Plans for National Lung Cancer Audit data validation

NLCA senior clinical lead Paul Beckett outlines the audit's plans to implement a local validation process in England to quality assure the registry data before publication of our next report.

At the beginning of this year, we published our annual report for patients who were diagnosed with lung cancer in 2015. In England, for the first time, this analysis was based on data from the National Cancer Registration and Analysis Service (NCRAS), as well as a number of data sources including the Cancer Outcomes and Services Dataset (COSD).

After the report was published, several lung cancer teams told us that the results that were reported for their organisation did not accurately reflect their performance. They said that the data quality was lower than they expected and that additional lung cancer cases that they did not recognise were included in the results.

We have developed and refined an algorithm to improve this process, and we believe that we now have 95% accuracy.

Paul Beckett, NLCA senior clinical lead

There have certainly been some local issues with transferring data from multidisciplinary team (MDT) software such as Somerset Cancer Register (SCR) and InfoFlex into the COSD, and our local data improvement leads have worked very hard with MDTs to rectify this.

We also carried out an in-depth audit in a number of trusts, which provided reassurance that the additional cases were real cases that were often unknown to the MDTs but were picked up by the cancer registry through its multiple data sources. These additional data sources include Patient Administration System (PAS) feeds, pathology reports and the National Radiotherapy Dataset (RTDS).

A further issue concerned the allocation of each case to individual organisations – we have developed and refined an algorithm to improve this process, and we believe that we now have 95% accuracy. As an audit team, we take the quality of the data very seriously, since they form the basis of quality improvement. For this reason, we have decided to implement a local validation process to quality assure the data that we use in our next report.

The main benefit of this validation will be to ensure that the allocated cases are real lung cancer cases, and that they have been allocated to the right trust.

Paul Beckett, NLCA senior clinical lead

All lung cancer MDTs have been contacted by the NCRAS and offered the opportunity to validate their data. Those that take up the offer will receive spreadsheets of patient-identifiable data on the cases that the registry currently believes should be allocated to that trust. Detailed instructions will be provided separately and will include the deadline for returning data, which will be necessarily short due to the need to keep within the timelines for publishing the report. When the NCRAS receives the updated spreadsheet, it will update the source data before they are supplied to the NLCA for analysis.

The main benefit of this validation will be to ensure that the allocated cases are real lung cancer cases, and that they have been allocated to the right trust. Most of this work can be undertaken by MDT coordinators and audit support staff rather than clinicians. There will also be the opportunity to update data on items such as performance status (PS), cancer stage, input from nurse specialists etc, which might require some clinical input.

We hope that this development will reassure our clinical community about our commitment to analysing and reporting on accurate data, and to moving the focus of the audit beyond data collection and onto improving the quality of care.

Paul Beckett, NLCA senior clinical lead

For the latest NLCA information follow us on twitter @RCP_NLCA or on the NLCA blog.