Given eAJKD is at the crossroads of nephrology and social media, live blogging from a social media in medicine perspective can be found in the next few posts. Dr. Joel Topf, eAJKD advisory board member will be sharing some key concepts.

Medicine 2.0 conference is currently in its 5th year. The conference began in 2007 in Toronto and this year (2012) it is on the campus of Harvard Medical School in Boston.

In the opening talk Gunther Eysenbach, the organizer, started the conference.  He noted that Boston was a logical location for the conference because it is at the crossroads of medical research and entrepreneurship (birthplace of Facebook). He then introduced the core concepts of Web 2.0 and how they apply to medicine:

        **Then he introduced Jamie Heywood, co-founder of Patients Like Me  who did a great job of laying bare the failures of modern research and clinical practice and offered an entirely novel approach to disease research. He is an MIT trained engineer who became focused on medical research following the diagnosis of ALS in his brother.

      **He started his talk by trying to define health as cohort of vectors that measure various aspects of health. The examples he used were happiness and mobility but imagined that by adding additional variables one could get a fuller picture of a patient’s whole body health.

       **He focused on a few therapies promoted for ALS, lithium and celecoxib. Both therapies were shown to be effective in randomized controlled trials.

          **Drachman et al. showed that Celecoxib was able to delay the progression of ALS in a mouse model. Heywood’s company, ALS Therapy Development institute , however was unable to reproduce these results despite multiple attempts. Heywood focused his attention on the original study’s control group and noted that it did particularly poorly. He came to the conclusion that the positive results were due to the poor survival of the control group.

Heywood’s data was born out when Drachman’s animal study was refuted in a human trial. While most of the editorials after the refutation focused on the failure of the animal model, Heywood felt that the mouse model was vindicated but he questioned how it was being used by scientists. Mainly he was concerned about the small size of the trials.

         **The talk then did a deep dive on how PatientsLikeMe tracked individual patients. Participants filled out frequent reports on how they are doing on a number of different axises of health. Heywood was able to use that data to match people based not only on demographic data but also on aspects of disease severity and progression. He used this matching ability to actually predict the disease course and was able to predict the month of death of an individual with ALS.

          **His group then used the same technique to run a virtual clinical trial on lithium. Prior to running the trial lithium had been shown to be effective for ALS in a randomized control trial. Heywood used the PatientsLikeMe data to predict the disease course of patients prior to taking lithium and watched to see if lithium bent the curve. Lithium failed to do that. Eventually the RCT data came in refuting the original report and lithium was shown to be ineffective . Heywood pointed out that by tracking people and measuring their disease on multiple axis he was able to find out a drug was ineffective earlier and cheaper than the standard model of randomized controlled trials.

         **Heywood concluded by advocating a wholesale change in how we do medical research. He wanted medicine to adopt the physics model with experimentalists and theoreticians. He verbalized a vision of experimentalists doing large trials, with longitudinal data that make all of their data public. Then a second cohort of theorists interpret the data and request future studies.

Joel Topf, MD
eAJKD advisory board member

 Check out all of eAJKD’s coverage of Medicine 2.0 here.