DSA with NHS Digital to enable the Cancer Alliance to access the CWT system

This agreement is for the Somerset, Wiltshire, Avon and Gloucestershire Cancer Alliance (SWAG) Cancer Alliance to access Cancer Waiting Times data.

However as the Cancer Alliance is not a legal entity, as staff are substantially employed by NHS England, who are therefore the lead organisation, and the data controller who processes data. In this agreement therefore, all references to accessing patient level data refer to the legal entity – NHS South Central and West Commissioning Support Unit (listed as a processor) are also part of the legal entity NHS England and are permitted to process the data.

Within NHS England are seven regions who support local systems to provide more joined up and sustainable care for patients. The regional teams are responsible for the quality, financial and operational performance of all NHS organisations in their region, drawing on the expertise and support of our corporate teams to improve services for patients and support local transformation.

Improvements for Cancer patients

The independent Cancer Taskforce set out an ambitious vision for improving services, care and outcomes for everyone with Cancer: fewer people getting Cancer, more people surviving Cancer, more people having a good experience of their treatment and care, whoever they are and wherever they live, and more people being supported to live as well as possible after treatment has finished.

Cancer Alliances

Cancer Alliances, which have been set up across England, are key to driving the change needed across the country to achieve the Taskforces vision. Bringing together local clinical and managerial leaders from providers and commissioners who represent the whole Cancer pathway, Cancer Alliances provide the opportunity for a different way of working to improve and transform Cancer services. Cancer Alliance partners will take a whole population, whole pathway approach to improving outcomes across their geographical footprints building on their relevant Sustainability and Transformation Plans (STPs). They will bring together influential local decision-makers and be responsible for directing funding to transform services and care across whole pathways, reducing variation in the availability of good care and treatment for all people with Cancer, and delivering continuous improvement and reduction in inequality of experience. They will particularly focus on leading transformations at scale to improve survival, early diagnosis, patient experience and long term quality of life. Successful delivery will be shown in improvements in ratings in the Clinical Commissioning Group (CCG) Improvement and Assessment Framework (IAF), including, importantly, in the 62 day wait from referral to first treatment standard.

Cancer Wait Times (CWT) system

The Cancer Wait Times (CWT) system collects and validates the National Cancer Waiting Times Monitoring Data Set (NCWTMDS), allowing performance to be measured against operational Cancer standards. Data is validated and records merged to the same pathway to cover the period from referral to first definitive treatment for Cancer and any additional subsequent treatments.

The CWT system then determines whether the operational standard(s) that apply were met or not for the patient and the accountable provider(s). The CWT system holds NCWTMDS in a series of pre-aggregated static reports. These reports are available monthly and quarterly data (aligned with the National Statistics for Cancer Waiting Times published by NHS England). Users can query the CWT system to generate reports to feedback on the progress towards meeting these targets.

Cancer alliances are also created to drive improvement in cancer outcomes. Align with the improvement trajectory set for cancer survival (also part of CCG IAF), cancer alliances are set to deliver the Faster Diagnostic Standards (FDS) from April 2021 (delayed from April 2020). FDS is part of CWT dataset, referring to the duration between urgent GP referral to patients being told whether they have a cancer diagnosis or not.

The National Cancer Programme has confirmed that FDS, along with 62-day wait, will be key metrics within the 10 year NHS Plan that Cancer Alliances will be held accountable to. Thus without access to the data as outlined in this request, the Cancer Alliance will not be able to deliver work programme as outlined by the National Cancer Programme. The Cancer Alliance will directly access the Cancer Waiting Times System on behalf of alliance member trusts and CCGs.

SWAG Cancer Alliance works with health organisations across Somerset, Wiltshire, Avon and Gloucestershire including 7 acute providers and 4 CCGs

Acute Providers

Gloucestershire Hospitals NHS Foundation Trust

North Bristol NHS Trust

Royal United Hospitals Bath NHS Foundation Trust

Salisbury NHS Foundation Trust

Somerset NHS Foundation Trust

Yeovil District Hospital NHS Foundation Trust

CCGs

NHS Bristol, North Somerset and South Gloucestershire CCG

NHS Bath, Swindon and Wiltshire CCG

NHS Gloucestershire CCG

NHS Somerset CCG

Data access

The CWT system provides the Data Controller / Processor representing each Cancer Alliance, with access to the following;

a) Aggregate reports (which may include unsuppressed small numbers)

b) Pseudonymised record level data - users can directly download this data from the CWT system

c) I-View Plus tool

The organisation will only access patient records which fall within the Cancer Alliances' footprint of responsibility based on the patients' CCG of responsibility.

A) Aggregate reports including small numbers

Aggregate data is available in the form of reports at Provider (Trust) and Clinical Commissioning Group (CCG) level. Small numbers may be included in the aggregate data reports and are essential for analyses carried out by lead organisations.

Investigating breaches

The Data Controller routinely monitors performance and standards using the CWT system, particularly in relation to breaches of the 62 day wait target. Due to the large number of potential Trust/CCG combinations, breach counts could result in small numbers as in some cases there are less than 6 breaches in a whole year. Given that financial penalties are linked to target breaches counts must accurately reflect the true percentage without suppression.

Mitigating risk of re-identification

Risk of disclosure is minimised as the dataset does not include patient demographics (increasing risk of re-identification) that may allow users to identify an individual e.g. there are no age, ethnic categories or geographic breakdowns based on patient postcode.

Additionally, the aggregation categories are such that the data is not at a lesser granular level e.g. the source NCWTMDS data collects information at ICD diagnosis code level, but the CWT system aggregates at tumour group level e.g. Head & Neck, Upper GI, Lower GI, Breast etc.

B) Pseudonymised record level extracts

Approved users will access record level pseudonymised data which includes the system generated pseudo CWT patient ID. Any record level data extracted from the system will not be processed outside of the authorised users of the system.

C) i-View Plus

iView Plus uses cube functionality to allow lead organisations to produce graphs, charts and tabulations from the data through the construction of queries. The data in iView plus is split by operational standard being measured and can then be analysed against a range of dimensions collected in the data and measures such as count, percentage and median. The outputs of iView Plus are aggregate, and no record level data can be obtained, however some queries may result in small numbers and these currently have limited disclosure control applied, see A) for further explanation. iView Plus holds published data, the lowest organisational granularity is trust level, data can also be aggregated to CCG level and other health hierarchies.

The Cancer Alliance will use the data to both monitor and improve performance against the Cancer Waiting Time standards and to inform wider Cancer pathway improvements.

The Cancer Alliance's use of the data will fall into two separate categories, each requiring different levels of suppression, and onward sharing both within the Cancer Alliance and with wider NHS stakeholders;

Purpose One - Aggregate local reports

Generation of routine Cancer Waiting Times reports at Provider (Trust) or CCG level. Lead organisations will access a summary of the totals for the Providers (Trust) and CCG's that are treating cancer patients where they have a commissioning responsibility for that patient (based on the CCG they are aligned to). This analysis would then be shared with the providers and commissioners and used to inform service improvement by providing benchmarked comparable data. The format of this report would be in a tabulated or graphical form (i.e. not record level) but may contain small numbers. An example of where small numbers would not be suppressed would be in relation to cases of breaches against a standard where small numbers would be essential to ensure the report is meaningful.

Examples of this type of analysis include:

a. Comparative Cancer Waiting Times performance at tumour group and individual tumour site (i.e. ICD10 code) level for Trusts and CCGs across the geography

b. Analysis of Cancer Waiting Times performance by treatment modality

c. Grouping length of waits for standards

d. Analysis of derived breach reason fields to identify trends in reasons for delays

e. To provide assurance through comparative analysis (e.g. orphan record identification, active monitoring proportions and validation of waiting list adjustments entered)

f. Analysis of flows of patients including analysis by provider trust site

g. Reviewing waits between surgery and radiotherapy for Head and Neck Cancer patients with a maximum recommended wait of 6 weeks

h. Reviewing routes to diagnosis of patients

i. Quantifying treatment volumes by provider organisation including analysis treatment rates

Purpose Two - Sharing of record level data with providers and commissioners responsible for direct patient care for that patient. This will be for local audit purposes.

The two broad purposes for this would be;

1) To support audit work

2) Investigate individual outliers to the national standards

Pathway analysis will be undertaken, identifying trends in reasons for breaches. The analysis will inform system wide pathway improvements and compliance to the national standards. Examples of potential changes to achieve this could be to support trusts in additional resources and processes and also to facilitate discuss between trusts for example in reaching agreement for diagnostics between trusts.

Examples of the types of reasons for this include;

a. Patients waiting excessively long period of time to seen of received treatment

b. Identification of 28 day standard exceptions - National guidance states patients who are diagnosed with cancer should be informed face to face, this would highlights numbers of patients who are not told in person by provider

c. Audits to review orphan records which require local providers to review local patients records

Record level data (pseudonymised) will be shared via NHS.net email accounts and access will be controlled by password protecting all files.

Access to the Cancer Wait Times (CWT) System will enable Cancer Alliances to undertake a wide range of locally determined and locally-specific analyses to support the Cancer Taskforce vision for improving services, care and outcomes for everyone with Cancer.

Only the lead organisation will directly access or download extracts from the Cancer Waiting Times system. Role Based Access Control prevents access to data downloads to employees outside of the analytical team responsible for producing outputs.

The CWT system is hosted by NHS Digital, access to and usage of the system is fully auditable. Users must comply with the use of the data as specified in this agreement. The CWT system complies with the requirements of NHS Digital Code of Practice on Confidential Information, the Caldicott Principles and other relevant statutory requirements and guidance to protect confidentiality.

Access to the CWT system will be granted to individual users only when a valid Data Usage Certificate (DUC) form is submitted to NHS Digital via the lead organisations Senior Information Risk Officer (SIRO), and where there is a valid Data Sharing Agreement between the lead organisation and NHS Digital.

Approved users will log into the system via a secure connection and will use a Single Sign-On (users are prompted to create a unique username and password).

Approved users will access:

a) Aggregate reports (which may include unsuppressed small numbers)

b) Pseudonymised record level data - users can directly download this data from the CWT system

c) I-View Plus tool (aggregated - access to produce graphs, charts/tabulations from the data through the construction of queries). This will give users access to run bespoke analysis on pre-defined measures and dimensions. It delivers the same data that is available through the reports and record level downloads (i.e. it will not contain patient identifiable data).

Any record level data extracted from the system will not be processed outside of the Data Controller or Data Processor unless otherwise specified in this agreement. Following completion of the analysis the record level data will be securely destroyed.

Users are not permitted to upload data into the system.

Data will only be shared with other members of the cancer alliance in aggregated form (without small number suppression).

Aggregate data/ graphical outputs may be shared via e-mail; for example as part of Alliance meeting papers.

Where record level data is shared with individual trusts these are shared only with trust(s) who were involved in the direct care of the patient, only via NHS.net email accounts.

As part of partnership working to improve Cancer Waiting Times performance, outputs may be shared with national/regional bodies including NHS Improvement and NHS England.

Training on the CWT system is not required as it is a data delivery system and it does not provide functionality to conduct bespoke detailed analysis. User guides are available for further assistance.

Access to the CWT system data is restricted to Cancer Alliance employees who are substantively employed by the Data Controller in fulfilment of their public health function.

The Cancer Alliances will use the data to produce a range of quantitative measures (counts, crude and standardised rates and ratios) that will form the basis for a range of statistical analyses of the fields contained in the supplied data.

Typical uses will include:

1) Analysis to support delivery of Cancer Waiting Times standard and identify variation, including clinical discussions to improve patient pathways

a. Comparative Cancer Waiting Times performance at tumour group and individual tumour site (i.e. ICD10 code) level for Trusts and CCGs. As well as the percentage of 62 Day performance, we will also need to look at number of activities, total numbers of patients treated, number of patients treated before and after Day 62

b. Analysis of Cancer Waiting Times performance by treatment modality to inform discussions

c. Grouping length of waits for standards to inform discussions on going beyond constitutional standards (e.g., activity and breach share by first seen trust and treatment trust, and by tumour site)

d. Analysis of derived breach reason fields to identify trends in reasons for delays.

e. To provide assurance through comparative analysis (e.g. orphan record identification, active monitoring proportions and validation of waiting list adjustments entered)

f. Analysis of flows of patients including analysis by provider trust site, by tumour site (e.g. median pathway durations, and the ability to track changes over time with "run charts" as per NHS Improvement requirements)

g. Outlier identification including exceptionally long waits to inform individual queries to providers

2) Cancer Waits analysis (not directly linked to constitutional standards) for the aim of identifying variation which may impact Cancer patients outcomes or patient experience. Examples for use of the data may include reviewing waits between surgery and radiotherapy for Head and Neck cancer patients with a maximum recommended wait of 6 weeks and using the data source to validate surgical numbers by provider trust.

Outputs fall into the following categories:

1) Analysis to support delivery of Cancer Waiting Times standard and identify variation, including clinical discussions to improve patient pathways

a. Comparative Cancer Waiting Times performance at tumour group and individual tumour site (i.e. ICD10 code) level for Trusts and CCGs.

b. Analysis of Cancer Waiting Times performance by treatment modality to inform discussions

c. Grouping length of waits for standards to inform discussions on going beyond constitutional standards

d. Analysis of derived breach reason fields to identify trends in reasons for delays.

e. To provide assurance through comparative analysis (e.g. orphan record identification, active monitoring proportions and validation of waiting list adjustments entered)

f. Analysis of flows of patients including analysis by provider trust site

g. Analysis of flows of patients across geography including analysis by tumour groups where specialist treatments are required only to be delivered by designated centres, aligning to Improving Outcome Guidance (NICE IOG)

h. Outlier identification including exceptionally long waits to inform individual queries to providers

2) Cancer Waits analysis (not directly linked to constitutional standards) for the aim of identifying variation which may impact Cancer patient􀍛s outcomes or patient experience. Examples for use of the data may include reviewing waits between surgery and radiotherapy for Head and Neck cancer patients with a maximum recommended wait of 6 weeks and using the data source to validate surgical numbers by provider trust.

1) Benefits type: Supporting delivery of CWT standards

The Cancer Waiting Times standards are key operational standards for the NHS, which aim to reduce the waits for diagnosis and treatment for Cancer patients, which will support improvements to survival rates and improve patient experience. This includes the new 28 day faster diagnosis standard being introduced as a standard from April 2021.

A key enabler to achieve these standards, and thus improve survival and patient experience is the role of Cancer Alliances locally to work with providers and commissioners to improve patient pathways. Access to the Cancer Waiting Times data as detailed in the above will enable Cancer Alliances to have informed discussions and allocate resources optimally to improve performance against these standards. It will also enable Cancer Alliances to work with local providers and commissioners to identify outliers against the standards, and mitigate the risk of similar delays for other patients.

Improvement would be expected on an on-going basis with standards already in place for nine standards:-

2 week wait urgent GP referral- 93%

2 week wait breast symptomatic -93%

31 day 1st treatment - 96%

31 day subsequent surgery-94%

31 day subsequent drugs-98%

31 day subsequent radiotherapy-94%

62 day (GP) referral to 1st treatment-85%

62 day (screening ) referral to 1st treatment-90%

62 day upgrade to 1st treatment locally agreed standard

In addition this access and use of data will be key in delivering the new 28 day faster diagnosis standard being

introduced from April 2021 (delayed from April 2020). Trusts are asked to ensure high level of data completeness for this item in 2019/20.

2) Benefits type: Improvements beyond constitutional standards

This access and resulting analysis will enable Cancer Alliances to undertake local analysis beyond the Cancer

Waiting times operational standards to support improvements to Cancer patients pathways beyond those already achieved by improving performance against standard set. This could include reviewing times between treatments, or treatment rates. The overall aim of this type of additional analysis would be to support improvements to Cancer patients survival and experience.

The Cancer Taskforce recommendation set out a number of ambitions to be met nationally and locally by 2020 including improving 1 year survival for Cancer to 75%, and improving the proportions of patients staged 1 or 2 to 62%. For both of these improvements to the diagnostic and treatment pathways are key, and require Cancer Alliances to be able to analyse the Cancer Waiting Times dataset to identify sub-optimum pathway and resulting improvements.

The overarching aim of all future analysis/outputs is to inform priorities and potential investment to improve Cancer pathways including reducing Cancer incidence and mortality, improving Cancer survival, improving patient experience, improving service efficiency and meeting national constitution standards relating to Cancer patients.

The Cancer Taskforce recommendation set out a number of ambitions to be met nationally and locally by 2020 including improving 1 year survival for Cancer to 75%, and improving the proportions of patients staged 1 or 2 to 62%. For both of these improvements to the diagnostic and treatment pathways are key, and require Cancer Alliances to be able to analyse the Cancer Waiting Times dataset to identify sub-optimum pathway and resulting improvements.

The publication of monthly and quarterly Official Statistics allows health organisations to assess performance

and benchmark, identifying outliers and policy developments required.

National timed pathways which promote best practice for patient care and faster diagnosis of cancer and noncancer patients have been developed based on performance of tumour groups identified in the Official Statistics.

The data enables local teams to analyse performance and initiate improvement work locally, and identify outliers which have resulted in performance improvement and improved patient care / experience.

The data also enables the ongoing identification of elements of good practice or challenge, for example the data is used to generate reports based around the 62-day intertrust referral pathways, identifying referral routes where the interprovider transfer is delayed either in referral using the 38-day interprovider transfer point as a marker, or treatment using the 24-day IPT to treatment point. This allows the alliance to support recovery where specific challenges are identified, improving future patients experience. The data enables breach analysis of the standards as well as breakdowns of treatment types and reports have been developed demonstrating completeness and compliance with the 28-day faster diagnosis standard, giving the opportunity to target areas of challenge during implementation.