Great idea, Maura! I have seen interest in OHDSI from so many different corners that it helps to hear what has brought everyone to the table.
So first, hello to everyone I haven't met. My name is Jon Duke and I am an internal medicine physician and researcher in medical informatics at a not-for-profit healthcare research organization called the Regenstrief Institute in Indianapolis. Note, there is no correct pronunciation for Regenstrief. Long e, short e, hard g, soft g, you can't go wrong. So don't hesitate to come visit for fear of pronunciation challenges.
Regenstrief does a lot of cool stuff that I sadly cannot take credit for-- such as LOINC and OpenMRS. We spend a lot of time thinking about health care data, how to make it better and how to deliver it to healthcare providers in useful ways.
My particular work centers around drug safety, which requires something of a holistic approach. By that I mean, even good evidence usually has minimal impact on provider decision-making due to longstanding challenges of information overload, poor specificity of recommendations, opaque outcome definitions, UI/UX problems, and other stuff that keeps informaticians busy at conferences. I've tried to come at these issues from both the data side (modeling, phenotypes, NLP) and the human side (HCI, provider behavior). Still have a long way to go on both, but OHDSI reflects the amazing potential of diverse talents to tackle big, seemingly intractable problems in healthcare through a strong common foundation.
So yes, I'd be ultimately psyched to see OHDSI be a standard for generating transparent, reproducible, high-quality evidence that actually succeeds in influencing decision-making around drug prescribing. But I also look forward to the many smaller steps along the way that will help people better capture, understand, and apply their healthcare data towards important problems.