Intro to Healthcare Interoperability

Examples you can use to learn the main ideas involved in HealthCare Interoperability using InterSystems IRIS for Health.

You can find more in-depth information in https://learning.intersystems.com.

👉 Slides available here Healthcare-Interop-Workshop.pdf

What do you need to install?

• Git
• Docker (if you are using Windows, make sure you set your Docker installation to use "Linux containers").
• Docker Compose
• Visual Studio Code + InterSystems ObjectScript VSCode Extension

Setup

Build the image we will use during the workshop:

git clone https://github.com/intersystems-ib/workshop-healthcare-interop
cd workshop-healthcare-interop

docker compose build

• Run the containers we will use in the workshop:

docker compose up -d

Open workshop-healthcare-interop folder in VS Code.

Basics

Start FindRateProduction

• Open the Management Portal.
• Login using the default superuser/ SYS account.
• Interoperability > Namespace INTEROP > List > Productions > Demo.Loan.FindRateProduction > Open
• Start Production
• Have a look at the Business Services, Business Processes and Business Operations.
• Click on the connector (green ball) to see how the components are linked.
• See the Legend to understand the meaning of the different colors of the components.

Test a Business Operation

• Click on Demo.Loan.WebOperations
• Go to Actions tab > Test > Choose a Demo.Loan.CreditRatingRequest message.
• Enter some input and see the output in the resulting Visual Trace.
• Have a look at the involved Business Operation and Messages in VS Code.

Make a sample loan request

• Open http://localhost:52773/csp/healthshare/interop/DemoLoanForm.csp and enter some data (you can test with different values).
• Go back to your production and open the Message Viewer.
• Have a look at the messages, go through some of the traces.
• Pay attention to the elements involved, sync/async calls, message contents and possible errors or alerts thrown.

Inspect a Business Process

• Back in Demo.Loan.FindRateProduction Production, click on Demo.Loan.FindRateDecisionProcessBPL.
• On the settings tab, click on the magnifyer icon on the Class name setting.
• Inspect the graphical BPL definition of the process.
• When you are done, stop the production.

Start Demo.HL7.MsgRouter.Production

• Interoperability > Namespace INTEROP > List > Productions > Demo.HL7.MsgRouter.Production > Open
• Start production.
• Have a look at the production, notice the prebuilt HL7 Business Services and Operations that are being used.
• Explore the settings on those services and operations (e.g. FilePath, etc.)

Process some sample HL7 messages

• In your VS Code with workshop-healthcare-interop opened, copy test/*.txt files into test/in subdirectory.
• Go back to the production and see Message Viewer.
• Explore some the new messages that have appeared. Notice the HL7 messages.

Explore routing rules and data transforms

• Back in Demo.HL7.MsgRouter.Production Production configuration page, click on XYZ_Router.
• Click on the magnifying glass on Business Rule Name in the Settings Tab.
• Notice the different routing rules based on the content of the HL7 messages.
• Find some of the rules that are using a Data Transform.
• Double-click on the Data Transform element and open the DTL Editor.
• Have a look at how can messages be transformed.

Extend HL7 production

Let's say that now you need to grab incoming ORM^O01 messages from XYZ facility and create files with some patient data to send to other systems.

So, you will need to figure out how to:

• Process incoming XYZ_HL7FileService ORM^O01 HL7 message
• Extract patient data: medical record number, name, etc.
• Write that information into a file

You can follow these guidelines and try to implement it on your own:

• Create a new message interop.msg.PatientDataReq that will contain your patient data. Check documentation about defining messages.

• Create a new Business Operation interop.bo.PatientFileOperation using EnsLib.File.OutboundAdapter. This operation will write your patient data to a file. Have a look at the documentation so you can see what's the structure that you need.

• Add your business operation to production and test it.

• Create a data transformation interop.dt.ormO01toPatientDataReq to transform ORM^O01 HL7 messages into interop.msg.PatientDataReq so you can extract patient information. Test your data transformation.

• Modify XYZ_Router rules so you can grab ORM^O01 messages, transform them using interop.dt.ormO01toPatientDataReq and finally sending them to your business operation.

FHIR Repository

Create FHIR endpoint

Create FHIR server in Health > FHIRREPO > FHIR Configuration > Server Configuration as:

• Endpoint: /csp/healthshare/fhirrepo/fhir/r4
• Core FHIR package: [email protected]

Load simple FHIR data

Load some sample FHIR data into repo using WebTerminal.

zn "FHIRREPO"
set sc = ##class(HS.FHIRServer.Tools.DataLoader).SubmitResourceFiles("/app/install/simple-fhir-data/","FHIRServer","/csp/healthshare/fhirrepo/fhir/r4")

Test FHIR repository

Open the Postman collection included in workshop-healthcare-interop.postman_collection.json and try some requests on FHIR Repo. Simple and FHIR Repo. Queries.

FHIR Interoperability

Scenario: FHIR client

Build a fhir client to send some requests to a FHIR server.

Have a look at the interop.Production production:

• interop.bs.FHIRFileService - Business Service that reads a file and creates a HS.FHIRServer.Interop.Request message.
• interop.bp.FileToFHIRService - Business Process that prepare HS.FHIRServer.Interop.Request from service before sending to external FHIR server.
• HS.FHIRServer.Interop.HTTPOperation - built-in Business Operation that sends a FHIR request to an external FHIR server.

Run some tests:

• Start the production
• Copy data/patient.json into data/fhir-input to process a sample file in your production.

Scenario: FHIR request processing

In this scenario you need to receive FHIR requests in an interoperability production, maybe manipulate them and finally send that requests to a FHIR server.

In case you need to forward a FHIR request to an external server you can use simple FHIR Interoperability Adapter in InterSystems IRIS or HealthShare Health Connect.

You can find more information in FHIR Interoperability Adapter.

Install adapter

You need to install the FHIR interoperability adapter before using it in a namespace. During the adapter installation it will create:

• A web application for your FHIR server endpoint.
• An InteropService and InteropOperation in your production.

Install the adapter using WebTerminal:

zn "INTEROP"
set status = ##class(HS.FHIRServer.Installer).InteropAdapterConfig("/myendpoint/r4")

Test your service using the Postman collection in workshop-healthcare-interop.postman_collection.json and running some requests in Interop directory.

Check some other features

Latest release of IRIS For Health introduces some interesting features such as FHIR Object Classes.

So keep in mind that you just can use objects to work with FHIR content as you need:

set file = ##class(%Stream.FileCharacter).%New()
set file.Filename = "/app/data/patient.json"
set json = {}.%FromJSON(file)
set patient =  ##class(HS.FHIRModel.R4.Patient).fromDao(json)
zw patient.name.get(0).family

FHIR Analytics: FHIR SQL Builder

What about running analytics on top of a FHIR repository? Well, FHIR model is a directed graph, so it's not trivial.

However, you can have a look at a new experimental feature: InterSystems FHIR SQL Builder.

FHIR SQL Builder is a tool that allows you to create your SQL schemas using data from your FHIR repository without moving the data to a separate SQL repository.

You will need to go through three simple steps:

Analyze your FHIR repository data

Access FHIR SQL Builder and create a New Analysis:

• New FHIR repository
• Name: fhirrepo
• Host: localhost
• Port: 52773
• Credentials: create new credentials using superuser / SYS
• FHIR repository endpoint: /csp/healthshare/fhirrepo/fhir/r4

Choose data you want to project to SQL

Create a New Transformation Specification.

Add some FHIR fields that will be projected as SQL, you can try the following:

• Patient.gender
• Observation.code.coding.code
• Observation.valueQuantity.value

Project data to SQL

Simply create a New Projection specifying the package you want to use for your projection (e.g. demo).

After that, you can access your data using SQL!

In your workshop you have included a JupyterLab Notebook to play with the data :)