High-quality Global Health data for research is more important than ever. But Researchers around the world face many obstacles, especially in resource-constrained settings.
So: We want to make OpenMRS-collected data easier to use for Researchers.
There are two main ways to get involved in this new project:
We want to talk to Researchers!
Are you a Health Researcher, or do you know someone who is?
We want to make sure we understand the experience and needs of Health Researchers!
We are looking to interview: Researchers and Program Decision Makers; e.g. Researchers like Epidemiologists, Health System Strengthening, Public Health, Whole-Health/OneHealth, etc; and Program Decision Makers like Population/Public Health leads.
If you or anyone you know fits that role, please book a 30-minute Zoom call with us at https://om.rs/meetdirectors! We will ask about their experience as a Researcher in general; and, if they have experience with OpenMRS, we can speak about OpenMRS in particular. (OpenMRS-data-experience is nice to have but not required.) These are our Interview Questions.
We would love to hear from you and learn from you. Or, if there are any other researchers you think we should speak with, please let us know!
The Technical Side
The OMOP Common Data Model (CDM)
Specifically, we plan to:
use/map OpenMRS’ custom data model to the OMOP CDM, and then;
set up a re-useable open-source tooling pipeline for data to be extracted for research question use.
This is because the OMOP ecosystem has many health data analysis tools for researchers, but these cannot be leveraged until data is converted into their expected data format.
@ibacher and @jayasanka have begun reviewing the technical considerations for this work and can speak more to their thoughts/findings so far. Full Project details here.
As usual, the OpenMRS community and ecosystem has already been a wealth of knowledge: We have reviewed Talk, Google, and Publication hubs and compiled this summary of Existing Work on OpenMRS ↔ OMOP. But we still have more to learn!
This is such an important opportunity! I am 1000% behind this effort and looking forward to the discussion. The OMOP CDM is another global good and really requires two pieces of ETL to work within the research community (OHDSI). The CDM itself is an information model that tells you where to put each data element from your source data… tells you which bucket to put the diagnoses, the labs, the dates, etc… The second really important requirement for the standardized tools to work with the CDM, is that the data itself needs to be mapped to standard codes. This is letting you know what is exactly IN the bucket. Is the diagnosis on the patient the same MALARIA that is on other patients’ diagnosis lists? The OpenMRS concept prevalence study is designed to help us standardize our data dictionary to work with project like OMOP/OHDSI. The CIEL concept dictionary is already part of OMOP but needs to be updated and enhanced to work with the LMIC use cases… so hope that you will participate in this worthy effort!
Even though you still need mappings done as you mentioned, having these synthetic databases and capacity to generate synthetic data I thought may be useful to start discussions around what sort of analyses or tools you may want to build.
set up a re-useable open-source tooling pipeline for data to be extracted for research question use.
In light of this goal, what sort of software tools are you looking to develop or build? I know within JuliaHealth, we’d be happy to help support where we can – I definitely have some members who’d be interested in helping with research software engineering as well as exploration. If I recall one of our conversations, you all considered automatic report generation. Is that still of interest?
Additionally, following up on @akanter 's points, have you been thinking of what the ETL pipeline could look like for OpenMRS? Not only in consideration of mapping vocabularies but also to convert the data over into the OMOP CDM? I’d have to poke around OHDSI tools at the moment but last I recall, there might be an opportunity here for a lot of innovation about building mapping/translation tools.
Social Side Thoughts
I’ll definitely share the goal of getting some more conversations started around my networks. Do you have any other social media posts (i.e. BlueSky, Twitter, LinkedIn, etc.) I could use to share?
Awesome thoughts from @thecedarprince and HUGE thank you to @johnblack for a very informative interview together today!
There’s so much to say here but just really quickly wanted to mention: For anyone wanting to follow the Engineering/Tech side of things, @jayasanka has done a great job summarizing the work thus far here: https://openmrs.atlassian.net/wiki/x/AQDDFQ
Thanks @johnblack . I have booked an appointment using the link above. @ronaldcornet is much more of an expert in OMOP; I am hoping someone from his team will join this effort, too. I’d be happy to help where I can.
and HUGE thank you to 