Initial Release
This project is an example of configuration and use of InterSystems IRIS Workflow functionality simulating the treatment of patients with high blood pressure.
What will you find in this project?
Angular application to use as a front-end.
InterSystems IRIS instance configured by default to create and run an interoperability production configured to work with Workflow engine.
Open a terminal in Visual Studio and just write:
docker-compose up -d
You will see 2 containers running on your Docker.
Now it’s time to configure IRIS.
Docker deployment will configure automatically the production wich manages the admission of HL7 messages and the creation of tasks, you don’t need to do anything.
HL7_File_IN is configured to read HL7 files from /shared/in folder and resend the HL7 messages to Workflow.BP.BloodPressurePlan where they will be processed.
Workflow.BP.BloodPressurePlan
This Business Process cover the entire functionality for the patient, it creates the user associated to the National Document of Identity and manages the tasks depending the values fetched from the HL7 messages.
AutomaticBloodPressureRole and ManualBloodPressureRole Business Operations will connect our Business Process with the Workflow engine, creating and returning the generated tasks.
This project is configured to deploy an Angular application in the following address: URL. This application has been created with Angular Material to simulate an application for smartphones, if you are using Google Chrome or Firefox you can change the visualization pressing CTRL + SHIFT + M
To log in Workflow UI application you can use the FakeDoctor account created during the Docker deployment, the username is FakeDoctor and the password is SYS
To launch the process you only have to copy and paste message_1_1.hl7 into /shared/in folder. With this action a new user will be created (07751332X/SYS) and two new tasks will be associated to him.
The task with no action associated will be automatically closed as soon as a new hl7 file (message_1_1.hl7) is received in the production meanwhile the task with an action will be waiting for the user confirmation.