Project - Standardized Patient Data

Google Summer of Code
,
#21

Hello @brady.miller Please I need some clarifications. We have always mentioned that having a means to import standardized patient data will help in the Instructional use of OpenEMR and also the use of OpenEMR for research purposes and in the data analytics field

  • For Instructional use I understand how standardized patient data import would be of help
  • For Research purposes, I do not understand how
  • For use in the Data Analytics field, I do not understand how too

Also in what format would imported data be made available for persons in both Research and Data analytics fields. Sorry I don’t have domain knowledge to know how all of these work.

Thanks.

#22

hi @prondubuisi ,

Check out this paper on Synthea to get an idea of importance of standardized patient data in those settings:

Regarding format, synthea provides both CCDA and FHIR formats.

2 Likes
#23

Thanks for the Link @brady.miller. It explains a lot. Also good to know that they have an API. Here is a sample GET request to get Synthetic Data for 100 patients https://syntheticmass.mitre.org/v1/fhir/Patient?_count=100&apikey=cIfiVBuJ4QWwTzrCy3qmTVunsZVd6zRA

1 Like
#24

Hi all,

I got set up with the synthea developer version and generated some synthetic patient data sets in CCDA, fhir v4, and csv.

Currently OpenEMR uses serve_ccda.js which hard codes to look for specific LOINC codes when there’s often a multi-to one mapping.

A good place to start might be to look at the top 2000 used LOINC codes and figure out how to represent them in the OpenEMR database or make changes so that the data/codes don’t get lost in translation.

#25

I’d like to participate in this project…
Any new updates on this?

#26

Hi @Khudhur can you reply with your email address? I can send you some documentation on how to parse fhir r4 files using PHP and look for specific codes. You may wish to familiarize yourself with a few code sets like LOINC, SNOMED, and ICD so you get a sense for how they work.

What complicates this task is that there are often multiple codes that represent similar things. E.g body temperature. Each temperature location has a specific LOINC code. You also need to look at the units of the value and convert if necessary.

I had been looking for a unit conversion php library but didn’t find any particularly active. So that might also be a nice place to start. It would be great if we had a single function that could convert from any given unit to a target unit.

#27

Hi @RachelEllison
Sent you a PM
I’ll try to familiarize myself with these code sets.

#28

(Updated my post after receiving great insight from two considerate community members on another post.)

I’m new here and a couple years late to the party… but came across this thread and am curious if this project is still ongoing or in need of additional support? I saw @brady.miller’s video OpenEMR Easy Docker Development - Create Random Patients - Video 2 - YouTube and seems like that covers many of the objectives and was very easy to follow setting up & implementing on a local OpenEMR instance.

Speaking as an embarrassingly developer, the one missing link to the Synthea video/Docker tutorials was explaining: What’s the easiest way to take the local instance of OpenEMR that you’ve created, place it on AWS via EC2 or Lightsail, and make sure you keep your instance of OpenEMR up-to-date..

Apologies for such a basic question, but I’m missing a necessary step or two from the OpenEMR Installation Guides - OpenEMR Project Wiki instructions to connect the dots… is it possible to use the OpenEMR Express install on AWS and Easy Dockers to generate the Synthea patient data? (And, if so, I’d any assistance pointing me in the direction on how to.)