Link Lite

Configuration

"ConnectionStrings": {
  // Default is the details for the local development db in docker
  "Omop": "Server=localhost;Username=postgres;Password=example;Database=omop"
},
"Serilog": {
    //...
},
"QueryPollingInterval": 5, // seconds
"RquestCollectionId": "",
"RquestConnectorApi": {
    "BaseUrl": "", // for the Connector API specifically, not just general RQUEST URL, e.g. `https://rquestserver.com/task/capi/`,
    "QueueStatusEndpoint": "queue",
    "FetchQueryEndpoint": "query",
    "SubmitResultEndpoint": "result",
},

App requirements

• .NET 5 SDK

OMOP

Postgres Instance

A docker-compose file is provided in omop/ for dev environment use. It runs a postgres instance on the default port (5432) with the credentials specified, and the "adminer" GUI tool on port 8080.

Structure

This tool works with the OMOP Common Data Model 5.3.x

https://github.com/OHDSI/CommonDataModel

Follow the instructions / use the scripts for Postgres.

Versioned script archives, with per platform guidance, can be downloaded from the GitHub Releases.

Notes:

• the 5.3.1 (latest 5.x at time of writing) postgres DDL script is broken
  • find and replace DATETIME2 with TIMESTAMP to fix.
  • OHDSI/CommonDataModel#256
• When it says to populate data, if you have a dataset (see below), do so.
• ⚠ For now, do not add foreign key constraints (see below).

Data

For proof of concept this project uses a test dataset (available from the project team). Because this dataset is all that is queried, there is no need to populate OMOP Vocabularies.

⚠ Because of this, you should not add the foreign key constraints to the schema.

Test Data issues / notes

• location schema mismatch (test vs 5.3.1 script)
  • extra country, longitude and latitude columns
    • ignore them
• person schema mismatch (test vs 5.3.1 script)
  • extra death_datetime column
    • ignore it

Future

As the app progresses beyond the proof of concept stage, the following OMOP Vocabulary data will be needed:

• SNOMED
• ICD9
• ICD10
• READ
• LOINC

The vocabulary data can be acquired here with a free account.

As the app is productionised, guidance for loading organisation data (initial data at point of installation, as well as updates) will be formed, and possibly additional tooling or services (such as a local API facade) provided.