Diabetes: A Guide for Designers
I put this guide together because I want to design better solutions for patients and families with diabetes. This is a primer, a starting point for learning more about this disease. It came about because I wanted a central place for providing information to individuals that I work with (students, trainees, designers, and developers) who want to create and design better solutions for individuals with diabetes.
My disclaimer: I am a physician and not a patient with diabetes. If you want to talk with the real experts, talk with individuals who live with the disease on a daily basis. I am thankful for the help of fabulous individuals like @Badshoe who have helped me to curate this site.
Please send me your comments, suggestions for additional slides/links and I will update this guide on a regular basis. If you have specific questions I will try and answer them. You can find me on Twitter at @joyclee or you can reach out by email.
1) What is diabetes, what is the difference between type 1 and type 2 diabetes, what is the overall burden of the disease, and how do you diagnose it?
Check out this slidedeck.
2) How do individuals perform blood glucose monitoring and administer insulin?
Check out this slidedeck.
3) Can I make my designs apply to all individuals with diabetes?
Although they both have the same name, type 1 (T1D) and type 2 diabetes (T2D) could be considered completely different diseases, as they require different treatment paradigms and affect different populations (see this article Beyond 'I'm a Diabetic,' Little Common Ground' from the New York Times for more background).
Also, elevated glucose is common to both, but the nature of glucose excursions is quite different. For example, when you have type 1 diabetes, your glucose can be like a rollercoaster, whereas in type 2 diabetes, blood sugars tend to be more steady. (Obviously these are generalizations to which there are exceptions). As a result, insulin management is a bigger piece of the management goal in type 1. In contrast, lifestyle modification (diet and exercise) and oral medication use are relatively more important for type 2 diabetes.
A word of advice: Define very specifically who your users are (individuals with type 1 vs. type 2) as this will alter quite significantly your design goals. In my opinion, type 1 diabetes management is much more complex.
4) What devices are used in diabetes?
All individuals will use a blood glucose meter which requires you to prick your finger or arm to measure your blood glucose level. Individuals with type 1 diabetes (T1D) check their blood sugars using a blood glucose monitor, on average 5 times a day, although some individuals do it less frequently and many check much more frequently (as much as 10-12 times a day). Approximately 50% of individuals with T1D use an insulin pump, and anywhere from 6-21% of individuals with T1D use a continuous glucose monitoring system (CGMS). These stats are based on data from the T1D Exchange, the largest US registry of adults and children with T1D. FYI, the insulin pump is not a cure for diabetes; it doesn't take away the work of diabetes; in fact it probably increases it. Also, the CGM measures interstitial glucose every 5 minutes which is different than what is measured using a blood glucose meter. CGM is not approved by FDA for making clinical decisions; you are still supposed to check a finger stick before dosing. However some of the CGMs are getting more and more accurate, so although it isn't supposed to replace the need for checking blood glucose, REPLACE-BG is a trial going on to see if you can skip the finger testing. If you are curious, Diabetes Forecast magazine has a nice comprehensive list of devices that you can explore here.
5) What are user tasks associated with Type 1 Diabetes?
If you think about user tasks associated with diabetes, there are too numerous to count. Coffen and Dahlquist describe these tasks, in a paper called Magnitude of Type 1 Diabetes Self-Management in Youth. They performed a task analysis which involved breaking down complex tasks into the component parks of the various domains of diabetes management. They documented over 600 tasks!! They write: "The full regimen task analysis covers 15 domains: Etiology, Pharmacology/Insulin, Insulin Technique, Monitoring, Healthy Eating, Being Active, Hypoglycemia, Hyperglycemia, Ketoacidosis, Stress/Illness, Traveling, Continuous Subcutaneous Insulin Infusion, Complications, General Knowledge, and Miscellaneous." For a complete list of the tasks click here.
Just to give you an example of the enormous number of tasks, Diabetic Danica, who has a whole series of fantastic educational videos, does some calculations about how many times she has checked her blood sugar since her diagnosis in her video "Just Bein' A Diabetes Nerd!"
6) What is it like to live with diabetes?
This piece by Kelly Close gives a really insightful and powerful perspective on the huge number of cognitive tasks that are required with management of T1D. She took part in the Boston University/Massachusetts General Hospital bionic pancreas study for 5 days, which basically took care of the diabetes for her, through an iPhone running a glucose control algorithm interpreting glucose data from a Dexcom G4 CGM, and directing insulin and glucagon dosing through two separate Tandem t:slim pumps. The paper describing the results of this study came out in the New England Journal of Medicine. She writes:
So what was it like to have a machine making all these decisions for me? Well...
- I was never hypoglycemic
- I never felt hypoglycemic
- I was never worrying about hypoglycemia
- I was never recovering from hypoglycemia
That's already enough of a game changer for me. But there was more...
- If I started veering low, my bionic pancreas figured it out and gave me the perfect amount of glucagon to make sure that hypoglycemia didn't occur
- I always felt safe during the week - at no time did I feel threatened or scared
- My glucoses were being watched and stayed perfectly in range overnight, every night. Wow.
- I counted zero carbs
- I never corrected
- I never thought about insulin sensitivity and how I couldn't figure that out
- I never thought about insulin to carb ratios
- I never bolused
- I was a nicer and kinder person the entire week with the bionic pancreas
Because the system uses both insulin and glucagon,* I got to see mini doses of glucagon in action for the first time. There's no question about it - glucagon is a cool hormone. It is ten times better than orange juice, glucose tabs, candy, gel, etc. It is SO precise and the power of having just the right amount was really cool. It was magic!
*Glucagon is a hormone secreted by the pancreas that raises blood glucose levels [which is opposite to insulin which lowers blood glucose levels]. We prescribe glucagon to our patients, so that it can be given in cases of severe hypoglycemia when an individual is unconscious or can't take glucose orally.
Scott also alludes to the burden of diabetes in this post titled "One Does Not Simply".
To learn more, there are a huge number of online resources created by the diabetes community. Check out this handout to get more information about the Diabetes Online Community, which includes the Diabetes Advocates, Diabetes Social Media Advocacy, online diabetes communities and diabetes blogs. Also, Chris, @iam_spartacus on twitter created the hashtag #dayofdiabetes on Twitter, in which individuals live-tweet their diabetes for a day to show what it is like to live with diabetes, and he has a Tumblr blog called Day of Diabetes where he curates the tweets.
7) What are the typical datastreams captured by individuals with diabetes?
Blood glucose, insulin doses, carbohydrate intake for starters. There are a lot of other datastreams (behavioral and contextual) that would likely be useful for informing diabetes management, including mood, physical activity, and location (as a proxy for social environment) but these datastreams are not typically collected or displayed by commercially available displays.
8) So collecting and using and interpreting health data in diabetes is easy right?
Not really, in great part due to the technology. There are a large number of blood glucose meters with their own proprietary software systems, which require users to plug their meters into PC computers, and often the data stays on the hard drive of the computer and doesn't ever reach the cloud. Scott Hanselman summarizes the state of the diabetes technology landscape in this fantastic article, titled, The Sad State of Diabetes Technology in 2012. He highlights major problems in the diabetes landscape, including a lack of data standards, problems with the usability of glucose meters (lack of Bluetooth, lack of information integration when using multiple diabetes devices from different manufacturers, the need for using a cord to download numbers).
A blog article reported on the iBGStar blood glucose meter for the iPhone and asked: "Could this be the beginning of mobile diabetes monitoring?" Scott cites this very insightful response to this question by Lloyd:
"As so many people above have stated, no, you moron. We've been monitoring blood sugar on the go for the past 30 years.
I've got Type 1 diabetes...and my current meter is smaller than the one shown here. I can plug it into my Mac via USB to download and visualize the data (& can control my insulin pump via bluetooth using the meter).
The only benefit with this particular iPhone-compatible meter would be enhanced, immediate visualization of results. Which might be easier to get, and might not, given the inconvenience of having to remove an iPhone case and plug in the meter. (Not to mention other issues - what if my iPhone's batteries are dead? Will it still work?)
Unfortunately, this product reminds me of 5 years ago, when someone would announce a new toaster, and the tech crowd wouldn't be impressed...unless it was a Bluetooth toaster. We're so focused on it being the hot new thing (it's compatible with iOS! Oooh!!) that we ignore the fact that there's nothing revolutionary being presented here.
The way I see it, this doesn't really change anything in terms of treatment. If it's a more accurate meter, great - sell based on that. Not on the bogus "we're taking blood glucose monitoring mobile" claims.
In addition to the technology there are other barriers to blood glucose monitoring, as Bill Polonsky describes in his article, Ten Good Reasons to Hate Blood Glucose Monitoring.
9) So all you need are apps and self-tracking tools to manage diabetes, right? A little bit of quantified self will solve the health outcomes problem in diabetes?
Not quite. We have had tools for self-tracking for a long time in diabetes; the kids I work with would prefer to not to have to quantify themselves on a regular basis. Check out this article by Thomas Goetz. And don't be an "app" hole.
11) Hypoglycemia - A Major Issue in Diabetes
A post about reasons for hypoglycemia and how to address them; what hypoglycemia feels like (low alarms, slammed, blood sugar tanking, down the stairs, clingy hypoglycemia, hypoglycemia hangover, there are days, mistreated carbs); the fear of hypoglycemia, and parental fear of hypoglycemia here and here.
12) Diabetes Management and Burnout:
Read this post about self-management, watch this video about crawling out of the primordial sludge of diabetes burnout, and how to deal with diabetes burnout by AllisonN, Hallie, Will, and Kerri.
13) Diabetes Design:
Amy Tenderich is a pioneer in thinking about design and diabetes, started the DiabetesMine Design Challenge, and in 2012 hosted the DiabetesMine Innovation Summit. Check out this video about a call for an open model of diabetes care, as well as her open letter to Steve Jobs. Here is a website with multiple posts describing the "Dream Diabetes Device".
14) Designing the T1D Care Delivery System
We have also been thinking about the design of the type 1 diabetes care delivery system with our work with colleagues from the @T1DExchange and the @C3N Project. Please view the Storify from design meeting 1 and design meeting 2.
15) Diabetes Software tools of interest that you should check out:
- MySugr. @MHoskins2179 has a great review of MySugr here.
- Tidepool - Fantastic open source platform for giving patients back their own data
- Nightscout Project - For a nice rundown check out this post, Kate Linebaugh wrote the piece in the WSJ about the project and check out the video from the presentation at the Diabetes Technology Meeting in 2014.
16) Diabetes FAQs.
17) My Diabetes Guide for Residents on the Inpatient Service working with New-onset Patients
18) Additional Diabetes Resources