Wednesday, September 05, 2012

Research idea

I've been thinking for a while about the internet being a vast source of information, but how a lot of the information that I want isn't available. And I keep thinking, what if a hundred or two hundred diabetics decided to get together and submit their own questions and data to each other and monitor it for years and then share it.

I think the project would go like this:
First of all, the people I want data on are people who are or have been on insulin for a significant amount of time. I'm going to say 10% of your life. I also want people who are going to be available for follow up, so people participating have to be volunteering their own data, not their kids'. And they have to commit to continuing to participate, and that means also giving me or somebody else their name to find an obituary if they die. So that's the inclusion criteria.
I then would need a whole lot of people to volunteer, in order to collect much meaningful data. Like I said, I'm thinking at least 100. Each person joining would need to submit their contact information, some statement to the effect that they want to join, and if they want to, would have the option of submitting the question(s) that they want the other people to be asked. The questions should be quantifiable type questions- not, "How did you feel?" but "Have you seen a therapist?" or "Have you been depressed?" not, how does your skin look, but "Do you have scar tissue at your injection sites?"
Then at some point we'd start the study, and probably for anonymity, assign or ask for a code name or number for each participant. Send out all the questions that everybody's submitted. Everybody returns the survey, with somebody badgering them until they do. Publically shared is the percentage answering each question in different ways, and maybe some large analysis; private to the members are any linkages that anybody asks for. For example, a participant might ask, of people who answered X1 to question number 14, how many answered Z1 to question 23? Of people who answered X2 to question number 14, how many answered Z1 to question 23?s
We would send out the same group of questions to the participants every year, except for questions like when did you go on insulin and how old are you. Participants would continue to be able to ask for linkage on past questions. For example, of people who answered Z2 to question 23 in the first year, but answered Z1 later, how many answered question 14 with X1 the first year?

Some questions I'd hope a study like this might answer are: what factors change the risk of lipodystrophy? What factors change the risk of scarring? How prevalent is lipodystrophy?
What questions would you want answered?

No comments: