for treating individual patients, thus lending itself only to very generalized treatment recommendations…a one-size-fits-all approach.  In reality many other factors play a major role in optimizing control such as monthly cycle, fat and protein content, glycemic index, physical activity, remembering medication, stress level, etc.

3)   They simply display the data but offer very little or nothing in the way of analysis assistance.  While graphs are helpful, they are insufficient to continuously analyze long-term patterns across multiple variables.

4)   They lack any means of patient engagement and empowerment.  Data entry is time consuming and often complicated.  Analysis is focused on patient error rather than rewards.

PROPOSAL:  In order to optimize diabetes management, we need to simplify and unify diabetes data analysis, while empowering and engaging patients.  Start by creating a single platform for integration of data from multiple diabetes devices. Then leverage the current Quantified Self, and Lifelogging trends to incorporate important variables via smartphone apps and wearable fitness devices that patients already use (many of which utilize an open API), allowing patients to create a custom diabetes management mashup of apps displayed alongside diabetes device data.  The individually tailored scope will have broad patient and doctor appeal, e.g., one patient may focus on testing BG 4 times a day, need a medication reminder, and passively track walking with a pedometer, while another may test BG 10-15 times a day, use a CGM, avidly track running, nutritional information, geolocation, mood, sleep patterns, and monthly cycle.

Enable doctor and patient data mining for potential meaningful and actionable correlations by applying prevailing 3D data visualization and comparative analysis techniques and algorithms (already in practice to mine big genomics data) e.g., compare several data sets to come up with key talking points for limited doctor-patient interface time such as, “75% of your nighttime lows occur during the week following your period in your menstrual cycle” or “On days when you walk after lunch, your BG is an average of 20% lower than on days when you don’t” or use a passive geolocation app to see what places you frequent when your sugars are higher/lower.

Finally, engage users in multiple ways.  The mashup format provides individual context for the data and will enable patients to set small, attainable goals that make sense for their lifestyle.  Incentivize both with rewards and positive feedback, e.g provide coupons and create a gamification aspect where you receive Target rewards as positive reinforcement, or earn badges for meeting your goals, or receive a text saying “Great job checking your BG 5 times today!”.  Integrate the site with social media, so users can share positive feedback and receive/provide support.   Integrate healthy meal suggestions with Target shopping lists.  Provide a Target kiosk to assist patients with uploading device data or setting up an account, etc.


Check out these related (less professional and more personal) posts:

My Fitbit Flex: A Delightful Surprise

Big Diabetes Data Requires Big Analysis