Cluster coding
Cluster coding is a mechanism of linking related diagnosis codes through use of a diagnosis cluster identifier (DCID) that will enhance the value of coded data.
Why is cluster coding necessary?
While clinical coders assign and sequence codes, the relationship between codes is not always evident when the coded data is analysed. Assigning codes in sequence does not always indicate a relationship and sometimes that sequencing is not retained when the data is processed and validated by health information systems.
With cluster coding, the relationship between codes becomes clearer which adds meaning and improves the utility of the data.
Clustering increases the understanding of activity data and provides more powerful information for end users by:
- identifying relationships between codes, for example multiple poisonings and injuries in the context of accompanying external cause codes
- enhancing safety and quality reporting, for example identifying hospital acquired complications in unique clusters
- enhancing reporting of chronic conditions, for example reporting all chronic conditions regardless of whether they meet ACS 0002 Additional diagnoses
- reducing assumptions when interpreting data
- eliminating the need to review episodes of care to establish relationships between codes
- supporting future funding models
- preparing for a potential future implementation of ICD-11, where clustering is a feature.
Further benefits of implementing the DCID will be realised over time, as the scope of conditions eligible for clustering is explored in future editions of ICD-10-AM and then ICD-11.
Learn more about ICD-11, IHACPA’s work to prepare for a potential implementation of ICD-11 in Australia and clustering in ICD-11 here.
How will cluster coding work?
Cluster coding is proposed to be implemented in a staged approach and stage one will apply for ICD-10-AM and ACS Thirteenth Edition. In stage one, conditions assigned with external cause codes, and supplementary codes for chronic conditions will be allocated to specific clusters. The remaining codes will be allocated a DCID denoting they are not part of a diagnosis cluster or chronic condition cluster.
Table 1: DCID allocation
| Clustering in Thirteenth Edition | DCID value allocated |
|---|---|
|
Diagnosis cluster
|
A - ZZ |
| Chronic condition cluster The chronic condition cluster identifies conditions assigned in accordance with ACS 0003 Supplementary codes for chronic conditions. |
0 |
| Code not clustered DCID 8 represents an ICD-10-AM code that has not been allocated to a diagnosis cluster or chronic condition cluster. |
8 |
An example of a coded episode of care is provided, demonstrating how cluster coding is applied.
Codes assigned with external cause codes are eligible for a diagnosis cluster:
S52.30 Fracture of shaft of radius, part unspecified
V03.1 Pedestrian injured in collision with car, pick-up truck or van, traffic accident
Y92.49 Unspecified public highway, street or road
U72 Leisure activity, not elsewhere classified
These codes are allocated to DCID A.
Chronic conditions (assigned in accordance with ACS 0003) are eligible for the chronic condition cluster:
U82.3 Hypertension
U86.2 Arthritis and osteoarthritis [primary]
These codes are allocated to DCID 0.
The remaining codes assigned are not eligible for a diagnosis cluster or chronic condition cluster:
J44.1 Chronic obstructive pulmonary disease with acute exacerbation, unspecified
Z72.0 Tobacco use, current
These codes are allocated to DCID 8.
Guidelines in the Australian Coding Standards (ACS)
ACS 0004 Diagnosis cluster identifier (DCID) has been created to provide guidance on the purpose and application of cluster coding. This will be released as part of Thirteenth Edition.
Pilot exercise to test the proposed ACS 0004 Diagnosis cluster identifier (DCID)
In March 2024, IHACPA conducted a pilot of the draft ACS 0004 Diagnosis cluster identifier (DCID) to test the proposed classification guidelines.
Clinical coders with various levels of expertise were engaged to apply the draft standard to 20 clinical scenarios. The pilot engaged 71 participants representing all jurisdictions from both the public and private sector.
The results and participant feedback were used to inform the final version of ACS 0004. Feedback from the pilot exercise will also inform focus areas for new edition education.
The outcomes of the pilot exercise are in the Diagnosis Cluster Identifier (DCID) pilot - final report.
Frequently asked questions (FAQs)
IHACPA has developed responses to commonly asked questions about the implementation and impact of cluster coding.
Data standards
IHACPA has developed data standards to support the implementation of cluster coding in consultation with the National Health Data and Information Standards Committee.
Additional support for implementation
- Enhancements to the electronic code lists (ECLs)
Thirteenth Edition ECLs will identify codes that require application of the DCID. Flagging relevant codes assists software vendors in the application of the DCID and allows data collection agencies to formulate edits that will enhance the quality and application of the DCID. - Engagement with stakeholders
IHACPA has engaged with stakeholders through public consultation and forums, and held an industry briefing in April 2024. -
Education
Education on cluster coding will be available to clinical coders, health information managers, data users, clinical coding auditors and educators, software vendors and tertiary education providers through IHACPA Learn.Subscribe to the Australian Classification Exchange (ACE) mailing list to keep updated on the implementation of cluster coding.