grouperserve

Microservice for the SwissDRG grouper. Provided as a REST JSON API over HTTP/HTTPS using the Java framework Spark.

Getting started

Copy grouper jar to the libs directory:

mkdir libs
cp grouper-X.Y.Z-all.jar libs 

Make sure gradle -v returns something similar to this:

------------------------------------------------------------
Gradle 3.0
------------------------------------------------------------

Build time:   2016-08-15 13:15:01 UTC
Revision:     ad76ba00f59ecb287bd3c037bd25fc3df13ca558

Groovy:       2.4.7
Ant:          Apache Ant(TM) version 1.9.6 compiled on June 29 2015
JVM:          1.8.0_101 (Oracle Corporation 25.101-b13)
OS:           Linux 3.19.0-59-generic amd64

Build jar:

gradle build

In order for the service to be functional you have to provide the JSON specifications (as provided by SwissDRG AG) for each system in the folder /grouperspecs with the folder names matching the field 'version' as obtained by the above call and stored in grouperspecs/systems.json.

Run grouper server:

java -cp build/libs/grouperserve-0.2.0.jar ch.eonum.grouperserve.GrouperServe

Test this URL in your browser: http://localhost:4567/systems You should obtain a list of all provided SwissDRG systems in a JSON array.

The patient case URL format

See [https://docs.swissdrg.org/BatchgrouperFormat2017.pdf]. Use '_' instead of ';', '-' instead of '|' and '$' instead of ':'.

API calls

If you test the API in your shell e.g. using curl: export ROOT_URL=http://localhost:4567

systems

Obtain a list of available SwissDRG systems.

URL: GET /systems

sample call using curl:

curl "$ROOT_URL/systems"

group

Group a patient case and obtain a grouper result and an effective cost weight. You can find a documentation of the grouper result and the effective cost weight in the documentation for the SwissDRG grouper.

URL: POST /group

Parameters:

• pc: patient case in the URL patient case format
• version: version identifier as provided by /systems
• pretty: (true|false) pretty print JSON (default is false)
• annotate: (true|false) return the annotated patient case with CCL values, validations and used flags (default is false)
• zegroup: (true|false) return Zusatzentgelte in field 'zeResult'

sample call using curl:

curl --header "Accept: application/json" --data "version=V5_A&pc=11_65_0_0_M__01__00_1_0_I481-Z921-F051_8954_&pretty=true" "$ROOT_URL/group"

group_many

Group more than one patient case in one request.

URL: POST /group_many

Parameters:

• pcs: patient cases JSON array with strings in the URL patient case format
• version, pretty, annotate: see /group

sample call using curl:

curl --header "Accept: application/json" --data "version=V5_A&pcs=[\"11_65_0_0_M__01__00_1_0_I481-Z921-F051_8954_\", \"12_65_0_0_M__01__00_1_0_I481__\"]&pretty=true" "http://localhost:4567/group_many"

Configure as Systemd service

sudo cp grouperserve.service /etc/systemd/system/
# adapt working directory
sudo systemctl daemon-reload
sudo systemctl enable grouperserve.service
sudo systemctl start grouperserve.service

Run as Docker container

Run the server:

docker run -it -v $PWD:/opt/grouperserve -p 4567:4567 --workdir /opt/grouperserve --rm openjdk:11 java -cp build/libs/grouperserve-0.2.0.jar ch.eonum.grouperserve.GrouperServe

Run in detached mode:

docker run -v $PWD:/opt/grouperserve -p 4567:4567 --workdir /opt/grouperserve --name=grouperserve --detach=true openjdk:11 java -cp build/libs/grouperserve-0.2.0.jar ch.eonum.grouperserve.GrouperServe

In detached mode you can stop, kill, remove or start the server as follows:

docker stop grouperserve
docker kill grouperserve
docker rm grouperserve
docker start grouperserve

Follow the logs:

docker logs -ft grouperserve

Manage container with SystemD

sudo cp docker-grouperserve.service /etc/systemd/system/
sudo systemctl daemon-reload
sudo systemctl start docker-grouperserve.service
sudo systemctl enable docker-grouperserve.service