Outlier management for National Clinical Audits
Outlier management can help to inform local quality improvement by targeting efforts and monitoring improvement or decline over time. All National Clinical Audits in the NCAPOP should:
1. Once established, undertake outlier analyses and notification for each national reporting cycle
2. Develop and follow a project-specific outlier policy which should be available publicly on their project webpage. This is normally based on the HQIP guidance (see guidance and implementation guides listed below for England and Wales)
The following outlier guidance applies to:
- Comparisons of healthcare providers (general practices, hospitals, trusts, clinical networks, but not individual practitioners) using batches of data collected over an appropriate defined period (and not continuous monitoring); and
- Both outcome and process measures of performance (referred to as ‘performance indicators’)
Building on successful examples of sharing ‘alert’ level data, HQIP has agreed to support extending ‘alert’ level data to further NCAPOP audits. The HQIP team will work with NCAPOP providers to support this on a case-by-case basis. For further details please see: ‘Briefing document for NCAPOP Providers: Sharing of alert level outliers, HQIP/CQC, July 2019’.
Current guidance for England
The following applies to any data from patient cohorts with a collection starting from 01/11/2017:
Current guidance for Wales
The following applies to any data from patient cohorts with a collection starting from 01/11/2018:
Previous guidance for England
The following was applied to data from patient cohorts with a collection starting before 01/11/2017 (i.e. it is superseded by the first two documents for England above):
Previous guidance for Wales
There was no previous guidance for Wales for data from patient cohorts with collection starting before 01/11/2018.
For further information, NCAPOP audits should consult the HQIP Provider Technical Manual.
 This is covered by the Clinical Outcomes Publication Technical Manual, HQIP, (2016)