tl;dr: We would like to join forces with UCSF & Jembi to maintain a (production-style) integrated reference implementation of OHRI, CDR and MPI, for the fictional country “Nakanda”.
The huge CDC TAP project has 3 main software product outputs:
- O3/OHRI (OpenMRS O3 Squad & UCSF)
- Appropriate MPI (JeMPI/SanteMPI/OpenCR) (Jembi)
- Reference CDR (Jembi)
Each partner tends to demo their technical work in silos: OpenMRS in the O3 refapp; OHRI in the OHRI/UCSF environment; DISI/HIE integrations in the Jembi environment. We’d like our software demos to reflect a closer-knit team vibe, and make it clearer to see how integrations and cross-site-EMR flows could work (or not) in the real world.
After talking about this challenge in Cape Town together 2 weeks ago, @eudson, @rcrichton, @mwaririm & I would like to set up an interesting shared demo environment we can use in our CDC TAP demos: A whole environment for an entirely fake “country” named “Nakanda” (name inspired by the previous CDC HQ Informatics director).
- We’d all use the same environment for demos
- Updates done by one partner can quickly be available to other partners (and if we break something for each other, we’ll know quickly)
- It’s easier to focus on one single “suite” for quality metrics. E.g. Have main workflow smoke tests (lab data sharing, case reports, patient matching, care visits), national-level load testing, and security audits.
@rcrichton & @eudson this is where I need your help. On the O3 side, I’m working w/ OMRS colleagues to get the O3 release pipeline smoothed out.
We need to:
- Deploy into servers in the sandbox - Jembi?
- Setup continuous integration - Jembi?
After this is done we can i.d. smoke tests we want to set up, and start regular demonstrations of the environment.
Thanks, that captures what we discussed well. I’d suggest we use the existing servers that we have already in the sandbox (if that works?) and maybe just work on changing the configuration to be around Nakanda. I believe that OHRI already has access to 2 sandbox servers that they are using and us at Jembi have a QA server that we are using to.
I think from here it might be good to work on the story of Nakanda, e.g.:
- What is the population? - maybe we use something representative of a large country? 100m?
- Do you want to name the two OHRI facilities?
- What is a number of total facilities that make sense? And how many people do they see daily? - this we can use to configure a data rate for test data.
Sounds great to me, and I believe both @eudson and I talked about this specific idea in Cape Town and we were both happy with this plan.
- Population: 100m sounds good to me, as it’s ~median of the population of many countries we’re imagining benefiting from TAP products.
- Naming OHRI Facilities: @eudson & @mwaririm what would you prefer? Clinic 1 and Clinic 2? Or a real-sounding name? I found this map of Marvel’s Wakanda, we could adapt some names with the “N”, like “N’Yan” and “Nashenga”…?
- No. of Facilities: 16,000? No. of Patients/Day at Facilities: average of 200? @wamz what do you think?
@grace 100 M sounds like a big country
i think it might be easier to model Nakanada around a real country:
Population: 40 Million
- Adults aged 15 and over living with HIV 1.4M
- Know their status 1.3M
- On ART 1.19 M
- Women aged 15 and over living with HIV 890,000 (new 21,000 Deaths 9,000 On ART 800,000)
- Men aged 15 and over living with HIV 470,000 (new 8,900 deaths10,000 on ART 330,000)
- Children aged 0 to 14 living with HIV 83,000 (new 5,200 deaths 31,000 on ART 66,000)
- Level 6 - n=10, pts/day: 1000
- Level 5 - n=50, pts/day: 800
- Level 4 - n=300, pts/day: 600
- Level 3 - n=1,800, pts/day: 400
- Level 2 - n=2,800, pts/day: 200
- Level 1 - n=5,000, pts/day: 100
@wamz will update the pts/day estimates. @rcrichton is this level of granularity okay?