Hi everyone,
I’m interested in starting an OHDSI regional node in Boston and I’d love advice from others in the community especially from anyone who has helped build local OHDSI communities.
Boston has an unusually dense ecosystem of hospitals, universities, researchers, biotech companies, and people working in life sciences. I think there’s an opportunity to bring some of those communities together around a simple goal:
Generate useful health evidence from observational data and put the results in the public domain.
I’m imagining this as more than a meetup or networking group. Ideally, it would become a local working community that identifies important questions, runs studies through a privacy-preserving federated network, and makes the resulting evidence publicly available in perpetuity.
A few principles I’d like to propose as a starting point:
- Public evidence. The underlying patient-level data remain with participating institutions. Study protocols, code implementing methods, and aggregate results should be openly available to anyone at no cost.
- Patient-centered questions. Patients and patient advocates should have a meaningful voice in deciding what questions are worth answering.
- Sustainable economics. Generating high-quality evidence requires people’s time, infrastructure, and expertise. I’d like us to think seriously about economic models that can sustain this work while keeping the resulting evidence public. I saw how successful a contract system was with data partners in Darwin and I foresee contracts and payment mechanisms being critical for incentivizing high throughput network evidence generation.
- Scientific rigor and reproducibility. We should take advantage of OHDSI’s emphasis on standardized methods, transparent analytics, diagnostics, and large-scale replication across databases. I would love to see Strategus API endpoints where we post our study specifications and results come back a few days later in a highly automated way.
- Open participation. I’d like this to bridge institutions and disciplines: researchers, clinicians, epidemiologists, statisticians, data engineers, software developers, patient advocates, students, biotech/pharma, technology companies, and others who want to contribute. A montly or bi-montly meetup where people share their research findings or tech stack would be fun and a great way to build community.
An early goal could be to get interested people in a room, identify one or two questions worth answering, find at least one data partner willing to run the study and see whether we can take a project all the way from patient/research question → protocol → federated analysis → publicly available evidence.
If that works then we can build up from there.
So a question for anyone reading this:
If you were starting an OHDSI chapter today, what would you do in the first six months and what mistakes would you avoid?
And if you’re in the Boston area and this sounds interesting, please reply or reach out. I’d be happy to organize an initial meeting and will post it on meetup.
tagging @rtmill @kzollove @Daniel_Sozonov (my co-conspirators 
)