What's new in COSD 2020
The new version of COSD v9.0 came into effect on the 1 April 2020. These changes continue to build on the work started in v7.0 and further rolled out through v8.1. They include some new amendments and re-alignments to the current data set, as well as a revision of the current schema specification in order to continue to meet the business objectives of the standard.
The COSD v9.0 amendments were made following extensive consultation over a 9 month period with 40 key groups or 150 clinical experts, including:
- site specific Expert Advisory Groups (EAGs) within PHE
- experts from within the National Cancer Registration and Analysis Service (NCRAS)
- clinical support and advice from the chair of the Royal College of Pathologists Working Group on Cancer Services
- cancer charities (Cancer Research UK, AMMF - The Cholangiocarcinoma Charity, Breast Cancer Care, Living Beyond Cancer and MacMillan)
- patient groups and individuals through ‘Use My Data’ and the EAGs
- national cancer audits
- cancer and pathology system suppliers
- NHS England
- Cancer Waiting Times (CWT)
- quality surveillance
- cancer taskforce transformation board
This revision allowed the data sets to be clinically reviewed, validated and updated by experts in all fields of cancer, and provides a clinically sound set of data to be collected from April 2020 onwards.
However, it was recognised that there were still further improvements required within this release to improve data quality and to be more reflective of current and future changes in cancer treatments, outcomes and clinical care.
Mandating key data items was vital in this release to improve data quality and reduce the burden of data processing within NCRAS. If a treatment record for example is submitted without a date or modality, it is of no use as a cancer registration event, however a registration officer would still have to review each incomplete record submitted, this multiplied by thousands of incomplete records per year was an unsustainable practice.
It also provides a clinically sound set of data which meets crucial recommendations in the Achieving World-Class Cancer Outcomes, A Strategy for England 2015-2020 (Cancer Taskforce Report).
Key changes to COSD
In total, 101 new data items (18 pathology specific) were added to meet the changing demands of cancer treatment and outcome data. In addition, there were 80 data items deleted, of which 12 are pathology specific. More importantly, this version has a strong emphasis on improving data quality and ascertainment along with formally separating pathology away from the patient pathway elements of a cancer record.
Many sections in the core now have mandatory elements within them, this will drive up data quality without creating undue burden on either the people collecting the data (within the Trusts) or those processing the data (within NCRAS).
In COSD v9.0 there has been an emphasis on improving data quality and ascertainment along with formally separating pathology away from the patient pathway elements of a cancer record.
Other new features in v9.0 include:
- 101 new data items (18 pathology specific) to meet the changing demands of cancer treatment and outcome data, e.g. in relation to acute oncology, HPV testing for head and neck cancers
- deletion of 80 data items, of which 12 are pathology specific
- certain key data items are being made mandatory, to improve data quality and reduce the burden of data processing
- the addition of choices to data items to enable clearer decision making and improve data quality
- an update of the pathology data items to align with the Royal College of Pathologists’ ‘Core’ data sets
- new data items within the Breast and Upper GI site specific sections
- new data items to support the Living With and Beyond Cancer campaign
- an improved mechanism for recording all ‘Non Primary Cancer Pathway’ to improve collection and data quality
- realignment or movement of data items, to ensure that data nests correctly within the XML
Many data have been re-aligned across the data set into the correct higher level groupings. The data set maintains its ability to be interpreted within the following 2 subsets of data.
This was part of the last standard and continues to be mandated across all Trusts to supply these data in COSD XML directly from their pathology departments.
It uses a different schema from COSD and has been formally separated within this version review and release.
This is the data excluding Pathology, which the Cancer Services Teams need to collect. By formally removing the pathology data from their workload, this could reduce their burden of data collection by up to 30% across the whole data set.
As these data are now collected and submitted by the pathology departments directly, it is a huge burden of duplication if we ask the Cancer Services (non-clinical) teams to transcribe the same data into COSD via a Trust's Cancer Information System.
Wherever possible, duplication across the data set has now been removed and full explanations of how to collect these data within the new structure provided within the changelog of the data set.
Where there were data that are no longer part of a linked national data set (for example, Royal College of Pathologists), these have also been removed from the data set. Expert advice and guidance was provided by the Royal College of Pathologists during the review of the COSD Pathology v4.0 data set.
Collaboration between the (CWT) data set developers and COSD have improved both data sets by removing or enhancing specific shared data items.
The downloads and additional resources section of this page contain updated versions of the COSD, Guidance and supporting documentation for both the main COSD and the final pathology specific data set. It also includes links to download the schema packs to support structured XML submissions.
See the NHS Digital website for all other documentation, including the new Specification and Implementation guide.