Seeding ONC Patient Matching Data
Last updated
Last updated
Once a docker image is setup - it can now be seeded or populated with data. In addition to running the installation qualification tool, you can seed the ONC dataset into the SanteMPI server which has just been setup.
It is recommended seeding the entire ONC dataset be done on a host machine with at least 8 GB of RAM, 4 CPU cores, and an SSD
In order to complete this tutorial you will require the following tools:
A setup version of SanteMPI (you can use the Quick Start Guide)
The ONC Patient Matching Challenge Data (https://github.com/onc-healthit/patient-matching)
The SanteDB SDK (you can see Operationalizing SanteDB)
SanteMPI Version 3.x contains the necessary configuration for the ONC patient matching data in a dataset file which can be downloaded from our GitHub page: https://github.com/santedb/santempi/blob/master/data/090-ONC%20Patient%20Matching%20Domains.dataset
Once downloaded, visit the SanteMPI administration console and import the dataset file via CDR Administration > Import Data
Once uploaded - administrators will see the dataaset scheduled for import:
Pressing the Run Scheduled will setup the identity domains.
Following this step, the ONC dataset files can be uploaded, when uploading a dataset file, be sure to set the type of file to ONC Patient Matching Challenge as pictured below:
Once the files are uploaded, you can run the import.
The seeding of the entire ONC patient dataset can take upwards of 12 hours depending on the complexity of your matching configuration.
First, we're going to need to prepare our SanteDB server to accept the ONC data, this includes:
Creating the necessary identity domains for the ONC spreadsheets
Setting the matching configuration appropriately
The ONC dataset has two identity domains which need to be created in the SanteDB server. To do this, log into the administrative portal and click on Reference Data then Identity Domains and press Create.
You will need to register two identity domains:
The ONC dataset is a large dataset where the source records (provided in the CSV files) can reliably represent a patient, because of this we can get an added performance bonus by changing our match configuration to source/source mode rather than source/master (since there is nothing that a master record would add in blocking).
To change the match configuration, click on SanteMPI then Match Settings locate the org.santedb.matching.patient.default entry and click View.
On the Blocking tab if we expand a blocking instruction we see the default method of matching is Source/Master:
We can change this by pressing Edit (pencil) and modifying both blocking instructions. After this step the configuration should appear as:
Now that our MPI is setup correctly we should navigate to the directory where the ONC patient source data was cloned. To seed the data more easily, create a new batch file (if on Windows - if on Unix you should create an equivalent shell file) to call the SDK:
The batch file will look at all the CSV files in the current directory and for each one it will:
Start a new process
Call the importer from the SDK with
The CSV file to be seeded
The realm on the local computer
Logged in as "Allison" and "Mohawk123" as password (i.e. this is Allison sending the data to the MPI)
Source format is ONC (opposed to FEBRL)
Enterprise ID is the ONC domain
Medical Record Number field maps to MRN domain
Social Security Number field maps to SSN domain
Now you can open a command prompt and type seed-onc.bat to start the seeding process. There will 9 consoles opened and 9 clients will begin seeding data to the MPI.
The overall health of your deployment can be monitored on the Probes page.
| Namespace | URL | OID | Unique |
|---|---|---|---|
ONC
http://onc.org/eid
2.25.4949384934
true
MRN
http://onc.org/mrn
2.25.17237262721
false
