Benchmarking Service to NHS organisations

University Hospitals Birmingham (UHB) NHS Foundation Trust requires access to national data sets for the purpose of providing objective insight and analysis that enables NHS organisations to deliver better services for patients.

The General Data Protection Regulation (GDPR) basis is that it is a public task to monitor healthcare services provided and it is in public interest that the care given is safe and effective. Namely:

a. Article 6 (lawfulness of processing):(e) Public task: the processing is necessary for you to perform a task in the public interest or for your official functions, and the task or function has a clear basis in law.

b. Article 9 (processing of special categories of personal data):(i) processing is necessary for reasons of public interest in the area of public health, such as protecting against serious cross-border threats to health or ensuring high standards of quality and safety of health care and of medicinal products or medical devices, on the basis of Union or Member State law which provides for suitable and specific measures to safeguard the rights and freedoms of the data subject, in particular professional secrecy.

The datasets held and requested under this Data Sharing Agreement (DSA) are: Hospital Episode Statistics (HES) Admitted Patient Care, Outpatients, Accident & Emergency, Emergency Care Data Set (ECDS), Critical Care and linked Civil registration deaths (secondary care cut) mortality data.

These data will be used to produce analytical outputs encompassing a vast range of metrics and topics pertinent to healthcare provision. These analytical outputs are used by NHS Organisations to understand their current performance and opportunities for improvement in order to target work plans strategically in an evidence based fashion.

For over a decade now, UHB has established a proven track record of using national data to provide analytical outputs that enable NHS Organisations to drive improvements to ultimately benefit patients. (Further details of examples are provided in section 5d.)

Such analytical outputs are provided via UHB’s online Healthcare Evaluation Data (HED) benchmarking platform and bespoke reports.

The HED platform provides access to a range of dashboards, alerts, reports and tools that allow NHS Organisations to benchmark across thousands of metrics.

The HED Tool is accessed online by third parties holding a subscription contract with UHB to use the HED system. Such third parties include:

1. NHS Organisations with various geographical locations across England including:

• NHS Acute Trusts (85%).

• Other NHS Organisations, including: NHS Integrated Care Boards (ICBs), NHS Commissioning Support Units (CSUs), Quality Observatories with legal entity under a NHS Trust For example; North East Quality Observatory (NEQOS) is hosted by the legal entity Cumbria, Northumberland, Tyne and Wear NHS Foundation Trust (12%).

2. Non-NHS Organisations such as Pricewaterhouse Coopers (PwC) & Grant Thornton (3%).

For any future non-NHS subscribers, UHB will inform NHS England in any renewal or amendment to this agreement.

NHS Organisations use the analytical outputs to efficiently make better informed decisions as to how they can make sustainable service improvements for the benefit of patients. The system allows NHS Organisations to actively understand performance trends and influences in order to identify improvement opportunities. The range of dashboards, alerts, reports and tools UHB create using these national data sets allow NHS Organisations to actively interrogate benchmarked and risk adjusted metrics in insightful ways. The insight provided is of significant value to NHS Organisations as it isn’t possible to produce such intelligence when solely using local information sources.

Non-NHS Organisations work within the healthcare space and are only permitted access to the HED tool when working to support NHS Organisations. Non-NHS Organisations only have access to aggregate, small number suppressed data in line with the HES Analysis guide. UHB has reviewed the NHS standard (ISB 1523) relating to anonymisation and can confirm compliance. The rationale for allowing Non-NHS Organisations access to the tool is that some NHS Organisations require additional, short-notice, specialist resource in their work.

Direct access enables:

• Staff to be equipped to provide immediate support to NHS Organisations, reducing lead time for projects which can be costly for the NHS, where it has already been determined that out-sourcing is required.

• An efficient way to disseminate information in support of NHS improvement work. Without which UHB would be actively encouraging NHS Organisations to export data outside the NHS.

• Autonomy when undertaking commissioned independent work beneficial to the NHS e.g. the Keogh Review.

The sole purpose of the work associated with this agreement is to support NHS Organisations (including UHB) in the on-going monitoring of clinical quality, patient safety and organisational effectiveness in order to deliver better services for patients.

To deliver the stated objective a wide range of healthcare indicators are calculated. The objective of having as wide an array of relevant indicators as possible is to give NHS managers and clinicians as complete a picture of hospital performance across topics such as mortality, readmissions, length of stay, emergency care, coding quality, etc. HES-Civil registration deaths (secondary care cut). linked data specifically will be used within this work to look at outcomes analysis and form analytical overviews relating to post-discharge mortality. Such overviews relate to standardised post-discharge mortality monitoring within distinct clinical cohorts and bespoke long-term survival monitoring. This work will increase the understanding of complete pathways of care.

Within this agreement, UHB are applying for Local Patient Identifier within Provider (LOPATIDs) a code which uniquely identifies an individual patient locally within a hospital that has provided that patient with treatment or care for the national datasets. There is a legal basis in place for this under Section 251 of the National Health Service Act 2006.

The LOPATID field will only be provided to NHS Hospital Staff, for patients that have been treated by their organisation, where the Caldicott Guardian of that organisation has given their approval.

LOPATID is required as:

• UHB via HED delivers a national benchmarking system that can provide assurance to NHS Hospitals that they are providing safe and high quality care and treatment while signposting them to any areas of concern. The HED system provides additional information to that available locally within a hospital trust by enabling the swift observation of patient cohorts that are statistical outliers and warrant further examination.

• Access to LOPATID and the ability to cross-reference local hospital records is an essential part of governance activities within a Hospital Trust. It enables hospitals to efficiently conduct internal governance audits on a case-by-case basis and therefore focus their finite attention where it is most needed. This directly enables organisations to monitor and manage services for the greater benefit of patients.

• Access to LOPATID is also critical to enable health care professionals to audit their data and clinical practice. This review of individual patient outcomes and experience can evidence the care provided is of a good quality and safe, and also provide assurance to trust boards. In addition, access to identifiers will also enable clinicians to review and audit deaths attributed to them in national mortality models.

• In summary, without this level of information, Trusts would not be able to conduct further analysis to determine the root cause, i.e., detailed reason, for any concerns identified in the high-level information. Such root-cause analyses serve to provide assurance internally and externally across a range of crucial measures.

To address the GDPR principle of data minimisation, UHB have only requested data that is deemed necessary to achieve the purpose described within this Agreement. In particular:

• Regarding identifiable data: UHB destroy patient identifiers for data older than 3 years.

• Regarding time periods held: The years of data required allow for organisations to perform multiple functions, such as being able to demonstrate service improvement over time and visualise trends.

• Regarding national data: The datasets requested under this DSA include data for the whole of England. This is required because when performing analyses, organisations need to be able to select peers based on case mix for more accurate benchmarking to assist with service improvement, and these organisations may not be local. Valid peers may also change for a selected service or indicator. As an example, some HED system users are regional providers of a specific service and others are specialised hospitals which require benchmarking across the whole of the UK. Furthermore, the current user base for the HED tool is spread across multiple regions of England and each organisation requires access to their own data. National data is also required to allow the calculation of standardised metrics.

There are no alternative less intrusive ways of achieving the purpose. National and peer specific benchmarking along with risk adjust metrics can only be provided using a national data set. As such it is in the interest of the public to use an established data set that is already being routinely submitted centrally.

The HED tool is produced and owned by UHB, an NHS Foundation Trust and registered charity. This work is commissioned and funded on an on-going basis by the sole data controller who also process data, UHB.

Income from subscriptions to the HED system goes back into the UHB Trust budget and so is subject to standard NHS financial governance.

All outputs will contain only data that is aggregated with small numbers suppressed in line with the HES Analysis Guide.

The outputs produced as a result of data processing are benchmarked or standardised healthcare indicators and analysis encompassing a wide range of topics pertinent to the provision of high quality healthcare within the NHS.

Such topics include: clinical quality, mortality monitoring, readmissions monitoring, patient safety, data quality, operational efficiency, and discharge and admission trends.

Such outputs are provided either via:

• the range of dashboards, alerts, reports and tools made available within the HED system, or

• bespoke reports including clinically-led, retrospective, small number suppressed reports that may be published in peer review journals

The outputs provided via the HED system will then be further utilised by end-users of the system, as described in section 5a, to produce novel outputs for routine service management/evaluation, audit etc. within the NHS. For example, Hospital Trusts using the system will use the outputs provided via HED to produce their own local analysis, dashboards, and reports including board reports.

Work streams are on-going and the outputs already form part of various NHS organisations’ monthly routine reporting and governance assurance processes.

For example, each month as the national data sets are released by NHS England, UHB then undertake the processing activities described in order to produce the range of benchmarked and standardised indicators described above. These outputs will then be used, both locally at UHB, and by NHS Organisations using the HED system, to produce their own local analysis, dashboards, and reports including board reports.

Each month the range of dashboards, alerts, reports and tools made available within the HED system allows NHS Organisations to identify areas of potential concern within their organisation. Such areas are then investigated further. When appropriate and authorised, the ability to identify a cohort of patients within the organisation and undertake appropriate clinical case note review is invaluable as a part of ensuring good hospital governance.

Some examples of yielded benefits include:

1) Several large acute trusts use the HED mortality modules to monitor and improve their trust’s mortality. Patient groups are identified that demonstrate persistent deviation from the expected mortality levels as they reach a pre-set trigger point. Consequently, the trust’s response to mortality reporting is much more instantaneous, allowing clinical and governance staff to review the deaths, put any actions in motion and share the outcome with the CQC considerably before any official regulatory notification.

Former NHS Medical Director Sir Bruce Keogh described mortality statistics as “smoke signals” for potential problems with care. Using Statistical Process Control (SPC) techniques, these signals are generated. As such, the benefit is derived from the ability to effectively use the data to highlight and investigate potential areas for further clinical review. This enables focused use of limited clinical resources, i.e. increased efficiency, which in itself benefits patients and the general public. This benefit has been experienced by many trusts using HED as they report the significant amount of time they have saved because of HED enabling them to pin-point priority areas, and automatically producing reports, increasing the time available for other hospital priorities.

Additionally, patients benefit as focused clinical review increases the safety and quality of the healthcare services provided.

In one acute Trust this use of HED mortality modules to undertake root cause analysis of patients has ultimately seen them move from special measures to a Care Quality Commission (CQC) rating of ‘Good’.

2) Several acute Trusts use HED to help optimise their services. One acute Trust in the South of England has used HED to generate Service Line Benchmarking packs for clinicians and service managers to help drive service transformation. These packs use the breadth and depth of information available in HED to allow service lines to examine their data and highlight potential opportunities. A measurable benefit from this was that the trauma and orthopaedic team were surprised to find 50% of local activity was being carried out by other providers outside of the main NHS teaching hospital. Consequently, referral to treatment times were targeted by the orthopaedic team working with radiology to develop a one-stop outpatient service where patients referred by their GPs can have their imaging and initial discussions on one day – and receive a same-day decision about whether they are suitable for surgery.

This change benefited patients as they were seen quicker, only had to make one trip to the hospital, and received a decision regarding their surgery faster.

Furthermore, HED delivers the ability for this information to be available to different staff groups and tailored to their individual needs so reducing reliance on local Information support staff.

3) One acute trust specifically used HED to investigate their performance of non-elective maternity re-admissions within 42 days of delivery as raised by the CQC. Analysis of clinical coding using HED identified that they had poor recording practices and evidenced that their subsequent change in coding more accurately represented their re-admission rate. The ability to use HED to undertake such analysis and so improve the accuracy of coding and therefore reported metrics, benefits the health service as accurate information is required to make high quality management decisions when running services.

Improving the accuracy of coding and reported metrics benefits patients as: a) it is important for them to have access to accurate information relating to the quality of the services they may receive, and b) high quality decisions taken using high quality information is critical to high quality services being provided to patients.

4) A specialist acute hospital was finding it difficult to analyse market share for one of their sites. As a specialist provider they required a specific set of hospitals to benchmark against. Using HED they have been able to isolate the Trusts and sites, so they are true peers. HED has helped them drive up their market share, thus promoting competition between hospitals, which in turn drives better care and service across the geographical area so benefiting patients.

5) Commissioning organisations using information from the HED system have reported back the ease with which their non-technical users can quickly pull-out information whilst simultaneously their analysts have the flexibility to answer more detailed questions. This diversity of topics and ease of use serves to spread the use of high-quality information empowering commissioners whilst allowing the analytical resource to be focused where required. Having effective, efficient, and evidence-based commissioning of health services directly benefits patients and the public.

This key reporting and case note review process is used by hospitals up and down the country. Without using the NHS England data it would be impossible to deliver these sorts of benefits to the NHS.

The above yielded benefits are ongoing monthly practices.

University Hospitals Birmingham (UHB) NHS Foundation Trust requires access to national data sets for the purpose of providing objective insight and analysis that enables NHS organisations to deliver better services for patients.

The General Data Protection Regulation (GDPR) basis is that it is a public task to monitor healthcare services provided and it is in public interest that the care given is safe and effective. Namely:

a. Article 6 (lawfulness of processing):(e) Public task: the processing is necessary for you to perform a task in the public interest or for your official functions, and the task or function has a clear basis in law.

b. Article 9 (processing of special categories of personal data):(i) processing is necessary for reasons of public interest in the area of public health, such as protecting against serious cross-border threats to health or ensuring high standards of quality and safety of health care and of medicinal products or medical devices, on the basis of Union or Member State law which provides for suitable and specific measures to safeguard the rights and freedoms of the data subject, in particular professional secrecy.

The datasets held and requested under this Data Sharing Agreement (DSA) are: Hospital Episode Statistics (HES) Admitted Patient Care, Outpatients, Accident & Emergency, Emergency Care Data Set (ECDS), Critical Care and linked Civil registration deaths (secondary care cut) mortality data.

These data will be used to produce analytical outputs encompassing a vast range of metrics and topics pertinent to healthcare provision. These analytical outputs are used by NHS Organisations to understand their current performance and opportunities for improvement in order to target work plans strategically in an evidence based fashion.

For over a decade now, UHB has established a proven track record of using national data to provide analytical outputs that enable NHS Organisations to drive improvements to ultimately benefit patients. (Further details of examples are provided in section 5d.)

Such analytical outputs are provided via UHB’s online Healthcare Evaluation Data (HED) benchmarking platform and bespoke reports.

The HED platform provides access to a range of dashboards, alerts, reports and tools that allow NHS Organisations to benchmark across thousands of metrics.

The HED Tool is accessed online by third parties holding a subscription contract with UHB to use the HED system. Such third parties include:

1. NHS Organisations with various geographical locations across England including:

• NHS Acute Trusts (80%).

• Other NHS Organisations, including: NHS Integrated Care Boards (ICBs), NHS Commissioning Support Units (CSUs), Quality Observatories with legal entity under a NHS Trust For example; North East Quality Observatory (NEQOS) is hosted by the legal entity Cumbria, Northumberland, Tyne and Wear NHS Foundation Trust (18%).

2. Non-NHS Organisations such as Pricewaterhouse Coopers (PwC) (2%).

For any future non-NHS subscribers, UHB will inform NHS Digital upfront and include specific details in any renewal or amendment to this agreement.

NHS Organisations use the analytical outputs to efficiently make better informed decisions as to how they can make sustainable service improvements for the benefit of patients. The system allows NHS Organisations to actively understand performance trends and influences in order to identify improvement opportunities. The range of dashboards, alerts, reports and tools UHB create using these national data sets allow NHS Organisations to actively interrogate benchmarked and risk adjusted metrics in insightful ways. The insight provided is of significant value to NHS Organisations as it isn’t possible to produce such intelligence when solely using local information sources.

PwC work within the healthcare space and are only permitted access to the HED tool when working to support NHS Organisations. PWC only have access to aggregate, small number suppressed data in line with the HES Analysis guide. UHB has reviewed the NHS standard (ISB 1523) relating to anonymisation and can confirm compliance. The rationale for allowing PwC access to the tool is that some NHS Organisations require additional, short-notice, specialist resource in their work.

Direct access enables:

• PWC staff to be equipped to provide immediate support to NHS Organisations, reducing lead time for projects which can be costly for the NHS, where it has already been determined that out-sourcing is required.

• An efficient way to disseminate information in support of NHS improvement work. Without which UHB would be actively encouraging NHS Organisations to export data outside the NHS.

• Autonomy when undertaking commissioned independent work beneficial to the NHS e.g. the Keogh Review.

The sole purpose of the work associated with this agreement is to support NHS Organisations (including UHB) in the on-going monitoring of clinical quality, patient safety and organisational effectiveness in order to deliver better services for patients.

To deliver the stated objective a wide range of healthcare indicators are calculated. The objective of having as wide an array of relevant indicators as possible is to give NHS managers and clinicians as complete a picture of hospital performance across topics such as mortality, readmissions, length of stay, emergency care, coding quality, etc. HES-Civil registration deaths (secondary care cut). linked data specifically will be used within this work to look at outcomes analysis and form analytical overviews relating to post-discharge mortality. Such overviews relate to standardised post-discharge mortality monitoring within distinct clinical cohorts and bespoke long-term survival monitoring. This work will increase the understanding of complete pathways of care.

Within this agreement, UHB are applying for Local Patient Identifier within Provider (LOPATIDs) a code which uniquely identifies an individual patient locally within a hospital that has provided that patient with treatment or care for the national datasets. There is a legal basis in place for this under Section 251 of the National Health Service Act 2006.

The LOPATID field will only be provided to NHS Hospital Staff, for patients that have been treated by their organisation, where the Caldicott Guardian of that organisation has given their approval.

LOPATID is required as:

• UHB via HED delivers a national benchmarking system that can provide assurance to NHS Hospitals that they are providing safe and high quality care and treatment while signposting them to any areas of concern. The HED system provides additional information to that available locally within a hospital trust by enabling the swift observation of patient cohorts that are statistical outliers and warrant further examination.

• Access to LOPATID and the ability to cross-reference local hospital records is an essential part of governance activities within a Hospital Trust. It enables hospitals to efficiently conduct internal governance audits on a case-by-case basis and therefore focus their finite attention where it is most needed. This directly enables organisations to monitor and manage services for the greater benefit of patients.

• Access to LOPATID is also critical to enable health care professionals to audit their data and clinical practice. This review of individual patient outcomes and experience can evidence the care provided is of a good quality and safe, and also provide assurance to trust boards. In addition, access to identifiers will also enable clinicians to review and audit deaths attributed to them in national mortality models.

• In summary, without this level of information, Trusts would not be able to conduct further analysis to determine the root cause, i.e., detailed reason, for any concerns identified in the high-level information. Such root-cause analyses serve to provide assurance internally and externally across a range of crucial measures.

To address the GDPR principle of data minimisation, UHB have only requested data that is deemed necessary to achieve the purpose described within this Agreement. In particular:

• Regarding identifiable data: UHB destroy patient identifiers for data older than 3 years.

• Regarding time periods held: The years of data required allow for organisations to perform multiple functions, such as being able to demonstrate service improvement over time and visualise trends.

• Regarding national data: The datasets requested under this DSA include data for the whole of England. This is required because when performing analyses, organisations need to be able to select peers based on case mix for more accurate benchmarking to assist with service improvement, and these organisations may not be local. Valid peers may also change for a selected service or indicator. As an example, some HED system users are regional providers of a specific service and others are specialised hospitals which require benchmarking across the whole of the UK. Furthermore, the current user base for the HED tool is spread across multiple regions of England and each organisation requires access to their own data. National data is also required to allow the calculation of standardised metrics.

There are no alternative less intrusive ways of achieving the purpose. National and peer specific benchmarking along with risk adjust metrics can only be provided using a national data set. As such it is in the interest of the public to use an established data set that is already being routinely submitted centrally.

The HED tool is produced and owned by UHB, an NHS Foundation Trust and registered charity. This work is commissioned and funded on an on-going basis by the sole data controller who also process data, UHB.

Income from subscriptions to the HED system goes back into the UHB Trust budget and so is subject to standard NHS financial governance.

All outputs will contain only data that is aggregated with small numbers suppressed in line with the HES Analysis Guide.

The outputs produced as a result of data processing are benchmarked or standardised healthcare indicators and analysis encompassing a wide range of topics pertinent to the provision of high quality healthcare within the NHS.

Such topics include: clinical quality, mortality monitoring, readmissions monitoring, patient safety, data quality, operational efficiency, and discharge and admission trends.

Such outputs are provided either via:

• the range of dashboards, alerts, reports and tools made available within the HED system, or

• bespoke reports including clinically-led, retrospective, small number suppressed reports that may be published in peer review journals

The outputs provided via the HED system will then be further utilised by end-users of the system, as described in section 5a, to produce novel outputs for routine service management/evaluation, audit etc. within the NHS. For example, Hospital Trusts using the system will use the outputs provided via HED to produce their own local analysis, dashboards, and reports including board reports.

Work streams are on-going and the outputs already form part of various NHS organisations’ monthly routine reporting and governance assurance processes.

For example, each month as the national data sets are released by NHS Digital, UHB then undertake the processing activities described in order to produce the range of benchmarked and standardised indicators described above. These outputs will then be used, both locally at UHB, and by NHS Organisations using the HED system, to produce their own local analysis, dashboards, and reports including board reports.

Each month the range of dashboards, alerts, reports and tools made available within the HED system allows NHS Organisations to identify areas of potential concern within their organisation. Such areas are then investigated further. When appropriate and authorised, the ability to identify a cohort of patients within the organisation and undertake appropriate clinical case note review is invaluable as a part of ensuring good hospital governance.

Some examples of yielded benefits include:

1) Several large acute trusts use the HED mortality modules to monitor and improve their trust’s mortality. Patient groups are identified that demonstrate persistent deviation from the expected mortality levels as they reach a pre-set trigger point. Consequently, the trust’s response to mortality reporting is much more instantaneous, allowing clinical and governance staff to review the deaths, put any actions in motion and share the outcome with the CQC considerably before any official regulatory notification.

Former NHS Medical Director Sir Bruce Keogh described mortality statistics as “smoke signals” for potential problems with care. Using Statistical Process Control (SPC) techniques, these signals are generated. As such, the benefit is derived from the ability to effectively use the data to highlight and investigate potential areas for further clinical review. This enables focused use of limited clinical resources, i.e. increased efficiency, which in itself benefits patients and the general public. This benefit has been experienced by many trusts using HED as they report the significant amount of time they have saved because of HED enabling them to pin-point priority areas, and automatically producing reports, increasing the time available for other hospital priorities.

Additionally, patients benefit as focused clinical review increases the safety and quality of the healthcare services provided.

In one acute Trust this use of HED mortality modules to undertake root cause analysis of patients has ultimately seen them move from special measures to a Care Quality Commission (CQC) rating of ‘Good’.

2) Several acute Trusts use HED to help optimise their services. One acute Trust in the South of England has used HED to generate Service Line Benchmarking packs for clinicians and service managers to help drive service transformation. These packs use the breadth and depth of information available in HED to allow service lines to examine their data and highlight potential opportunities. A measurable benefit from this was that the trauma and orthopaedic team were surprised to find 50% of local activity was being carried out by other providers outside of the main NHS teaching hospital. Consequently, referral to treatment times were targeted by the orthopaedic team working with radiology to develop a one-stop outpatient service where patients referred by their GPs can have their imaging and initial discussions on one day – and receive a same-day decision about whether they are suitable for surgery.

This change benefited patients as they were seen quicker, only had to make one trip to the hospital, and received a decision regarding their surgery faster.

Furthermore, HED delivers the ability for this information to be available to different staff groups and tailored to their individual needs so reducing reliance on local Information support staff.

3) One acute trust specifically used HED to investigate their performance of non-elective maternity re-admissions within 42 days of delivery as raised by the CQC. Analysis of clinical coding using HED identified that they had poor recording practices and evidenced that their subsequent change in coding more accurately represented their re-admission rate. The ability to use HED to undertake such analysis and so improve the accuracy of coding and therefore reported metrics, benefits the health service as accurate information is required to make high quality management decisions when running services.

Improving the accuracy of coding and reported metrics benefits patients as: a) it is important for them to have access to accurate information relating to the quality of the services they may receive, and b) high quality decisions taken using high quality information is critical to high quality services being provided to patients.

4) A specialist acute hospital was finding it difficult to analyse market share for one of their sites. As a specialist provider they required a specific set of hospitals to benchmark against. Using HED they have been able to isolate the Trusts and sites, so they are true peers. HED has helped them drive up their market share, thus promoting competition between hospitals, which in turn drives better care and service across the geographical area so benefiting patients.

5) Commissioning organisations using information from the HED system have reported back the ease with which their non-technical users can quickly pull-out information whilst simultaneously their analysts have the flexibility to answer more detailed questions. This diversity of topics and ease of use serves to spread the use of high-quality information empowering commissioners whilst allowing the analytical resource to be focused where required. Having effective, efficient, and evidence-based commissioning of health services directly benefits patients and the public.

This key reporting and case note review process is used by hospitals up and down the country. Without using the NHS Digital data it would be impossible to deliver these sorts of benefits to the NHS.

The objective is to provide quality and benchmarking analysis that will enable NHS organisations to deliver better services for patients.

Such analysis is solely provided either: via the online Healthcare Evaluation Data (HED) tool, or via bespoke reports.

This work is commissioned and funded on an ongoing basis by the data controller, University Hospitals Birmingham NHS Foundation Trust (UHB) and produced by the Health Informatics Department within the Hospital.

The sole objective of this work it is to support both UHB and other NHS Trusts and commissioners in the ongoing monitoring of clinical quality and organisational effectiveness. This purpose is fulfilled either:

a) Directly, the NHS Trust holds a subscription to use the HED system,

b) Indirectly, analytics are provided via a non-NHS organisation, e.g. Pricewaterhouse Coopers (PwC), who hold a subscription to use aggregate small number suppressed data within the system only with NHS organisations. UHB has reviewed the NHS standard (ISB 1523) relating to anonymisation and can confirm the systems are compliant with this.

c) through clinically-led bespoke reports based on retrospective, pseudonymised data which if requested are published in peer review journals with small numbers suppressed.

The GDPR basis is:

a. Article 6 (lawfulness of processing):

(e) Public task: the processing is necessary for you to perform a task in the public interest or for your official functions, and the task or function has a clear basis in law.

b. Article 9 (processing of special categories of personal data):

(i) processing is necessary for reasons of public interest in the area of public health, such as protecting against serious cross-border threats to health or ensuring high standards of quality and safety of health care and of medicinal products or medical devices, on the basis of Union or Member State law which provides for suitable and specific measures to safeguard the rights and freedoms of the data subject, in particular professional secrecy.

All subscriber organisations are NHS Organisations, except for two.

The two non-NHS organisations are PwC and Carnall Farrar.

These non-NHS organisations work within the healthcare space and have access to the system solely for the purpose of assisting NHS organisations. They will only access anonymous data, i.e. aggregate level data with small numbers suppressed in line with HES Analysis guide.

For any future non-NHS subscribers, UHB would seek permission from NHS Digital first and update the DSA accordingly.

NHS organisations are limited to the below list only. The geographical range of these organisations is spread across England.

• NHS Trusts (between 50 to 75 hospitals)

• Clinical Commissioning Groups (CCG - Less than 10)

• Commissioning Support Units (CSU - Less than 10)

• NHS England and NHS Improvement

• Quality Observatories (QO - Less than 10)

AQUA – legal entity is under Salford Royal NHS Foundation Trust

NEQOS – legal entity is under Northumberland Tyne and Wear and South Tees NHS Foundation Trusts

• Non-NHS Organisations - (PwC and Carnall Farrar)

Patient data is required for all England, for two reasons. Firstly, because when performing analyses, organisations need to be able to select peers based on case mix for more accurate benchmarking to assist with service improvement and these organisations may not be local or regional based. This is true for both UHB and customers of the UHB HED analytical tool. As an example, some HED customers are regional providers of a specific service and others are specialised hospitals which require benchmarking across the whole of the UK. Secondly, the current client base for the HED tool is spread across multiple regions of England and therefore require access to their own data. Further, data for all organisations is required to allow for valid peers to be selected for a given service and indicator. National data is also required to allow the calculation of standardised metric. The years of data required allow for organisations to perform multiple functions, such as being able to demonstrate service improvement over time and visualize trends.

To address the GDPR principle of data minimisation, UHB have only requested data that is deemed necessary to achieve the purpose described within this Agreement. Identifiable data is deleted each month for the period in question. UHB do not hold identifiable data older than 3 years.

To deliver the stated objective a wide range of healthcare indicators are calculated (over 100) and as such various HES data sets including Admitted Patient Care, Outpatients, Accident & Emergency as well as linked mortality data are required. This includes access to the Emergency Care Data Set (ECDS). The objective in having as wide an array of relevant indicators as possible is to give NHS managers and clinicians as complete a picture of hospital performance as possible. Therefore, the whole dataset is needed and cannot be compressed to certain fields.

HES-Civil registration mortality linked data specifically will be used within this work to look at outcomes analysis and form analytical overviews relating to post-discharge mortality. Such overviews relate to standardised post-discharge mortality monitoring within distinct clinical cohorts and bespoke long-term survival monitoring. This work will increase the understanding of complete pathways of care. Any analysis produced using Civil Registration mortality data will not be made available to non-NHS organisations.

Local patient identifier (LOAPTID) is required within the HED system for direct patient care and there is a legal basis in place under Section 251 of the National Health Service Act 2006. Subscribers will only be able to access this record level Identifiable data for their own patient records (i.e. patients within their Trust).

HED delivers a national benchmarking system that can provide assurance to hospitals they are providing safe and high-quality care and treatment, or signpost areas of concern. The HED system enables Trusts to easily identify particular patient cohorts that are statistical outliers and warrant further examination. Local patient identifiers are an essential requirement linking areas that require investigation, to hospital records. Without them Trusts would not be able to identify their patients and conduct root-cause analyses both for internal governance and also to provide assurance to external regulatory authorities across a range of key indicators i.e. HSMR, SHMI. Historically, the Secretary of State for Health had ordered a review of avoidable deaths. Ensuring Trusts have access to their local patient identifiers through HED enables them to conduct these audits and therefore focuses attention on eradicating mortalities that could have been prevented.

HED does not just include HES and Mortality datasets. Other data sources used within the HED system are Healthcare Resource Group (HRG) National Tariff, Venous Thromboembolism Risk Assessment data, publicly available NHS England datasets, National Reporting and Learning System (NRLS) patient safety incident reports, PROMS, Safety Thermometer and Infection Control. There is no international data within HED. Furthermore, HES/ECDS/Civil Registration mortality data is not linked to international data and is not used outside of England/Wales.

It is essential for root-cause analysis that patients can be considered on a case-by-case basis. The ability to be able to identify patients via HED and then subsequently interact with other datasets and clinical notes held locally is vital to detect required clinical quality improvements. A specific example of this is via Mortality reviews, where HED directly enables organisations to monitor and manage services so that no avoidable harm comes to patients whilst in their care.

HED specifically empowers healthcare managers and clinicians to measure patient experience and outcomes benchmarked against their peers (both local and nationally) e.g. Length of stay, Mortality, Readmissions. This information is not available locally and delivers clinically relevant outcome data and comparative information to clinicians. Access to local patient identifiers is critical to enable health care professionals to audit their data and clinical practice. This review of individual patient outcomes and experience can evidence the care provided is of a good quality and safe, and also provide assurance to trust boards. In addition, access to identifiers will also enable clinicians to review and audit deaths attributed to them in national mortality models.

Record-level patient information is only available to organisations who deliver the care. A protocol including Caldicott authorisation form has been reviewed previously by DAAG (DAAG reference: 240412-a) for controlling access to such sensitive items. This established process ensures that access to sensitive items is restricted to authorised hospital trust staff and was found to be robust during a recent NHS Digital Audit.

The rationale for allowing PwC access to aggregate-level small-number suppressed analytics:

As some NHS organisations require additional specialist resource to deliver the benefits of using benchmarking information, therefore subscription to the HED tool is required by PwC as:

1. This enables PwC to have people equipped to provide immediate support to NHS organisations.

2. Providing PwC with aggregate-level information via the tool is the most efficient way of disseminating information in support of this work – the alternative described directly below would clearly create large inefficiencies.

3. It allows such organisations to be autonomous in undertaking work that requires a level of independence and is beneficial to the NHS e.g. the Keogh review. In this instance PwC were commissioned to complete this review independently of any engagement of NHS Trusts involved. It would therefore have been inappropriate for them to ask the Hospitals for the information required to undertake this review.

The rationale for allowing Carnall Farrar access to aggregate-level small-number-supressed analytics:

Carnall Farrar is a management consultancy dedicated to improving health, care and public services. They identify where clinical improvements can be made and the best ways to make change happen. This leads to better use of the scarce NHS resources of workforce, facilities and beds. Carnall Farrar’s objective is to look for clinical solutions to streamline and improve patient care, allowing more people access to a quality service.

NHS clients of Carnall Farrar have to meet the costs associated with collecting, processing and analysing that data. The ability for Carnall Farrar to access analysis directly within HED would streamline the data collection process and save time, therefore lowering the cost for the NHS organisation and saving NHS management teams the time it takes to respond to data requests. The ability to undertake such clinical improvement projects allows patient care to be streamlined and improved so allowing more people to receive the best quality clinical care.

Please note: The level of access given to PwC and Carnall Farrar will be the same i.e. same levels of security and censoring.

An alternative would be for NHS organisations working with PwC and Carnall Farrar to provide data directly to the non-NHS organisations. This arrangement would have the following detrimental effect on the NHS:

1. It would actively be encouraging NHS-organisations to export data at varying levels (while only aggregate-level small-number-supressed analytics will be provided) from the HED system and send it to non-NHS organisations. By these two non-NHS organisations having direct access UHB can monitor which modules are accessed when and by whom. This negates the need for NHS organisations to export isolated aggregate data and email it outside the NHS.

2. It would introduce a longer lead time for projects which would ultimately cost the NHS more.

The sole outputs are benchmarked or standardised healthcare indicators such as measures of mortality, survival, discharge and admission trends, readmissions, length of stay, patient safety etc.

Such outputs are solely provided either via:

• the range of Dashboards and Modules made available within the HED system, or

• aggregate small number censored reports

Within the HED system:

Dashboards will only contain aggregate level information. Modules can contain aggregate level information and low-level information. As explained, the level of data that can be viewed within a module depends on the access level of the named individual user and which organisation they are working for. As such, access to low-level information, including small numbers, is strictly controlled in line with the access controls outlined in the above section on ‘Processing activities’

Outputs are to be used solely for the purpose of assisting the NHS.

Outputs will be used by NHS Clinicians and Managers to:

• Assure and manage clinical quality and patient safety within NHS Organisations

• Identify trends requiring a clinical review of patient pathways. (Hospital based users with Caldicott approval are able to investigate nationally standardised metrics and ‘drill-down’ to patient level information for patients treated at that trust, including local patient identifiers, in order to conduct clinical case note review and route cause analysis)

• Increase the understanding of patient outcomes

• Identify potential areas for improvement in clinical quality or operational efficiency either within a Hospital or a local healthcare economy

• Identify areas of best practice either within hospital Trusts or local healthcare economies

• Provide consultants with the information necessary for consultant revalidation

All of the above will serve to increase the understanding of patient outcomes in regard to quality, safety, productivity and efficiency benchmarking within the NHS.

As mentioned, aggregate-level small-number-suppressed analytics produced for the HED tool are being made available to two non-NHS organisations.

In the case of PwC, UHB are working with this subscriber to support NHS organisations. Some examples of the specific outcomes of this access are:

• Aggregate-level analytics have been used to undertake due diligence for both Monitor and CQC. A good example being production of the Keogh review information packs. This information was vital to the extensive work undertaken as part of the Keogh review which has culminated in the majority of the hospitals originally identified being taken out of special measures by the CQC.

• A further example is use of aggregate level analytics on elderly care readmissions with a CCG in order to support work helping them to understand how they could reduce avoidable readmissions within their region.

In the case of Carnall Farrar, the specific output is continuation of the development and implementation of local solutions that address long-standing challenges within clinical service delivery and the achievement of financial targets. This support is provided to STPs, hospital Trusts and CCGs.

It is not appropriate to set a target date for this work as the work streams are ongoing and the outputs already form part of various NHS organisations’ monthly reporting and governance assurance processes.

Each month a range of modules are produced which allow NHS hospitals to identify areas of potential concern within their organisation. Such areas are then investigated further. When appropriate and authorised, the ability to identify a cohort of patients within the organisation and undertake appropriate clinical case note review is invaluable as a part of ensuring good hospital governance.

• University Hospitals Birmingham (UHB) NHS Foundation Trust’s Clinical Quality Monitoring Group use the HED CUSUM model as a continual assessment of mortality. Diagnosis groups are identified that demonstrate persistent deviation from the expected mortality levels and reach a pre-set trigger point. Consequently, the trust approach to mortality is proactive rather than reactive and enables UHB to review the deaths and share the outcome with CQC considerably before the official regulatory notification. Former NHS Medical Director Sir Bruce Keogh described mortality statistics as “smoke signals” for potential problems with care. Using Statistical Process Control (SPC) techniques, these signals are only generated by ‘out-of control’ organisations, both significantly higher or lower than expected. As such, the benefit is derived from the ability to effectively use the numbers to highlight and investigate potential areas for further clinical review.

• George Eliot Hospital NHS Trust have used HED mortality modules to undertake root cause analysis of patients which has ultimately seen them move from special measures to a CQC rating of ‘Good’.

• Royal Liverpool and Broadgreen University Hospital NHS Trust uses HED to report benchmarking and mortality to their Trust Board. A measurable benefit they have observed, and that is echoed in other Trusts, is the significant amount of time they have saved with HED automatically producing these reports for them, freeing their time for other hospital priorities.

• Calderdale and Huddersfield have used HED to look at their clinical coding to evidence areas of improvement in data quality and ensure they are not over- or under-coding compared to local and national peers. The system has provided reassurance to the Calderdale and Huddersfield NHS Outpatients (CHFT) Exec Board and Local Commissioning Groups while providing benchmarking insight into coding within different specialty areas.

• HED provides direct intelligent feedback to service areas that can change hospital performance for the better. However, this is not the only tool that will influence behaviour in a hospital, so it is difficult to attribute changes solely to HED. As a marker of the sizeable impact, it would have were HED not able to provide this information, Derby Teaching Hospitals NHS Foundation Trust, Royal Cornwall Hospital NHS Trust, Bradford Teaching Hospitals NHS Foundation Trust and The Royal Wolverhampton NHS Trust are just a few examples of the many Trusts that have contacted HED over the last few months unable to access LOPATID information for case note reviews. Hospitals who are unable to access LOPATID have noticed instantly the impact this missing data is having on their ability to interrogate and improve their trust’s mortality levels.

• Plymouth Hospitals NHS Trust have used HED to generate Service Line Benchmarking packs for clinicians and service managers to help inform decisions about transformation. The packs bring together data on finances, operational efficiency, safety and quality indicators benchmarked against relevant peers. This allows service lines to examine their data and highlight potential financial opportunities. A measurable benefit from this was that the trauma and orthopaedic team were surprised to find 50% of local activity was being carried out by other providers outside of the main NHS teaching hospital. Consequently, referral to treatment times were targeted by the orthopaedic team working with radiology to develop a one-stop outpatient service where patients referred by their GPs can have their imaging and initial discussions on one day – and receive a same-day decision about whether they are suitable for surgery.

• Sherwood Forest Hospitals NHS Foundation Trust predominantly use HED for market share and benchmarking of length of stay, day case rates, readmissions, DNA rates and new to follow up ratios against peers. HED is also used to provide assurances around the HSMR/SHMI and highlight areas of focus to the Mortality Surveillance group. Specific measurable benefits have been the delivery of information to service line meetings and to the business unit. This has been used to set targets, interpret meaningful information and provide context by measuring against local peers and nationally – which the trust had not otherwise been able to do for a number of years. Furthermore, HED delivers the ability for this information to be available to all staff groups – without the reliance on the individual trust’s Information Team.

• One acute trust specifically used HED to investigate their performance of non-elective maternity re-admissions within 42 days of delivery as raised by the CQC. Analysis of clinical coding using HED identified that they had poor recording practices and evidenced that their subsequent change in coding more accurately represented their re-admission rate.

• A specialist acute hospital was finding it difficult to analyse market share for one of their sites. As a specialist provider they required a specific set of hospitals to benchmark against. Using HED they have been able to isolate the Trusts and sites, so they are true peers. HED has helped them drive up their market share, thus promoting competition between hospitals, which in turn drives better care and service across the geographical area.

This key reporting and case note review process is used by hospitals up and down the country, most commonly in the areas of mortality and readmissions management. Without using the HES data it would be impossible to deliver these sorts of benefits to the NHS.

The yielded benefit of HED is the continual feedback process to Trusts that, through benchmarking, enables best practice, clinical excellence, efficiency and financial savings. Most outcomes are intertwined with other ongoing hospital monitoring programmes and initiatives.

The objective is to provide quality and benchmarking analysis that will enable NHS organisations to deliver better services for patients.

Such analysis is solely provided either: via the online Healthcare Evaluation Data (HED) tool, or via bespoke reports.

This work is commissioned and funded on an ongoing basis by the data controller, University Hospitals Birmingham NHS Foundation Trust (UHB) and produced by the data processor, Health Informatics Department within the Hospital.

The sole objective of this work it is to support both UHB and other NHS Trusts and commissioners in the ongoing monitoring of clinical quality and organisational effectiveness. This purpose is fulfilled either:

a) Directly, the NHS Trust holds a subscription to use the HED system,

b) Indirectly, analytics are provided via a non-NHS organisation, e.g.. Pricewaterhouse Coopers (PwC), who hold a subscription to use aggregate small number suppressed data within the system only with NHS organisations. UHB has reviewed the NHS standard (ISB 1523) relating to anonymisation and can confirm the systems are compliant with this.

c) through clinically-led bespoke reports based on retrospective, pseudonymised data which if requested are published in peer review journals with small numbers suppressed.

The GDPR basis is:

a. Article 6 (lawfulness of processing):

(e) Public task: the processing is necessary for you to perform a task in the public interest or for your official functions, and the task or function has a clear basis in law.

b. Article 9 (processing of special categories of personal data):

(i) processing is necessary for reasons of public interest in the area of public health, such as protecting against serious cross-border threats to health or ensuring high standards of quality and safety of health care and of medicinal products or medical devices, on the basis of Union or Member State law which provides for suitable and specific measures to safeguard the rights and freedoms of the data subject, in particular professional secrecy.

All subscriber organisations are NHS Organisations, except for two.

The two non-NHS organisations are PwC and Carnall Farrar.

These non-NHS organisations work within the healthcare space and have access to the system solely for the purpose of assisting NHS organisations. They will only access anonymous data, i.e. aggregate level data with small numbers suppressed in line with HES Analysis guide.

For any future non-NHS subscribers, UHB would seek permission from NHS Digital first and update the DSA accordingly.

NHS organisations are limited to the below list only. The geographical range of these organisations spread across England, and are not concentrated in one region.

• NHS Trusts (between 50 to 75 hospitals)

• Clinical Commissioning Groups (CCG - Less than 10)

• Commissioning Support Units (CSU - Less than 10)

• NHS Improvement

• NHS England

• Quality Observatories (QO - Less than 10)

AQUA – legal entity is under Salford Royal NHS Foundation Trust

NEQOS – legal entity is under Northumberland Tyne and Wear and South Tees NHS Foundation Trusts

• Non-NHS Organisations - (PwC and Carnall Farrar)

Identifiable patient data is required for all England, for two reasons. Firstly, because when performing analyses, organisations need to be able to select peers based on case mix for more accurate benchmarking to assist with service improvement and these organisations may not be local or regional based. This is true for both UHB and customers of the UHB HED analytical tool. As an example, some HED customers are regional providers of a specific service and others are specialised hospitals which require benchmarking across the whole of the UK. Secondly, the current client base for the HED tool is spread across multiple regions of England and therefore require access to their own data. Further, data for all organisations is required to allow for valid peers to be selected for a given service and indicator. National data is also required to allow the calculation of standardised metric. The years of data required allow for organisations to perform multiple functions, such as being able to demonstrate service improvement over time and visualize trends.

To address the GDPR principle of data minimisation, UHB have only requested data that is deemed necessary to achieve the purpose described within this Agreement. Identifiable data is deleted each month for the period in question. UHB do not hold identifiable data older than 3 years.

To deliver the stated objective a wide range of healthcare indicators are calculated (over 100) and as such various HES data sets including Admitted Patient Care, Outpatients, Accident & Emergency as well as linked mortality data are required. This includes access to the Emergency Care Data Set (ECDS). The objective in having as wide an array of relevant indicators as possible is to give NHS managers and clinicians as complete a picture of hospital performance as possible. Therefore, the whole dataset is needed and cannot be compressed to certain fields.

HES-Civil registration mortality linked data specifically will be used within this work to look at outcomes analysis and form analytical overviews relating to post-discharge mortality. Such overviews relate to standardised post-discharge mortality monitoring within distinct clinical cohorts and bespoke long-term survival monitoring. This work will increase the understanding of complete pathways of care. Any analysis produced using Civil Registration mortality data will not be made available to non-NHS organisations.

Local patient identifier (LOAPTID) is required within the HED system for direct patient care and there is a legal basis in place under Section 251 of the National Health Service Act 2006. Subscribers will only be able to access this record level Identifiable data for their own patient records (i.e. patients within their Trust).

HED delivers a national benchmarking system that can provide assurance to hospitals they are providing safe and high-quality care and treatment, or signpost areas of concern. The HED system enables trusts to easily identify particular patient cohorts that are statistical outliers and warrant further examination. Local patient identifiers are an essential requirement linking areas that require investigation, to hospital records. Without them trusts would not be able to identify their patients and conduct root-cause analyses both for internal governance and also to provide assurance to external regulatory authorities across a range of key indicators i.e. HSMR, SHMI. Historically, the Secretary of State for Health had ordered a review of avoidable deaths. Ensuring trusts have access to their local patient identifiers through HED enables them to conduct these audits and therefore focuses attention on eradicating mortalities that could have been prevented.

HED does not just include HES and Mortality datasets. Other data sources used within the HED system are Healthcare Resource Group (HRG) National Tariff, Venous Thromboembolism Risk Assessment data, publicly available NHS England datasets, National Reporting and Learning System (NRLS) patient safety incident reports, PROMS, Safety Thermometer and Infection Control. There is no international data within HED. Furthermore, HES/ECDS/Civil Registration mortality data is not linked to international data and is not used outside of England/Wales.

It is essential for root-cause analysis that patients can be considered on a case-by-case basis. The ability to be able to identify patients via HED and then subsequently interact with other datasets and clinical notes held locally is vital to detect required clinical quality improvements. A specific example of this is via Mortality reviews, where HED directly enables organisations to monitor and manage services so that no avoidable harm comes to patients whilst in their care.

HED specifically empowers healthcare managers and clinicians to measure patient experience and outcomes benchmarked against their peers (both local and nationally) e.g. Length of stay, Mortality, Readmissions. This information is not available locally and delivers clinically relevant outcome data and comparative information to clinicians. Access to local patient identifiers is critical to enable health care professionals to audit their data and clinical practice. This review of individual patient outcomes and experience can evidence the care provided is of a good quality and safe, and also provide assurance to trust boards. In addition, access to identifiers will also enable clinicians to review and audit deaths attributed to them in national mortality models.

Record-level patient information is only available to organisations who deliver the care. A protocol including Caldicott authorisation form has been reviewed previously by DAAG (DAAG reference: 240412-a) for controlling access to such sensitive items. This established process ensures that access to sensitive items is restricted to authorised hospital trust staff and was found to be robust during a recent NHS Digital Audit.

The rationale for allowing PwC access to aggregate-level small-number suppressed analytics:

As some NHS organisations require additional specialist resource to deliver the benefits of using benchmarking information, therefore subscription to the HED tool is required by PwC as:

1. This enables PwC to have people equipped to provide immediate support to NHS organisations.

2. Providing PwC with aggregate-level information via the tool is the most efficient way of disseminating information in support of this work – the alternative described directly below would clearly create large inefficiencies.

3. It allows such organisations to be autonomous in undertaking work that requires a level of independence and is beneficial to the NHS e.g. the Keogh review. In this instance PwC were commissioned to complete this review independently of any engagement of NHS Trusts involved. It would therefore have been inappropriate for them to ask the Hospitals for the information required to undertake this review.

The rationale for allowing Carnall Farrar access to aggregate-level small-number-supressed analytics:

Carnall Farrar is a management consultancy dedicated to improving health, care and public services. They identify where clinical improvements can be made and the best ways to make change happen. This leads to better use of the scarce NHS resources of workforce, facilities and beds. Carnall Farrar’s objective is to look for clinical solutions to streamline and improve patient care, allowing more people access to a quality service.

Historically, NHS clients of Carnall Farrar have to meet the costs associated with collecting, processing and analysing that data. The ability for Carnall Farrar to access analysis directly within HED would streamline the data collection process and save time, therefore lowering the cost for the NHS organisation and saving NHS management teams the time it takes to respond to data requests. The ability to undertake such clinical improvement projects allows patient care to be streamlined and improved so allowing more people to receive the best quality clinical care.

Please note: The level of access given to PwC and Carnall Farrar will be the same i.e. same levels of security and censoring.

The alternative would be for NHS organisations working with PwC and Carnall Farrar to provide data directly to the non-NHS organisations. This arrangement would have the following detrimental effect on the NHS:

1. It would actively be encouraging NHS-organisations to export data at varying levels (while only aggregate-level small-number-supressed analytics will be provided) from the HED system and send it to non-NHS organisations. By these two non-NHS organisations having direct access UHB can monitor which modules are accessed when and by whom. This negates the need for NHS organisations to export isolated aggregate data and email it outside the NHS.

2. It would introduce a longer lead time for projects which would ultimately cost the NHS more.

The sole outputs are benchmarked or standardised healthcare indicators such as measures of mortality, survival, discharge and admission trends, readmissions, length of stay, patient safety etc.

Such outputs are solely provided either via:

• the range of Dashboards and Modules made available within the HED system, or

• aggregate small number censored reports

Within the HED system:

Dashboards will only contain aggregate level information. Modules can contain aggregate level information and low-level information. As explained, the level of data that can be viewed within a module depends on the access level of the named individual user and which organisation they are working for. As such, access to low-level information, including small numbers, is strictly controlled in line with the access controls outlined in the above section on ‘Processing activities’.

Outputs are to be used solely for the purpose of assisting the NHS.

Outputs will be used by NHS Clinicians and Managers to:

• Assure and manage clinical quality and patient safety within NHS Organisations

• Identify trends requiring a clinical review of patient pathways. (Hospital based users with Caldicott approval are able to investigate nationally standardised metrics and ‘drill-down’ to patient level information for patients treated at that trust, including local patient identifiers, in order to conduct clinical case note review and route cause analysis)

• Increase the understanding of patient outcomes

• Identify potential areas for improvement in clinical quality or operational efficiency either within a Hospital or a local healthcare economy

• Identify areas of best practice either within hospital trusts or local healthcare economies

• Provide consultants with the information necessary for consultant revalidation

All of the above will serve to increase the understanding of patient outcomes in regard to quality, safety, productivity and efficiency benchmarking within the NHS.

As mentioned, aggregate-level small-number-suppressed analytics produced for the HED tool are being made available to two non-NHS organisations.

In the case of PwC, UHB are working with this subscriber to support NHS organisations. Some examples of the specific outcomes of this access are:

• Aggregate-level analytics have been used to undertake due diligence for both Monitor and CQC. A good example being production of the Keogh review information packs. This information was vital to the extensive work undertaken as part of the Keogh review which has culminated in the majority of the hospitals originally identified being taken out of special measures by the CQC.

• A further example is use of aggregate level analytics on elderly care readmissions with a CCG in order to support work helping them to understand how they could reduce avoidable readmissions within their region.

In the case of Carnall Farrar, the specific output is continuation of the development and implementation of local solutions that address long-standing challenges within clinical service delivery and the achievement of financial targets. This support is provided to STPs, hospital trusts and CCGs.

It is not appropriate to set a target date for this work as the work streams are ongoing and the outputs already form part of various NHS organisations’ monthly reporting and governance assurance processes.

Each month a range of modules are produced which allow NHS hospitals to identify areas of potential concern within their organisation. Such areas are then investigated further. When appropriate & authorised, the ability to identify a cohort of patients within the organisation and undertake appropriate clinical case note review is invaluable as a part of ensuring good hospital governance.

• University Hospitals Birmingham (UHB) NHS Foundation Trust’s Clinical Quality Monitoring Group use the HED CUSUM model as a continual assessment of mortality. Diagnosis groups are identified that demonstrate persistent deviation from the expected mortality levels and reach a pre-set trigger point. Consequently, the trust approach to mortality is proactive rather than reactive and enables UHB to review the deaths and share the outcome with CQC considerably before the official regulatory notification. Former NHS Medical Director Sir Bruce Keogh described mortality statistics as “smoke signals” for potential problems with care. Using Statistical Process Control (SPC) techniques, these signals are only generated by ‘out-of control’ organisations, both significantly higher or lower than expected. As such, the benefit is derived from the ability to effectively use the numbers to highlight and investigate potential areas for further clinical review.

• George Eliot Hospital NHS Trust have used HED mortality modules to undertake root cause analysis of patients which has ultimately seen them move from special measures to a CQC rating of ‘Good’.

• Royal Liverpool and Broadgreen University Hospital NHS Trust uses HED to report benchmarking and mortality to their Trust Board. A measurable benefit they have observed, and that is echoed in other trusts, is the significant amount of time they have saved with HED automatically producing these reports for them, freeing their time for other hospital priorities.

• Calderdale and Huddersfield have used HED to look at their clinical coding to evidence areas of improvement in data quality and ensure they are not over- or under-coding compared to local and national peers. The system has provided reassurance to the Calderdale and Huddersfield NHS Outpatients (CHFT) Exec Board and Local Commissioning Groups while providing benchmarking insight into coding within different specialty areas.

• HED provides direct intelligent feedback to service areas that can change hospital performance for the better. However, this is not the only tool that will influence behaviour in a hospital so it is difficult to attribute changes solely to HED. As a marker of the sizeable impact it would have were HED not able to provide this information, Derby Teaching Hospitals NHS Foundation Trust, Royal Cornwall Hospital NHS Trust, Bradford Teaching Hospitals NHS Foundation Trust and The Royal Wolverhampton NHS Trust are just a few examples of the many trusts that have contacted HED over the last few months unable to access LOPATID information for case note reviews. Hospitals who are unable to access LOPATID have noticed instantly the impact this missing data is having on their ability to interrogate and improve their trust’s mortality levels.

• Plymouth Hospitals NHS Trust have used HED to generate Service Line Benchmarking packs for clinicians and service managers to help inform decisions about transformation. The packs bring together data on finances, operational efficiency, safety and quality indicators benchmarked against relevant peers. This allows service lines to examine their data and highlight potential financial opportunities. A measurable benefit from this was that the trauma and orthopaedic team were surprised to find 50% of local activity was being carried out by other providers outside of the main NHS teaching hospital. Consequently, referral to treatment times were targeted by the orthopaedic team working with radiology to develop a one-stop outpatient service where patients referred by their GPs can have their imaging and initial discussions on one day – and receive a same-day decision about whether they are suitable for surgery.

• Sherwood Forest Hospitals NHS Foundation Trust predominantly use HED for market share and benchmarking of length of stay, day case rates, readmissions, DNA rates and new to follow up ratios against peers. HED is also used to provide assurances around the HSMR/SHMI and highlight areas of focus to the Mortality Surveillance group. Specific measurable benefits have been the delivery of information to service line meetings and to the business unit. This has been used to set targets, interpret meaningful information and provide context by measuring against local peers and nationally – which the trust had not otherwise been able to do for a number of years. Furthermore, HED delivers the ability for this information to be available to all staff groups – without the reliance on the individual trust’s Information Team.

• One acute trust specifically used HED to investigate their performance of non-elective maternity re-admissions within 42 days of delivery as raised by the CQC. Analysis of clinical coding using HED identified that they had poor recording practices and evidenced that their subsequent change in coding more accurately represented their re-admission rate.

• A specialist acute hospital was finding it difficult to analyse market share for one of their sites. As a specialist provider they required a specific set of hospitals to benchmark against. Using HED they have been able to isolate the trusts and sites so they are true peers. HED has helped them drive up their marketshare, thus promoting competition between hospitals, which in turn drives better care and service across the geographical area.

This key reporting and case note review process is used by hospitals up and down the country, most commonly in the areas of mortality and readmissions management. Without using the HES data it would be impossible to deliver these sorts of benefits to the NHS.

The yielded benefit of HED is the continual feedback process to trusts that, through benchmarking, enables best practice, clinical excellence, efficiency and financial savings. Most outcomes are intertwined with other ongoing hospital monitoring programmes and initiatives.

The objective is to provide quality and benchmarking analysis that will enable NHS organisations to deliver better services for patients.

Such analysis is solely provided either: via the online Healthcare Evaluation Data (HED) tool, or via bespoke reports.

This work is commissioned and funded on an ongoing basis by the data controller, University Hospitals Birmingham NHS Foundation Trust (UHB) and produced by the data processor, Health Informatics Department within the Hospital.

The sole objective of this work it is to support both UHB and other NHS Trusts and commissioners in the ongoing monitoring of clinical quality and organisational effectiveness. This purpose is fulfilled either:

a) Directly, the NHS Trust holds a subscription to use the HED system,

b) Indirectly, analytics are provided via a non-NHS organisation, e.g.. Pricewaterhouse Coopers (PwC), who hold a subscription to use aggregate small number suppressed data within the system only with NHS organisations. UHB has reviewed the NHS standard (ISB 1523) relating to anonymisation and can confirm the systems are compliant with this.

c) through clinically-led bespoke reports based on retrospective, pseudonymised data which if requested are published in peer review journals with small numbers suppressed.

The GDPR basis is:

a. Article 6 (lawfulness of processing):

(e) Public task: the processing is necessary for you to perform a task in the public interest or for your official functions, and the task or function has a clear basis in law.

b. Article 9 (processing of special categories of personal data):

(i) processing is necessary for reasons of public interest in the area of public health, such as protecting against serious cross-border threats to health or ensuring high standards of quality and safety of health care and of medicinal products or medical devices, on the basis of Union or Member State law which provides for suitable and specific measures to safeguard the rights and freedoms of the data subject, in particular professional secrecy.

All subscriber organisations are NHS Organisations, except for two.

The two non-NHS organisations are PwC and Carnall Farrar.

These non-NHS organisations work within the healthcare space and have access to the system solely for the purpose of assisting NHS organisations. They will only access anonymous data, i.e. aggregate level data with small numbers suppressed in line with HES Analysis guide.

For any future non-NHS subscribers, UHB would seek permission from NHS Digital first and update the DSA accordingly.

NHS organisations are limited to the below list only. The geographical range of these organisations spread across England, and are not concentrated in one region.

• NHS Trusts (between 50 to 75 hospitals)

• Clinical Commissioning Groups (CCG - Less than 10)

• Commissioning Support Units (CSU - Less than 10)

• NHS Improvement

• NHS England

• Quality Observatories (QO - Less than 10)

AQUA – legal entity is under Salford Royal NHS Foundation Trust

NEQOS – legal entity is under Northumberland Tyne and Wear and South Tees NHS Foundation Trusts

• Non-NHS Organisations - (PwC and Carnall Farrar)

Data is required for all England, for two reasons. Firstly, because when performing analyses, organisations need to be able to select peers based on case mix for more accurate benchmarking to assist with service improvement and these organisations may not be local or regional based. This is true for both UHB and customers of the UHB HED analytical tool. As an example, some HED customers are regional providers of a specific service and others are specialised hospitals which require benchmarking across the whole of the UK. Secondly, the current client base for the HED tool is spread across multiple regions of England and therefore require access to their own data. Further, data for all organisations is required in order to allow for valid peers to be selected for a given service and indicator. National data is also required in order to allow the calculation of standardised metric. The years of data required allow for organisations to perform multiple functions, such as being able to demonstrate service improvement over time and visualize trends.

To deliver the stated objective a wide range of healthcare indicators are calculated (over 100) and as such various HES data sets including Admitted Patient Care, Outpatients, Accident & Emergency as well as linked mortality data are required. This includes access to the Emergency Care Data Set (ECDS). The objective in having as wide an array of relevant indicators as possible is to give NHS managers and clinicians as complete a picture of hospital performance as possible. Therefore, the whole dataset is needed and cannot be compressed to certain fields.

HES-Civil registration mortality linked data specifically will be used within this work to look at outcomes analysis and form analytical overviews relating to post-discharge mortality. Such overviews relate to standardised post-discharge mortality monitoring within distinct clinical cohorts and bespoke long-term survival monitoring. This work will increase the understanding of complete pathways of care. Any analysis produced using Civil Registration mortality data will not be made available to non-NHS organisations.

Local patient identifier (LOAPTID) is required within the HED system for direct patient care and there is a legal basis in place under Section 251 of the National Health Service Act 2006. Subscribers will only be able to access this record level Identifiable data for their own patient records (i.e. patients within their Trust).

HED delivers a national benchmarking system that can provide assurance to hospitals they are providing safe and high-quality care and treatment, or signpost areas of concern. The HED system enables trusts to easily identify particular patient cohorts that are statistical outliers and warrant further examination. Local patient identifiers are an essential requirement linking areas that require investigation, to hospital records. Without them trusts would not be able to identify their patients and conduct root-cause analyses both for internal governance and also to provide assurance to external regulatory authorities across a range of key indicators i.e. HSMR, SHMI. Historically, the Secretary of State for Health had ordered a review of avoidable deaths. Ensuring trusts have access to their local patient identifiers through HED enables them to conduct these audits and therefore focuses attention on eradicating mortalities that could have been prevented.

HED does not just include HES and Mortality datasets. Other data sources used within the HED system are Healthcare Resource Group (HRG) National Tariff, Venous Thromboembolism Risk Assessment data, publicly available NHS England datasets, National Reporting and Learning System (NRLS) patient safety incident reports, PROMS, Safety Thermometer and Infection Control. There is no international data within HED. Furthermore, HES/ECDS/Civil Registration mortality data is not linked to international data and is not used outside of England/Wales.

It is essential for root-cause analysis that patients can be considered on a case-by-case basis. The ability to be able to identify patients via HED and then subsequently interact with other datasets and clinical notes held locally is vital to detect required clinical quality improvements. A specific example of this is via Mortality reviews, where HED directly enables organisations to monitor and manage services so that no avoidable harm comes to patients whilst in their care.

HED specifically empowers healthcare managers and clinicians to measure patient experience and outcomes benchmarked against their peers (both local and nationally) e.g. Length of stay, Mortality, Readmissions. This information is not available locally and delivers clinically relevant outcome data and comparative information to clinicians. Access to local patient identifiers is critical to enable health care professionals to audit their data and clinical practice. This review of individual patient outcomes and experience can evidence the care provided is of a good quality and safe, and also provide assurance to trust boards. In addition, access to identifiers will also enable clinicians to review and audit deaths attributed to them in national mortality models.

Record-level patient information is only available to organisations who deliver the care. A protocol including Caldicott authorisation form has been reviewed previously by DAAG (DAAG reference: 240412-a) for controlling access to such sensitive items. This established process ensures that access to sensitive items is restricted to authorised hospital trust staff and was found to be robust during a recent NHS Digital Audit.

The rationale for allowing PwC access to aggregate-level small-number suppressed analytics:

As some NHS organisations require additional specialist resource to deliver the benefits of using benchmarking information, therefore subscription to the HED tool is required by PwC as:

1. This enables PwC to have people equipped to provide immediate support to NHS organisations.

2. Providing PwC with aggregate-level information via the tool is the most efficient way of disseminating information in support of this work – the alternative described directly below would clearly create large inefficiencies.

3. It allows such organisations to be autonomous in undertaking work that requires a level of independence and is beneficial to the NHS e.g. the Keogh review. In this instance PwC were commissioned to complete this review independently of any engagement of NHS Trusts involved. It would therefore have been inappropriate for them to ask the Hospitals for the information required to undertake this review.

The rationale for allowing Carnall Farrar access to aggregate-level small-number-supressed analytics:

Carnall Farrar is a management consultancy dedicated to improving health, care and public services. They identify where clinical improvements can be made and the best ways to make change happen. This leads to better use of the scarce NHS resources of workforce, facilities and beds. Carnall Farrar’s objective is to look for clinical solutions to streamline and improve patient care, allowing more people access to a quality service.

Historically, NHS clients of Carnall Farrar have to meet the costs associated with collecting, processing and analysing that data. The ability for Carnall Farrar to access analysis directly within HED would streamline the data collection process and save time, therefore lowering the cost for the NHS organisation and saving NHS management teams the time it takes to respond to data requests. The ability to undertake such clinical improvement projects allows patient care to be streamlined and improved so allowing more people to receive the best quality clinical care.

Please note: The level of access given to PwC and Carnall Farrar will be exactly the same i.e. same levels of security and censoring.

The alternative would be for NHS organisations working with PwC and Carnall Farrar to provide data directly to the non-NHS organisations. This arrangement would have the following detrimental effect on the NHS:

1. It would actively be encouraging NHS-organisations to export data at varying levels (while only aggregate-level small-number-supressed analytics will be provided) from the HED system and send it to non-NHS organisations. By these two non-NHS organisations having direct access UHB are able to monitor which modules are accessed when and by whom. This negates the need for NHS organisations to export isolated aggregate data and email it outside the NHS.

2. It would introduce a longer lead time for projects which would ultimately cost the NHS more.

The sole outputs are benchmarked or standardised healthcare indicators such as measures of mortality, survival, discharge and admission trends, readmissions, length of stay, patient safety etc.

Such outputs are solely provided either via:

• the range of Dashboards and Modules made available within the HED system, or

• aggregate small number censored reports

Within the HED system:

Dashboards will only contain aggregate level information. Modules can contain aggregate level information and low-level information. As explained, the level of data that can be viewed within a module depends on the access level of the named individual user and which organisation they are working for. As such, access to low-level information, including small numbers, is strictly controlled in line with the access controls outlined in the above section on ‘Processing activities’.

Outputs are to be used solely for the purpose of assisting the NHS.

Outputs will be used by NHS Clinicians and Managers to:

• Assure and manage clinical quality and patient safety within NHS Organisations

• Identify trends requiring a clinical review of patient pathways. (Hospital based users with Caldicott approval are able to investigate nationally standardised metrics and ‘drill-down’ to patient level information for patients treated at that trust, including local patient identifiers, in order to conduct clinical case note review and route cause analysis)

• Increase the understanding of patient outcomes

• Identify potential areas for improvement in clinical quality or operational efficiency either within a Hospital or a local healthcare economy

• Identify areas of best practice either within hospital trusts or local healthcare economies

• Provide consultants with the information necessary for consultant revalidation

All of the above will serve to increase the understanding of patient outcomes in regard to quality, safety, productivity and efficiency benchmarking within the NHS.

As mentioned, aggregate-level small-number-suppressed analytics produced for the HED tool are being made available to two non-NHS organisations.

In the case of PwC, UHB are working with this subscriber to support NHS organisations. Some examples of the specific outcomes of this access are:

• Aggregate-level analytics have been used to undertake due diligence for both Monitor and CQC. A good example being production of the Keogh review information packs. This information was vital to the extensive work undertaken as part of the Keogh review which has culminated in the majority of the hospitals originally identified being taken out of special measures by the CQC.

• A further example is use of aggregate level analytics on elderly care readmissions with a CCG in order to support work helping them to understand how they could reduce avoidable readmissions within their region.

In the case of Carnall Farrar, the specific output is continuation of the development and implementation of local solutions that address long-standing challenges within clinical service delivery and the achievement of financial targets. This support is provided to STPs, hospital trusts and CCGs.

It is not appropriate to set a target date for this work as the work streams are ongoing and the outputs already form part of various NHS organisations’ monthly reporting and governance assurance processes.

Each month a range of modules are produced which allow NHS hospitals to identify areas of potential concern within their organisation. Such areas are then investigated further. When appropriate & authorised, the ability to identify a cohort of patients within the organisation and undertake appropriate clinical case note review is invaluable as a part of ensuring good hospital governance.

• University Hospitals Birmingham (UHB) NHS Foundation Trust’s Clinical Quality Monitoring Group use the HED CUSUM model as a continual assessment of mortality. Diagnosis groups are identified that demonstrate persistent deviation from the expected mortality levels and reach a pre-set trigger point. Consequently, the trust approach to mortality is proactive rather than reactive and enables UHB to review the deaths and share the outcome with CQC considerably before the official regulatory notification. Former NHS Medical Director Sir Bruce Keogh described mortality statistics as “smoke signals” for potential problems with care. Using Statistical Process Control (SPC) techniques, these signals are only generated by ‘out-of control’ organisations, both significantly higher or lower than expected. As such, the benefit is derived from the ability to effectively use the numbers to highlight and investigate potential areas for further clinical review.

• George Eliot Hospital NHS Trust have used HED mortality modules to undertake root cause analysis of patients which has ultimately seen them move from special measures to a CQC rating of ‘Good’.

• Royal Liverpool and Broadgreen University Hospital NHS Trust uses HED to report benchmarking and mortality to their Trust Board. A measurable benefit they have observed, and that is echoed in other trusts, is the significant amount of time they have saved with HED automatically producing these reports for them, freeing their time for other hospital priorities.

• Calderdale and Huddersfield have used HED to look at their clinical coding to evidence areas of improvement in data quality and ensure they are not over- or under-coding compared to local and national peers. The system has provided reassurance to the Calderdale and Huddersfield NHS Outpatients (CHFT) Exec Board and Local Commissioning Groups while providing benchmarking insight into coding within different specialty areas.

• HED provides direct intelligent feedback to service areas that can change hospital performance for the better. However, this is not the only tool that will influence behaviour in a hospital so it is difficult to attribute changes solely to HED. As a marker of the sizeable impact it would have were HED not able to provide this information, Derby Teaching Hospitals NHS Foundation Trust, Royal Cornwall Hospital NHS Trust, Bradford Teaching Hospitals NHS Foundation Trust and The Royal Wolverhampton NHS Trust are just a few examples of the many trusts that have contacted HED over the last few months unable to access LOPATID information for case note reviews. Hospitals who are unable to access LOPATID have noticed instantly the impact this missing data is having on their ability to interrogate and improve their trust’s mortality levels.

• Plymouth Hospitals NHS Trust have used HED to generate Service Line Benchmarking packs for clinicians and service managers to help inform decisions about transformation. The packs bring together data on finances, operational efficiency, safety and quality indicators benchmarked against relevant peers. This allows service lines to examine their data and highlight potential financial opportunities. A measurable benefit from this was that the trauma and orthopaedic team were surprised to find 50% of local activity was being carried out by other providers outside of the main NHS teaching hospital. Consequently, referral to treatment times were targeted by the orthopaedic team working with radiology to develop a one-stop outpatient service where patients referred by their GPs can have their imaging and initial discussions on one day – and receive a same-day decision about whether they are suitable for surgery.

• Sherwood Forest Hospitals NHS Foundation Trust predominantly use HED for market share and benchmarking of length of stay, day case rates, readmissions, DNA rates and new to follow up ratios against peers. HED is also used to provide assurances around the HSMR/SHMI and highlight areas of focus to the Mortality Surveillance group. Specific measurable benefits have been the delivery of information to service line meetings and to the business unit. This has been used to set targets, interpret meaningful information and provide context by measuring against local peers and nationally – which the trust had not otherwise been able to do for a number of years. Furthermore, HED delivers the ability for this information to be available to all staff groups – without the reliance on the individual trust’s Information Team.

This key reporting and case note review process is used by hospitals up and down the country, most commonly in the areas of mortality and readmissions management. Without using the HES data it would be impossible to deliver these sorts of benefits to the NHS.