Skip to end of metadata

• Created by Michiel Scheepmaker, last modified on Jun 07, 2019

Go to start of metadata

Private-preview application link: http://re-uat-terminology-app-01.gel.zone/ 

Contents

• Contents
• 1. Introduction
• 2. Data Sources
• 3. Search Participants
  • Data Model
  • Step 1: Specify search criteria
  • Step 2: Browse search results
  • Step 3: Download results
• 4. Code Systems
• 5. FHIR API
• 6. Clinical Data Mapping
  • Condition
  • Observation
  • Procedure
  • Encounter

1. Introduction

The purpose of the Terminology Toolset is to help you work effectively with the various clinical coding standards that are in use within the Research Environment:

• Explore the structure of different code systems, such as SNOMED CT, ICD-10, OPCS-4 and HPO
• Utilise subsumption, mappings, and concept attributes
• Create subsets of codes for your specific purpose 

With the Terminology Toolset you can search for Main Programme participants, using terminology resources to specify your search criteria:

• Specify search criteria for participants in terms of conditions, observations and/or procedures
• Lookup terms/concepts based on partially entered codes, descriptions or aliases 
• Include descendants of selected codes in your search (subsumption-based searching)
• Include mapped concepts from other code systems in your search 
• See descriptions of the codes that are present in the search result
• Download search results to a CSV file

2. Data Sources

Terminology reference data is provided by a FHIR terminology server (https://genomics.ontoserver.csiro.au/fhir)

Participant clinical data is obtained from the Main Programme Data Release version 6. 

3. Search Participants

Data Model

The diagram below depicts the clinical data model used for specifying search criteria and for displaying the results. This model is based on the HL7 FHIR resource model.

Please click here, to see how tables and columns in the Main Programme Data Release are mapped to this model.

Step 1: Specify search criteria

Specify a combination of conditions, observations and procedures to search for matching participants.

Notes:

• Click the "+" icon to add a new row to the search criteria. Each row has the form "<high-level concept> IS ANY OF <one or more parent concepts>".
• Select the high-level clinical concept (i.e. "Condition", "Observation" or "Procedure") to search for. 

• Search for terms from the available code systems, by entering part of the code or description. Notes:

  • The search will only return results when terms in the search criteria are from the same code system as the underlying data. For example, ICD-10 coded participant data can only be found if the search criteria include ICD-10 terms. 

  • You can use the "Automatically include mapped concepts" feature (described below) to find equivalent terms across different code systems.

  • For an overview of fields and code systems, see Terminology Service Tool: Clinical Data Mapping.

• Select on or more terms from the list. Notes:

  • Descendants of selected terms are automatically included in the search.

  • Searching with high-level terms, which can have thousands of descendants, may slow down the search.

  • It is not currently possible to search with terms that have more than 10000 descendants. 

• Add more rows using the "+" icon. The result will only include participants that match all the search criteria (AND-logic).
• Modify existing rows using the edit icon at the end of each row.
• To remove a row completely, click the edit icon followed by the bin icon.

• Select "Automatically include mapped concepts" to find and add mapped concepts from other code systems to the search criteria. For example, after selecting a SNOMED CT concept, the search can automatically include ICD-10 terms that have the same meaning as the SNOMED CT concept in the search, thus potentially finding more participants. Notes:

  • Mapped concepts may not have the exact same meaning as the selected concepts, so this feature must be used carefully! 

  • Mapped concepts and their descendants can be distinguished in the UI by their colour (green) and a special icon.

  • SNOMED CT to ICD10 and SNOMED to OPCS4 mappings are provided by NHSD. Targets that are equivalent or narrower are included. Targets that require combining multiple wider concepts have been excluded. 

  • SNOMED CT to HPO mappings are provided by CSIRO.

• When your search criteria are complete, click on "Continue" to see the search results.

Step 2: Browse search results

For each participant that matches the search criteria, the following fields are displayed:

• Participant ID
• Programme
• Type (Proband or Relative)
• Recruited Disease (possible multiple values)
• Year of Birth
• Phenotypic Sex
• Ethnic Category
• Consent Form
• Genome Build (possible multiple values)

Notes:

• To see why a participant matched the query, click on the participant row to see details of the conditions, procedures and/or observations that match the query, grouped by encounter.
• Results can be sorted on different columns by clicking on the column header.
• To modify your search criteria, click the "Back" button. To proceed to downloading the results, click the "Continue" button.

Step 3: Download results

Select columns to include and download a CSV file with the search results.

Notes:

• The result only includes unique rows.
• When including columns that can have multiple values per participant, the result may contain multiple rows per participant.
• When the participant ID column is not included, a column that contains the count of unique participants per row is automatically included. 

4. Code Systems

Browse and explore the structure of relevant code systems.

5. FHIR API

Behind this application is an HL7 FHIR-compliant terminology server, which is also available for your own scripts and applications to implement terminology-enabled functionality.

FHIR API endpoint: https://genomics.ontoserver.csiro.au/fhir

Disclaimer: this API is currently provided for testing purposes only and without any guarantees regarding availability or content quality.

Please note: the SSL certificate of the API endpoint may trigger a security warning when accessing the URL through a browser. In the case of the terminology server, this should not pose a threat and the risk can be accepted. If you understand and accept the risk, please click the “Accept the Risk and Continue” button. The issue is being addressed and will be resolved in the future.

To learn about the FHIR terminology server standard, see http://hl7.org/fhir/stu3/terminology-module.html

For details about our specific implementation of the terminology server and more examples, please refer to https://ontoserver.csiro.au/docs/5.1/api-fhir.html#.

Example API requests:

• Lookup ICD-10 code H356
• Lookup OPCS code X352
• Lookup HPO code HP:0000488
• Find ICD-10 codes where the description matches "epilepsy" (first 10 results))
• Find all SNOMED CT concepts subsumed by General Anaesthesia Procedure (expansion of an implicit value set)

6. Clinical Data Mapping

This section describes how the standard concepts of Condition, Observation, Procedure and Encounter map to the table structure of the Genomics England Main Programme data release. 

More fields will be included in future releases. 

Condition

The following table shows which fields from the Main Programme data model map to conditions, and the coding standards used.

Observation

The following table shows which fields from the Main Programme data model map to observations, and the coding standards used.

Procedure

The following table shows which fields from the Main Programme data model map to procedures, and the coding standards used.

Encounter

The following table shows which fields from the Main Programme data model map to encounters.