I’ve been as vocal as anyone complaining about the lack of solid diabetes data reporting in this country, and how that hampers our cause on many levels. Now D-community leader Manny Hernandez and his wonderful crew at the non-profit Diabetes Hands Foundation are taking this problem into their own hands, so to speak. They’ve struck up a partnership to launch a remarkable new reporting application. To explain it all, I’ve asked Manny to join us today:
A guest report by Manny Hernandez
As some of you may have heard, TuDiabetes.org has partnered with Children’s Hospital Boston to develop an innovative new A1C mapping tool called TuAnalyze with support from the US Centers for Disease Control and Prevention (CDC). TuAnalyze was officially launched on the TuDiabetes site last Wednesday.
The application enables members to submit their Hemoglobin A1C data. The information submitted gets displayed in a community map on TuDiabetes, with states lighting up according to the aggregate A1C data once a threshold of participants in each state is reached. As of this writing, only California (verify this before posting) had lit up, but whatever the colors, we hope to light up the entire US map by the end of May!
We plan to explore additional metrics in the future, and move beyond the United States to map data from around the world collected through TuAnalyze.
But, beyond the cool effect: What’s the point of mapping diabetes data?
• In the short term, the application certainly offers you a convenient place to track your own A1C data. Is this something you can do elsewhere? Absolutely. But…
• Once the states start to light up, you can also see how your own numbers stack up against other people entering their A1C in your own state. You can view the total number of entries for the state, plus the average, low and high values and how the data entered are spread throughout the spectrum.
• Also, as pointed out by Ginger Vieira from Diabeteens, “it’s hard to feel alone… when you can look at a map lit up with A1Cs of all kinds and ranges across the entire country!”
In the mid-to-long-term, as we start discovering correlations and learning from the data, there could be valuable things for us all to learn. Just as clinical studies can indicate the connection between the intake of this or that food or medication and changes in one or more biometrics for people with diabetes, we expect similar useful studies to result from the analysis of data collected through TuAnalyze.
We could also identify, for example, trends or a correlation between people’s participation in health-related social networks and their level of diabetes management. This is where the benefits of the application begin to transcend helping individuals into paths that may inform public health endeavors and research.
Are there possible negative implications from self-reported data?
Of course, all data sources have flaws. With TuAnalyze, we seek to complement the strengths of other data sources (CDC, NIH) while supplementing the weaknesses they may have. We also want to learn about participation and selection biases (what makes people be more inclined to enter their diabetes data vs. not doing it?).
We also hope to understand whether and how the research process itself can be accelerated through apps like TuAnalyze, helping reduce costs, complexity and cutting time.
Where does the TuAnalyze data get stored and how is it handled?
• Members of the TuDiabetes social network contribute their data safely and anonymously via TuAnalyze, a highly secure application developed by researchers in the Children’s Hospital Informatics Program and based on the Indivo Personally Controlled Health Record. Indivo is currently in use as a personal health platform by the Children’s Hospital Boston along with the member companies of the Dossia consortium.
• Through your “sharing settings” in the TuAnalyze application you select how much information about your A1C values to share – if any.
• You may choose to have your data used for research purposes, unidentified and anonymous; have your A1C values grouped with the values of other users and made available for academic research, online charts, graphs and maps displayed on TuDiabetes; or make your data visible to whomever can see your TuDiabetes profile page.
Going back to my first conversations about the need for better diabetes data reporting with the Children’s Hospital Boston team in August 2008, I can only be proud of the carefully thought-out path we’ve traveled since then, to bring us to this new tool today. Where will this lead us? I don’t know for sure, but it is my firm belief that TuAnalyze will significantly aid the Diabetic Community to build a shared knowledge that’s bigger than any one of us.
****
Post-Script from the guest author:
I want to thank Amy Tenderich for the opportunity to guest write on her prestigious blog. I also want to thank the Children’s team, in particular Dr. Kenneth D. Mandl, Elissa R. Weitzman and Ben Adida — without their leadership and determination, this project would not been possible.
The problem with this tool is that it’s only going to capture data from computer-savvy people, who are more likely than average to use other tools. A large number of people don’t even know what an A1c is, and those are the ones who are apt to have the very high A1cs. So the numbers on the chart are apt to be lower than they should be.
If labs or health care people put in the data, it might be more accurate. But then there would be questions about privacy.
@Gretchen – That wouldn’t be such a bad start, considering that more & more people are becoming computer-savvy, no…? And the privacy issue is pretty well covered, I think, since you can easily post your data anonymously with this tool.
I have been in IT for 30 years and to sign up was very difficult. No link directly to the sign up I had to look in several places before I dound it. What a way to really dirve PWD’s away.
I agree it’s a start. I just thought people should take it with a grain of salt.
BTW, the security words are so difficult to see it takes two or three tries to get a comment accepted. Sometimes I just give up.
Hi Manny,
This is a great start. One consideration. What is the range for A1c values for a non-diabetic over an entire lifetime. Second, what is the movement from a non-diabetic to a diabetic where type is not the factor. What this does is to allow the ability to compare and contrast a measured value. Thanks for getting this started. Type 1 and Type 2 diabetics present a different factor. The sandard variables are food, exercise and meds. A more direct section for type 1′s is the question of their individual insulin to carbs ratio. AS\s a person with a 1 to 30 ratio it is very difficult for me to hit a “normal range” for A1c range. Hope these thoughts and comments help. As always have a great day.
Dan
Thanks to everyone for all the feedback! All great points.
@Amy: once more thanks for letting me share TuAnalyze on your blog with all your readers!
@Gretchen: you make some good points. I don’t think we can do this through HC practitioners, as you point out, w/o breaching some serious privacy issues. We are relying for now on self-reporting, which presents the issues mentioned on the blog post.
@Russell: would you mind sharing with us at tuanalyze@chip.org any feedback you might have as to how to make the app easier to use/find?
all I can say is: FANTASTIC!!!!
Quest Diagnostics (LabOne) makes use of HbA1c data and aggregates it anonymously. If you pay enough money, they’ll share this information. They use it to track (“anonymously”) people’s HbA1c over time, and use the diagnostic codes indicated by your doctor to relate A1c levels of non-diabetics.
I’m sure Quest would be able to provide scientific data based on state as to average HbA1c’s if they were simply asked. With their millions of data points, they could analyze the average drift as seasons roll through (do A1c levels go up when it’s cold outside?). This kind of health care analytics is already being done, but one must throw money at it to get the data from the company that owns it.
@Jason: it’s not only about sharing information (which I am not sure that Quest is able to do legally, even though they may have the data). It’s about studying the data, which you need researchers for. That is not the business model of Quest, to the best of my understanding.
So it’s not all about throwing money at it. There’s more than data collection being done on TuAnalyze (hence the name).