Technical note on the calculation of the baseline for the Detect Cancer Early HEAT target

3. Data sources

3.1 There are two principal datasets used to record stage of diagnosis in patients with cancer. These are cancer registry and cancer audit data.

3.2 Cancer registry is compiled by ISD staff and uses multiple sources of electronic data such as hospital discharge records, pathology records, death records, etc. to create individual cancer registry records. It is population-based and does not gather an over-ambitious amount of data as this may risk reducing the data quality of the most important variables. Another important benefit is that the data set remains reasonably stable over time. These characteristics (as well as a very clear set of guiding principles) also support international comparisons.

3.3 Considerable effort is expended to maximise case ascertainment including cases that may never undergo any treatment. There are a relatively high proportion of records in this data set where staging is unknown due to the complexity of pulling together the different information sources.

3.4 The intricate nature of this task also means it takes a long time to compile the data in to cancer registries. The data is therefore not able to be published until in excess of one year following derivation - for example, data for 2010 was published in Summer 2012 (and staging data was not published with this release). Such a delay in the publication of information is not ideal for use in monitoring progress against the HEAT target. Real time information is more appropriate for the operation of the DCE Programme so the impact of interventions such as the national and local public awareness campaigns and the initiatives to improve cancer screening uptake can be assessed and improved upon.

3.5 Cancer registry data have to meet the needs of many users including policy makers, epidemiologists, researchers, geneticists, and health services managers and clinicians. The data are used for a very wide array of purposes, and are collected according to international standards to facilitate international comparisons of incidence and survival.

3.6 More information on cancer registry data can be found here:

http://www.isdscotland.org/Health-Topics/Cancer/Scottish-Cancer-Registry/

3.7 Cancer audit data have only been gathered much more recently and are collected directly by NHS Boards. Due to the close working relationships between clinicians and audit staff staging data is more complete for cancer audit data than for cancer registry.

3.8 A further benefit of cancer audit data is that it focuses on selected single types of cancer (and does not necessarily cover every histological subtype of cancer). This makes it easier to collect a more detailed data set without compromising the integrity of the most important data items. However, as the focus of the data collection is around monitoring performance and informing clinical decisions around treatment there is not the same requirement to collect information on all cases where there has been a diagnosis of cancer. So, for example, people diagnosed for the first time at post-mortem would not be included in the data set.

3.9 Although there is considerable overlap with data items within the Cancer Registry data set, the additional need for in-depth clinical and pathological information means the reason for collecting it is quite different in nature. The data is also used to inform Quality Performance Indicators (3), a critical work stream for the Scottish Cancer Taskforce.

3.10 The differences between cancer audit and cancer registry data are summarised in Annex A.