Educational Goals

(Kevin Yeh) #1

In trying to identify next steps for advancing the use of LibreHealth in education, here’s a new thread for discussion.

Rather than focusing on functionality let’s talk about goals/possibilities. Some questions/thoughts for discussion. Who is the audience? There are many different kinds of people with: clinicians such as Doctors, Nurses, Allied Health Professionals and students and instructors for each of these ; non-clinicians, office staff, billing departments. Information Technology Professionals .

What sort of lessons/curriculum need to be taught? Having complete records to look at and examine is certainly a great start as it provides at platform for “self-exploration.” But are there specific concepts that need to be taught?

One random idea for Radiology. There are many case websites that provide images, but these are typically provide in web image formats rather than DICOM so there is loss of fidelity. Also the ability to adjust image parameters (contrast?) isn’t available. I remember different things happening to the image when dragging left and right with the mouse button down vs. dragging up and down. Slice by slice review of CT scans from web pages also seems problematic. I presume radiology departments have was to share interesting cases internally, but could be build a better collaborative educational environment?

Best way to log in to "practice"
(Kevin Yeh) #2

Another possible teaching scenario/use case would be in Clinical Skills Tests (similar to the USMLE Step 2 CS).

Histories for the simulated patients would be in the EHR. Students could then review them before going into the testing rooms and document the findings from their physical exams and interviews electronically.

(Saptarshi Purkayastha) #3

I think students in HI or med school who want to understand EHRs are the audience. Off course, this might be through a tutor/instructor/educational program in/outside universities. @rhoyt mentioned that we could focus on informatics students, but in the future could be medical, nursing or other students too.

I think you highlight an important point. We need to create additional teaching/training material that will guide how the students to understand EHRs using LibreHealth. So, these would be really useful for students to self-teach, as well as instructor guided material.

This needs a real DICOM viewer. Tools that are used by radiologists in actual practice. But we could also have radiology reports that are interpretations of studies (Xray/CT etc) come back to the EHR and those are shown to the clinicians. So the educational EHR could show this workflow, with order (EHR) -> image capture (PACS) -> reading (DICOM viewer) -> interpretation (RIS) -> radiology report (EHR)

NHANES data in LibreEHR
(Bob Hoyt) #4

Thanks for creating this new category “Education Goals” as we now need to migrate towards making this prototype available to instructors/students in the not too distant future. At this point I have two HIM schools and one HI school. Today a professor from Brazil (Renato Sabattini) stated he used OpenEMR and is very interested in our project.

To me, the big picture is to create a platform that could be used for EHR training, health sciences education and research down the road.

I think ( but could be wrong) that the way forward is to enhance 20 patients who represent a decent cross section of patients (age, gender, diseases) and populate their records as much as possible. I thought I might peruse the NHANES data and extract past surgeries and allergies for these patients, but it now occurs to me I don’t know who Jimmy Jennings really is in the NHANES data.

Several people seem willing to develop and share teaching scenarios and in fact in the AMIA Educational Working Group meeting today we mentioned that one of our upcoming webinars should be devoted to sharing best teaching case scenarios and how we might link the educational EHR to simulations.

Personally, I think we can wait on true DICOM images and just begin with images (jpegs) posted under “documents”. I hope we can get some type of grant to take this to a much higher level. Thanks for everyone’s contributions. We have come a long way but have a long way to go. When do we announce the demo and make it available? I now have a list serve of 454 health informatics instructors so they can provide us more input should we need it

(Bob Hoyt) #5

One other thought. Should we get medical librarians or graduate students to search the web and literature to find teaching scenarios for ambulatory EHRs? I suspect there are cases out there we can use. Unfortunately, I am the process of launching the seventh edition of my textbook, so my time is somewhat limited. It is the summer semester so don’t know if there are any available students. Thoughts?

(Judy Gichoya) #6

I will take a stab at doing the literature search and then ask a librarian to look at it – but the more the merrier

On your 3rd edition - Do you ever provide updates to links … Do you think if we cleaned up the content here we could get to a point where we could have a link that says – Additional content available at XXXX with a LibreHealth EHR use case … i know its a long shot but i thought i would put the idea out here … maybe for 4th edition

Kudos @rhoyt on a job well done

(Bob Hoyt) #7

I’m not sure I understood your post. Are you asking if LibreHealth EHR could be in our seventh edition of the textbook? It should be out at the end of 2017 The answer is yes

(Bob Hoyt) #8

A UWF medical library science person is actively researching case scenarios, as well as people in our nursing department. I’ll post what we find.

I will also identify a child with either asthma or type 1 diabetes to make a sample patient. I’ll also identify a pregnant woman and create some type of scenario. I don’t think we can tease out complications, like pre-eclampsia from this data set so let me know if you have any other thoughts about stereotypical patients you would like to use as “sample patients”

(Bob Hoyt) #9

I just completed a new encounter on Andre Cardenas (age 6) who has mild asthma. Easy to add flow rates, CXR and so forth.

In the process of trying to create a sample pregnant woman I ran into the following problem. According to the Demo G CSV file and the RIDEXPRG attribute (pregnancy = 1) we have 28 pregnant women in our data set. The problem is how to identify them in the EHR. It doesn’t appear that the numbering in the CSV file correlates with the PID. Perhaps Kevin has already mentioned how to map patients in the EHR with patients in the CSV data set. I can use age and gender but that is very crude. Help

(Kevin Yeh) #10

The SEQN in the nhanes files maps to the SSN field in the EHR

(Bob Hoyt) #11

Many thanks. I should have known that you had a system to keep everyone straight. That explains why the SEQN is as long as it is

(Bob Hoyt) #12

Progress Report:

  1. Used Sonja Bryan #67504 to create a pre-eclampsia scenario in a pregnant women, using a real encounter I found on the Internet
  2. Used demographics from the SyntheticMass files for patients, so everyone is located around Boston MA
  3. Populated the medical diagnoses drop down list so we now have the top 20 common diseases
  4. Medical librarian is looking for more scenarios we can use to create quality patient scenarios
  5. I found numerous “EHR scavenger hunts” on the Internet we can use to get students started
(Kevin Yeh) #13

In terms of “interesting patients” here’s a list of people with 10 or more different diagnoses. Designing cases around these complex patients might be another way to go.

13 66294 Brad Sharp 13 63765 Ernest Bean 12 68626 Bettie Huang 12 70298 Marsha Haynes 12 69377 Charles Mejia 11 64206 Hazel Holland 11 62474 Jasmine Wilcox 10 69347 Freddie Yang 10 70321 Keith Oconnor 10 67520 Tyrone Castaneda 10 65114 Jane Hunter 10 63082 Kimberly Henry 10 64497 Fernando Whitaker 10 64930 Terry Jones 10 70739 Pedro Mayer 10 68645 Dena Whitehead 10 63801 Brianna Rubio 10 66744 Roy Raymond 10 62934 James Murillo 10 71861 Alana Irwin

(Bob Hoyt) #14

Excellent thought. As an Internal Medicine doc, we live and breathe chronic complicated patients. I’ll take a look and see if I can create some realistic scenarios with these folks.

(Judy Gichoya) #15

Are we making backups of the enhanced cohort of patients cc @r0bby

(Robby O'Connor) #16

Currently, no we are not.

(Kevin Yeh) #17

https://nhanes.librehealth.io/download/nhanesbackup20170616.sql.gz

I don’t think we’re at a point where we need to do automated backups, but here’s a dump with Dr. Hoyt’s additions. That others can also download and test out for themselves if interested.

(Robby O'Connor) #18

I’m doing backups on Rackspace servers currently.

Services hosted on rackspace:

  • Forums
  • Radiology Demo
  • Toolkit Demo
  • PACS server
  • Gitlab CI runner

I’m planning on moving Chat and EHR/NHANES to Rackspace at some point. Moving Chat will require downtime so I can get a consistent database – so that’s on hold right now. Once those are done, we’ll have some backups. Both servers are created.

(Saptarshi Purkayastha) #19

I looked at the few patients who have some data for them and it looks quite genuine. I also like the aggregate information that can be extracted from the NHANES imported cohort. This is excellent work and I would like to congratulate and thank @rhoyt and @yehster for the work. What we need now is some cases…

I also dont quite get the user-interface. I need a tutorial or user documentation to get used to this EMR. Are there resources in OpenEMR or some documentation for the user interface?

(Bob Hoyt) #20

The problem summary list is good but could be better. None of the problems have dates because that was not part of the import. In the case of Bettie Huang I added fake dates to all diagnoses and dosage and frequency to all of her meds. When I get a chance I’ll add surgeries, allergies and immunizations.

I have a friend working on HIM exercises and Kevin and I are definitely working on providing research tools so students can potentially analyze the NHANES data. What is probably missing is exercises/scenarios for health informatics students. In my EHR class I had the grad students create and send HL7 messages, CCDA messages and code and bill a patient so they understood E&M. Clinical students could probably look at these records and pick up the mistakes (missing meds and labs, for example). BTW, I have reviewed all of the complicated patients Kevin gave me and two said they had diabetes and didn’t and one was recorded as being on no meds with multiple problems. One patient only had one lab test done.

Still no word back from Bellevue or UT-Memphis about demo access I gave them. They might provide input regarding exercises

From what I can tell there is no user manual for LibreHealth EHR. We will need something for universities who want to beta test the educational version.

I might also mention in the demo I can’t add clinic doctors and can’t add immunizations or future appointments