A Rails engine providing integration with Diaverum dialysers.

Development

Running

The gem has a built in test Renalware app you can run

bundle install
bundle exec rake db:create
bundle exec rake db:migrate
bundle exec rake db:seed
bin/web

Credentials are superkch renalware

Configuration

Incoming

For each HD patient dialysing at a Diaverum unit, Diaverum will SFTP us an XML file (probably overnight) containing HD Sessions for the last 30 days. We import these sessions into Renalware.

Running the import script

bundle exec rake diaverum:ingest
or, if running in development
bundle exec rake app:diaverum:ingest

Configuration

Configure in a host application using:

Renalware::Diaverum.configure do |config|
  config.xxx = "yyy"
end

The default configuration is in lib/renalware/diaverum/configuration.rb and is driven by an ENV variable of the same name:

diaverum_incoming_skip_session_save	Set to "false" to actually import sessions rather then doing a dry run.
DIAVERUM_INCOMING_SKIP_SESSION_SAVE
diaverum_go_live_date	e.g. "2018-11-01". Sessions (XML 'Treatments') earlier than this date will not be imported.
DIAVERUM_GO_LIVE_DATE
honour_treatment_deleted_flag	> If set to "true" then for each patient XML file, in addition to importing new sessions,
HONOUR_TREATMENT_DELETED_FLAG	we look at the backlog of sessions in the file (most of which should have already been imported)
	and if we find an imported session that now has a Deleted=1, we soft delete that session.
	This allow for the Diaverum to mark accidentally-created or erroneous sessions for deletion.

Outgoing

HL7

Currently the only data we send to Diaverum are HL7 pathology results.

Whenever a new HL7 message arrives in Renalware, the Diaverum::Outgoing::PathologyListener has the chance to inspect it; if it is for an HD patient dialysing at a Diaverum unit, it drops (via a background job) a file containing the HL7 message into the Diaverum out folder (probably a remote, mounted folder). The same file is also dropped into a local archive folder so we can move the files again if there is a communications issue.

We also log result, payload, patient and hospital unit in the diaverum.transmissions table, so that is the first port of call when debugging issues.

Note that although we drop the HL7 files in real-time into the SFTP folder, ready to send, this does not mean they are received in real-time at Diaverum. A cron job would normally schedule the SFTP sync.

Other data

Earlier in the development we sent patient XML files to Diaverum containing demographics and pathology data. Path data is not sent via HL7, and we no longer send anything else. If the need arises to send extended patient data to Diaverum then some the files are in this repo - either look for .dead files or check the git history.