Definitions for the SACT data set

Within SACT, it is important that field naming is consistent within hospital systems and the definitions of the fields are unambiguous and applied by all providers.

All field naming and definitions must either be aligned with or approved by the NHSE Data Model and Dictionary Service or by the Data Design Authority. This allows for consistency across the NHS and prevents any unnecessary burden of data collection.

The following are definitions used throughout the SACT data set and user guide:

• the term ‘Regimen’ is used to identify a standard combination of drugs
• the term ‘Cycle’ is used to identify treatment intervals within a regimen
• the term ‘Administration’ is used to identify the physical administration of drugs

The relationships between programmes, regimens, cycles and administration dates are shown in the accompanying graphics and examples of data set structures.

Regimen:

A SACT regimen identifies a standard for a combination of drugs (or single drug) given in a planned schedule.

A regimen can be:

• standard
• part of a trial
• specifically designed for an individual treatment plan

The SACT drug regimen title will be as agreed by the SACT team and an NHS support pharmacists’ group, as they maintain the regimen list, and this will inform the OPCS Guidance (Classification of Interventions and Procedures).

Cycle:

Apart from continuous SACT, a regimen normally contains identifiable repeating elements, and each repeat should be identified and numbered. Some regimens have alternating repeating elements, and some have consecutive sets of repeating elements. In all these cases the term 'Cycle' would be equally valid and help to identify the stage of progress of the patient through SACT.

Cycle number:

These will be numbered sequentially within a regimen and the option to start from any number must be available to allow for prior management not recorded on the current system. The cycle number is now mandatory from v4.0.

Administration date:

In v4.0, the ‘Administration Date’ has changed after discussions with NHS England’s (NHSE) Data Design Authority and the Data Model and Dictionary teams.

From v4.0, the administration date has been split into two distinct choices, each having a different reporting format and at least one must be submitted within the data set record, as follows:

Choice 1:

Administration Timestamp (Infusion): For recording the date and time when the anti-cancer drug was administered to a patient (an infusion commenced).

Choice 2:

Administration Date (Oral Drug Dispensed): For recording continuous oral chemotherapy, the administration date will be the first day of the nominal cycle, or the date on which an oral drug was dispensed to the patient.

For these dates there is no change to the v3.0 format, in that only the date is required, and this must be recorded using the standard date format [ccyy-mm-dd].

The following data model explains the relationship between the data and the treatment process (Regimen and Cycle), provided to a patient:

The data structures are described below using three examples:

For the purposes of SACT the term ‘cancer’ relates to all conditions defined as registerable by the UK and Ireland Association of Cancer Registries (UKIACR) and these are listed in Appendix B.

These are in addition to Appendix A - Cancer Waiting Times ICD10 Codes and Tumour Groups for Primary Diagnoses. SACT requires that all treatments for new diagnoses and secondary/metastatic cancer are recorded.

All treatments for recurrences diagnosed at each Trust must also be included.

As with previous versions of SACT, v4.0 will not be converted to .xml reporting. Therefore, all Trusts should continue to report using the formatted .csv uploads as usual.

Choices

Within v4.0, we have introduced choices throughout the data set, these are to improve the quality of data submitted by the Trust and will be clearly explained within this user guide.

Sections

Within v4.0, we have introduced a series of sections throughout the data set, which link specific data items or defined attributes with their additional linked data item. This improves data quality and allows for the flow of data to be more logical in its approach. These will be clearly explained within this user guide.

Mandatory

The ‘Linkage’ data items are mandatory and must be submitted for all records. It is vital that these are always available so that the correct information can be linked to the right patient and the correct tumour. A record will not be able to be submitted if any mandatory data item is missing. These records must be added to the main file otherwise the whole file will not pass validation tests.

In some cases, certain data items have been made mandatory within sections to improve the quality of the data submitted. In these cases, no data within that section can be submitted without these mandatory data items being completed. These sections however are required; therefore, a missing section will not affect the submission of all other data.

Required

Most other data-items are set as ‘Required’. This means that if they are applicable to the reported tumour or patient pathway, they must be completed and treated as a mandatory item. Not every data-item however will be applicable to every patient, tumour, or treatment pathway. By using ‘Required’, this allows for a more accurate and inclusive collection of data. Therefore, all applicable data in each section marked as ‘required’ must be submitted for each record as soon as available.

Optional

Optional data items can be submitted by any Trust, but there is no requirement to enforce this data collection at this point. All optional data-items would remain under review and may change in future version controls of SACT.

Note:

• there are no optional data-items in v4

Meaning of 'Not known' value

'Not known' includes both 'not recorded' and for example 'test not done'. This is usually coded 9 or 99 (depending on the data item format).

Pilot

There are no pilot data items in SACT v4.0.

It is important to remember that each section and data item within the data set is controlled by its own published cardinality rule. This will help you understand the relationship each section or data item has within the overall file. There is a range of cardinality within the data set, these are at both the section and data item level.

Section level cardinality:

• (1..1)   this indicates that the section is mandatory and ‘must be one occurrence per record’
• (1..2)   this indicates that the section is mandatory and there ‘must be at least one of the following choices per record’
• (0..*)   this indicates that the section is required and ‘may be multiple occurrences per submission’
• (0..1)   this indicates that the section is required and ‘may be up to one occurrence per record’

Data item level cardinality:

• (1..1)   this indicates that the data item is mandatory and must be submitted
• (1..*)   this indicates that the data is mandatory and that multiple attributes are allowed to be reported
• (0..1)   this indicates that the data item is required/optional
• (0..*)   this indicates that the data is required/optional, but that multiple attributes are allowed to be reported