1 posts tagged “health statistics”

Giving Health Statistics a Good Name

Posted June 29th, 2011 by

Mark Twain popularized the notion that there are three kinds of lies: “lies, damned lies, and statistics.” But there’s a cure for bad statistics, and that cure is openness.

When our study of the effects of lithium in ALS was published in Nature Biotechnology, we not only presented the usual background and analysis, but we also included all the data that were used to support the analysis. (The data were de-identified, of course.)

A Chart from the PatientsLikeMe ALS and Lithium Study Published in Nature Biotechnology

This “data openness” was one aspect of our study – as well as the ensuing media coverage in the Wall Street Journal and other outlets – that caught the eye of folks at the National Institute of Statistical Sciences (NISS). They were in the midst of organizing a workshop on observational medical studies (like our study), and they invited PatientsLikeMe to discuss our lithium research there. As one of the study authors, I went to represent PatientsLikeMe and give a talk entitled “Drug Efficacy in the Wild.” (Click here to download a PowerPoint of my presentation.)

Sitting in a roomful of statisticians for two days might be considered a form of torture by some, but it’s the sort of thing I enjoy as a Research Scientist. I look forward to sharing ideas with my colleagues. Besides, the motto of NISS is “The Statistics Community Serving the Nation,” a theme that dovetails nicely with what we’re trying to do here at PatientsLikeMe. We believe it’s through real-world data and statistics that you can put your experience in context as a patient.

Ultimately, the NISS workshop did not disappoint. I heard some interesting talks, and I learned some new techniques for future work. My own talk was well received, and our approach to the lithium analysis was called “refreshing.” There were a few people there who were not yet aware of PatientsLikeMe, and they were particularly fascinated by this new way in which patients could gather information and work together to learn about treating their illness.

Of course, none of this would be possible without you. Your shared data is what drives our research capabilities and helps us all answer the questions that we have about disease. So, as always, thank you, and keep on sharing!

PatientsLikeMe member tvaughan