| OHDSI Home | Forums | Wiki | Github |
HI Everyone,
I’m doing an evaluation of health data models available.
Although I am very impressed by the OMOP data model, I’m curious about good alternatives for claims and clinical data (both commercial & open source).
Can anyone suggest other models available in the market?
Best,
Mauricio A.
There is a data model from mini-Sentinel. It is essentially a subset of the OMOP version (though not specifically designed as such). For example, it does not include costs, and it is less flexible in terms of how it handles “observations” that are not laboratory measurements. It is located here.
Kaiser has their own model, but I don’t know of any public details about it.
I2B2 might have something that I have never looked into carefully. Others know more than I do about other data models, but I thought I would respond in the hope that others might provide additional clarifications. I think @Daniella_Meeker or @schillil might have done some detailed investigations and published them.
Hi @rmalarc :
There’s a few papers out there that compare alternative models for observational health data, some of which have been led by members of the OHDSI community.
@schillil and @mgkahn published a paper that compared OMOP CDM with i2b2 and VDW/MiniSentinel: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3824370/
@Daniella_Meeker was involved in another paper that compared OMOP/i2b2/VDW for comparative effectiveness research: http://www.ncbi.nlm.nih.gov/pubmed/23774519
@Vojtech_Huser published a paper looking at these models: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3900207/
Folks that are actively involved in the standards community will also bring up the HL7 RIM model and the CDISC BRIDG modeling effort, but I’m not aware of any groups who have actually implemented these models for use in storage and/or analyses for observational health databases. Mitra Rocca presented on some work she was leading that connected up OMOP CDM with BRIDG, the poster is available here: http://omop.org/sites/default/files/07_Rocca_FDA_Harminization%20of%20the%20OMOP%20CDM%20with%20BRIDG%20Model.pdf
I hope that helps. As you move forward with your comparison and evaluation, I’m sure the OHDSI community would quite like to hear your findings, so please share when you’re done!
Slainte!
Patrick
@Mark_Danese and @Patrick_Ryan.
This is awesome!. Thank you very much for the feedback!.
Best,
Mauricio A.
Probably not useful anymore given the elapsed time but
Are there any active groups that have used OMOP, pcornet CDM and VDW? I understand there is ETL to transform from OMOP to PCORNET – wondering if something exists for VDW as well or what implementers might recommend if an organization was going to implement all of these…??
Thank you,
Beth Lindholm/aka pocketbeagle
We transform from OMOP (previously v4, currently rewriting for v5) to PCORnet, but are not going to VDW. You’re welcome to have at look at our transformation code, which lives at https://github.com/PEDSnet/pedsnetcdm_to_pcornetcdm. Enjoy!
We are working on an intermediate data model for ETL focused on Medicare and similar datasets. Because they are all so similar, we might get some efficiencies this way. It is still very a very new project for us, but when we get it going, we will make it publicly available. It might also work for PCORnet. But right now we have to finish it and see if it actually works first. 
If it does, I will post back here (time scale is probably months though)
We have an opportunity to start down a new path – ideally it would be a common format that could be used for OMOP, VDW, Pcornet. If you were to start from scratch, is there something you wish you had done?
I think the future will be in how we leverage the vocabularies. That is where the power is. The structure doesn’t matter as much. Not that structure doesn’t matter. Just that you can create virtual tables with the vocabularies, if you need them. (I don’t know if this answers your question.)
Thank you! this is very helpful!