Question for Implementers: What challenges did you face adapting OpenMRS for COVID-19?

Dear community, We’ve been hearing a range of feedback about teams’ experiences adapting their implementation for COVID-19. We’d love to learn more about the issues you ran into when trying to adapt your EMR on the OpenMRS platform for COVID. If we can understand the problems, this might impact our roadmap, especially as we’re thinking about future pandemic response.

For example, we’ve seen that some teams adapted their forms and were ready to release in under 10 days, while others had bigger work broken down over several sprints. I’m posting here with permission from the PIH team who are interested in sharing what they ran into, and we’d like to hear from others as well.

What were you trying to accomplish, and what problems did you run into?

@ball @mseaton @ddesimone @ibacher @michaelbontyes @suruchi

1 Like

Hi Grace,

A few quick comments based on what I observed, which is by no means exhaustive:

  • Requirements around use of tablets was a big use case, similar to what drove this during Ebola - as tablets were preferred for isolation units. The “recent” investments made to incorporate Bootstrap into the referenceapplication for a more responsive UI paid off here. There was a moderate level of work needed to adapt several aspects of our GSP UI to utilize the responsive features of bootstrap to improve usability on tablets and phones, so this took some extra time, but the patterns established to do this in the refapp made this relatively straightforward.

  • I noticed many implementations trying to get up and running with content by using the Initializer module and openmrs configurations. This seems to have made getting up and running with the right metadata much easier and quicker, as well as more straightforward and transparent for sharing this metadata across institutions. This demonstrates the continued investments we should be making in tools and standards around configuration like this

  • Many of the same lessons and requirements from Ebola came up again and highlighted the need to invest time and resources to harvest and learn from these experiences. Everything from effective tablet-friendly UIs that include order entry to solutions (and integrations) that are already geared up to support the use cases around case finding, contact tracing, monitoring, reporting, etc.



cc @ssmusoke

@mseaton This feedback has totally given me a great start to the second half of 2020, given the 2 year long drive to get Bootstrap into the Reference Application

@mksd just wanted to touch base here, is Initializer able to read files along the classpath so that it can handle metadata loading for custom modules


@ssmusoke are you asking if Iniz can add resources to the classpath?

If so the answer is no, but we could consider adding this feature. Typically this would be useful for packaging HFE forms as a config that would be made of the XMLs and the resources that go with it (CSS, JS files… etc). Is that what you are thinking of? See also this thread: Initializer to load HTML forms?

@mksd I meant reading I am thinking of leveraging it to replace DataExchange module for loading metadata such concepts, attributes, identifiers and other metata in a custom module from the resources folder

@ssmusoke sure, that’s what Iniz was always about :slight_smile:

@mksd earlier Iniz could not read configs packaged in custom modules, needed to be in the OpenMRS application folder

Okay, sorry I misunderstood you. It cannot do that, indeed.

Iniz is about maintaining configurations outside of modules.