Upon learning about the possible future of medicine, including self-tracking and social networks for patients, you might wonder whether these things really work or if they are just nice ideas. Well, now there are at least some indications that they are useful.
The Decision Tree quotes a report from the the Kaiser Permanente Center for Health Research which says that people who kept track of how much food they ate lost twice as much weight as people who didn’t in a study about weight loss. As a press release puts it, “It seems that the simple act of writing down what you eat encourages people to consume fewer calories.” The study was published in the American Journal of Preventive Medicine in August. However, the effect is much more powerful if people exercise and self-track together, in line with ideas about “social contagion” that I have discussed before.
(By the way, the last several posts at The Decision Tree about the Health2.0 conference and the Kaiser Permanente’s HealthCamp “unconference” make for interesting reading about everything that is brewing in this field.)
What about “social medicine”? A couple of weeks ago, Alexandra Carmichael from CureTogether gave a talk at the Mayo Clinic and revealed that the company has achieved its first statistically significant finding based solely on self-reported data from users of their site. According to these “patient-generated data”, people with infertility are twice as likely to have asthma. This has also been found before in controlled clinical studies. It’s a sign that aggregated self-reported data has the potential to uncover many known and unknown correlations.
On the theme of patient data, there is a new piece, Owning Your Health Information – An Inalienable Right by Leslie Saxon, which is worth a read.
And a final reading tip: a Nature opinion piece (with Craig Venter as one of the authors) called An agenda for personalized medicine. It discusses how disease risk assessments obtained from two direct-to-consumer genetic testing companies can vary quite a lot in some cases. The authors suggest a number of “best practices” to improve disease risk predictions.